Professional Documents
Culture Documents
PROGRESS IN OPTICS
EDITORIAL ADVISORY BOARD
PROGRESS IN OPTICS
Edited by
E. WOLF
University of Rochester, NY, USA
Contributors
Ayman Alfalou, Mario Bertolotti, Fabio Bovino,
Christian Brosseau, Natalie A. Cartwright,
Rafal Demkowicz-Dobrzanski, Marcin Jarzyna,
Brandon A. Kemp, Jan Koodynski, Yuri A. Obod,
Kurt E. Oughstun, Alexander B. Shvartsburg,
Concita Sibilia, Oleg D. Volpian
No part of this publication may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, recording, or any information storage and
retrieval system, without permission in writing from the publisher. Details on how to seek
permission, further information about the Publishers permissions policies and our
arrangements with organizations such as the Copyright Clearance Center and the Copyright
Licensing Agency, can be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by
the Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and
experience broaden our understanding, changes in research methods, professional practices,
or medical treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in
evaluating and using any information, methods, compounds, or experiments described
herein. In using such information or methods they should be mindful of their own safety and
the safety of others, including parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors,
assume any liability for any injury and/or damage to persons or property as a matter of
products liability, negligence or otherwise, or from any use or operation of any methods,
products, instructions, or ideas contained in the material herein.
ISBN: 978-0-12-802284-9
ISSN: 0079-6638
Contributors ix
Preface xi
Acknowledgement xiii
v
vi Contents
1. Introduction 437
2. Classical Electrodynamics 440
3. Optical Force, Momentum, and Stress 452
4. Momentum and Stress in Dispersive Media 467
5. Discussion 472
Appendix A. Scattering by a Cylinder 478
Appendix B. Scattering by a Sphere 481
Acknowledgments 485
References 485
Ayman Alfalou
Vision ISEN-BREST Laboratory L@BISEN, Institut Superieur de lElectronique et du
Numerique, Brest, France
Mario Bertolotti
Dipartimento SBAI, Universita` di Roma La Sapienza, Rome, Italy
Fabio Bovino
Quantum Technologies Laboratory, SELEX-ES, Genoa, Italy
Christian Brosseau
Laboratoire des Sciences et Techniques de lInformation, de la Communication et de la
Connaissance, Universite de Brest, Brest, France
Natalie A. Cartwright
Department of Mathematics, State University of New York, New Paltz, New York, USA
Rafal Demkowicz-Dobrza nski
Faculty of Physics, University of Warsaw, ul. Pasteura 5, Warszawa, Poland
Marcin Jarzyna
Faculty of Physics, University of Warsaw, ul. Pasteura 5, Warszawa, Poland
Brandon A. Kemp
College of Engineering, Arkansas State University, Jonesboro, Arkansas, USA
Jan Koody nski
Faculty of Physics, University of Warsaw, ul. Pasteura 5, Warszawa, Poland
Yuri A. Obod
R&D Company Fotron Auto, Moscow, Russia
Kurt E. Oughstun
College of Engineering and Mathematics, University of Vermont, Burlington, Vermont,
USA
Alexander B. Shvartsburg
Joint Institute for High Temperatures, Russian Academy of Sciences; Institute for Space
Researches Russian Academy of Sciences, Moscow, and Far Eastern Federal University,
Vladivostok, Russia
Concita Sibilia
Dipartimento SBAI, Universita` di Roma La Sapienza, Rome, Italy
Oleg D. Volpian
R&D Company Fotron Auto, Moscow, Russia
ix
This page intentionally left blank
PREFACE
In this 60th volume of Progress in Optics, six review articles are presented.
Chapter 1, contributed by Bertolotti, Bovino, and Sibilia, takes a histor-
ical approach to single-photon sources. They discuss photon statistics and
entangled states. Recently developed plasmonic sources and possible appli-
cations in quantum information processing are also described.
In Chapter 2, Alfalou and Brosseau discuss a variety of image processing
techniques. They compare the more traditional digital approach with newly
emerging, all optical setups. Because of their parallel nature, these can pro-
duce real-time results.
Chapter 3, by Cartwright and Oughstun, is a continuation of their article
in Volume 59. In this second part, they discuss recent developments in the
theory of pulse propagation through dispersive media. One of these is the
extension of the classic SommerfeldBrillouin theory to both the weak
and singular dispersion limits.
In Chapter 4, written by Demkowicz-Dobrza nski, Jarzyna, and
Koody nski, the authors describe how nonclassical light states can be used
to enhance the performance of optical interferometers. This is especially
important, for example, for the ongoing search for gravitational waves.
Chapter 5 is a contribution by Kemp. Starting with the Minkowski
Abraham controversy, different expressions for the electromagnetic force
density are analyzed. This ongoing discussion has direct consequences for
the growing number of optical trapping applications.
The final chapter, Chapter 6, is written by Shvartsburg, Obod, and
Volpian. They discuss how the classical effect of tunneling through optical
barriers takes on new and unexpected forms when the barrier consists
of a dielectric gradient metamaterial. Because of the scalability of these
effects, these also apply to microwave tunneling phenomena in trans-
mission lines.
It has been my honor to serve as editor of Progress in Optics for over
50 years. But, as I used to tell my children when they were young, All good
things must come to an end. I would like to thank the members of the
editorial advisory board for their help, advice, and friendship over these
many years. Beginning with Volume 61, Dr. Taco Visser will take over
xi
xii Preface
as editor of this series. I have complete confidence that he will keep readers
of Progress in Optics well informed of the most important advances being
made in the field. I wish him well.
EMIL WOLF
Rochester, NY
April 2015
ACKNOWLEDGEMENT
The editorial staff would like to extend a special thanks for the support and
expert assistance provided by Dr. Taco Visser during the compilation of
this work.
xiii
This page intentionally left blank
CHAPTER ONE
Contents
1. Introduction 1
2. Fock States 3
3. The Problem of Localizing Photons 6
4. Antibunching of Single-Photon States 7
5. Photon Statistics and Spectral Purity 9
6. The Purcell Effect and the Control of Emission of Electromagnetic Radiation 12
7. Preparation of Single-Photon States: Quantum Engineering 13
8. Realization of Single-Photon Sources 20
8.1 Strongly Attenuated Sources 21
8.2 Single-Photon Sources Emitting one Photon in a Random Way 23
8.3 Single-Photon Sources on Demand 69
9. Entangled States 70
10. Plasmonic Sources 82
11. Application to Quantum Information Processing 86
12. Conclusions 95
Addendum 95
References 96
1. INTRODUCTION
The introduction of quantization of energy to discuss the interaction
of electromagnetic radiation with matter was done in 1900 by Planck (1900a,
1900b). Einstein (1905) surmised that also free radiation had a granular struc-
ture. The name photon was later proposed by Lewis (1926). The quantization
of the electromagnetic field was made by Dirac (1926, 1927). A review of
some historical papers on the subject has been made by Keller (2007).
What a photon is, exactly nobody can say. It is the quantum represen-
tation of a mode of the electromagnetic field and is an exclusively quantum
concept. With this definition, photons have associated plane waves of def-
inite wave vector k and definite polarization s. A monochromatic wave
implies delocalization in time and space; in practice, a single photon local-
ized to some degree in time and space can be described as superposition of
monochromatic photon modes.
When Glauber (1963a, 1963b) completed the model of radiation detec-
tion, discussing from a quantum point of view the interaction of radiation
and matter, and constructed a quantum theory of coherence, a number of
interesting states of radiation received a full reconnaissance as useful and pos-
sible states: among them coherent and single-photon states are perhaps the most
interesting, together with squeezed and entangled ones. Full description of
these states may be found in many excellent textbooks, like, e.g., Mandel
and Wolf (1995).
The generation of quantum states of the radiation field started to
receive great attention from the 1980s. Single-photon states, in particular,
are studied because of possible applications in quantum communication,
quantum lithography, quantum metrology, information processing, and
quantum computing, such as quantum random-number generation,
quantum networks, secure quantum communications, and quantum cryp-
tography (see, for example, Beveratos, Brouri, et al., 2002; Cerf &
Flurasek, 2006; Dusek, Lutkenhas, & Hendrych, 2006; Gisin, Ribordy,
Tittel, & Zbinden, 2002; Gisin & Thew, 2007; Grangier & Abram, 2003;
Kilin, 2001).
For example, the security in some schemes of quantum cryptography
is based on the fact that each bit of information is coded on a single photon.
The fundamental impossibility of duplicating the complete quantum state
of a single particle (no cloning theorem; Cerf & Flurasek, 2006; Diecks,
1982; Ghirardi & Weber, 1983; Wootters & Zurek, 1982) prevents any
potential eavesdropper from intercepting the message without the receivers
noticing.
An ideal single-photon source would produce exactly one photon in a
definite quantum state, in contrast with a classical source, such as attenu-
ated laser pulses, for which the photon number follows a Poisson distribu-
tion. A more stringent request would be to have the single-photon
generation on demand, that is, at a determined time. Additional requests
could be room temperature operation, high repetition frequency, high
efficient extraction into free space or fiber, good coherence, and
Quantum State Engineering 3
2. FOCK STATES
States with a prescribed number of photons are called number states or
Fock states.
They were first introduced and discussed by Fock (1932) (see also
Faddeev, Khalfin, & Komarov, 2004).
A Fock state is strictly quantum mechanical and contains a precisely def-
inite number of quanta of field excitation; hence, its phase is completely
undefined.
As well known, the Hamiltonian for the free electromagnetic field can be
written as
H ~a{ , ~a ~a{ ~a + 1=2 (1)
4 Mario Bertolotti et al.
n~ ~a{ ~a (3)
The eigenstate of the photon number operator n
n~jni njni (4)
has a perfectly fixed photon number n. Since n is a Hermitian operator, the
number n is real.
If jni is an eigenstate of n, then ajni must be an eigenstate as well, with the
eigenvalue n 1. In fact,
n~~ajni ~a{ ~a2 jni ~a~a{ ~a ~a jni n 1~ajni (5)
In a similar way, ajni is an eigenstate of n with eigenvalue n + 1. So there
are the fundamental relations
and orthonormal
Quantum State Engineering 5
hnjn0 i nn (10)
In momentum representation, it is
o p 1=4 exp p2 =2 (12)
The expectation value of the electric field and its square are
hnjEjni 0 (13)
and
2
njE jn =o V n + 1=2 (14)
~aji ji (16)
where d2 dRedIm:
The other states of interest here are squeezed and entangled states. The real
and imaginary part of the complex amplitude of the electromagnetic field
fluctuate with equal dispersion in a coherent state. The phenomenon of vac-
uum fluctuations is a manifestation of this effect because the vacuum state is
an example of a particular coherent state. In a squeezed state, one part of the
field fluctuates less and the other part fluctuates more than in the vacuum
state. A number state can be considered a special case of a squeezed state.
Entangled states will be discussed in Section 9.
average a time a half a Rabi period1 has to elapse to have a finite probability
for the molecule to be in the excited state and emit a second photon.
Antibunching was first observed in pioneering experiments (Kimble,
Dagenais, & Mandel, 1977; Walls, 1979).
In the nanosecond time regime within a single bunch of photons, the
emitted photons from a single-quantum system are expected to show anti-
bunching; that is, the probability for two photons to arrive at the same time
is zero.
To observe antibunching correlations, the second-order correlation
function g(2)() is generally measured by determining the distribution of time
delays N() between the arrival of successive photons in a dual beam detector
g2 0 1 1=n (21)
Half-silvered mirror
Liquid filter Interference filter
Photomultiplier tube
Lens
Mercury arc Rectangular aperture Slide
Photomultiplier tube
Correlator
Amplifier Amplifier
Integrating motor
Figure 1 The Hanbury Brown and Twiss experimental arrangement.
1
The Rabi frequency j12 Ej/, where 12 is the transition dipole moment and E is the laser field
with which the atom is excited.
Quantum State Engineering 9
120
2.0
1.5
80
g(2)(t)
c(t)
1.0
40
0.5
0.0 0
20 0 20 40
t (ns)
Figure 2 An example of antibunching. From Beveratos, Brouri, Gacoin, Poizat, and
Grangier (2001).
Z
W I t0 dt0 (23)
where : : represent the normal order of the field operator, PN(W) is the
quasi-probability distribution of the normal integral intensity W, that may
have also negative values, and is the density matrix.
We briefly recall the possibility of describing the most general statistical
quantum system by means of the so-called density matrix . This is a
Hermitian operator, time independent in the Heisenberg picture, such that
the quantum and statistical expectation value of any operator A is expressed as
Quantum mechanically
hn~i ~a{ ~a
2 2 (31)
n~ < ~a{ ~a > ~a{ ~a + ~a{2 ~a2
The variance of the number of absorbed photons ( 1) is expressed by
~
n2 hn~i + W 2 N (32)
For a coherent field h(W)2iN 0 and h(n)2i n2, i.e., such a field is
Poissonian. Classical fields have h(W)2iN > 0 and they are super-Poissonian,
for example, for the chaotic field of a natural source h(W)2i hIi2T2 and
n2 hni1 + hni (33)
corresponding to the BoseEinstein distribution (28).
For quantum fields having no classical analogues, it may be h(W)2iN < 0
and h(n)2i < hni, and such light is sub-Poissonian.
The photon number distribution can be calculated using the density
matrix
pm hmjjmi (34)
From an experimental point of view p(m) is related to p(n,T,t) by the
Bernoulli transformation
X
pn, T, t m!=n!m n!n 1 mn pm (35)
Introducing the so-called Fano factor Fn n2 =hni, the fluctuation
formula (32) can be written in terms of the photocount number n and the
photon number m in the form of the quantum Burgess variance theorem
(Perina, 1984)
Fn 1 Fm 1 (36)
For the Fock state Fm 0.
Although both effects are nonclassical, are exhibited only by a quantum
field, and are often associated, sub-Poissonian statistics and antibunching are
distinct effects that need not necessarily occur together; one may occur
without the other (Singh, 1983).
However for a stationary single-photon source, the nonclassical nature of
the emitted radiation would lead to sub-Poissonian photon statistics with
g(2)(0) < 1.
12 Mario Bertolotti et al.
h2 vc jh f jHjiij2 (38)
where H is the interaction Hamiltonian and (c) is the density of states at the
transition frequency c, that for radiation in free space is
2
4vc
vc 2 (39)
c3
The spontaneous emission rate can therefore be changed if an atomic sys-
tem finds itself in a space region where (c) is modified; for example, it is
placed close to a metal surface or a dielectric interface. First experimental
demonstrations were carried out by Drexhage, Kuhn, and Schafer (1968),
Quantum State Engineering 13
Kuhn (1970), and Chance, Prock, and Silbey (1974). Approximate solutions
of the electromagnetic boundary value problem were reviewed, for exam-
ple, by Drexhage (1974). More detailed calculations are in Lukosz and Kunz
(1977a, 1977b) and references therein. In a more effective way, the radiation
emitted by a source can be altered by suitably modifying the surrounding
vacuum fields in a cavity.
The application to a small cavity for which the density of modes may be
modified was considered by Kleppner (1981). In particular, when the tran-
sition frequency c is near resonance with a mode eigenfrequency, the spon-
taneous emission rate can be considerably increased.
The effect was experimentally observed (Goy, Raimond, Gross, &
Haroche, 1983) with a sodium Rydberg atom set through a resonant sup-
erconducting cavity. Also inhibited spontaneous emission was observed by
studying the cyclotron motion of a single electron (Gabrielse & Dehmelt,
1985). Therefore, a reduction of the two-photon probability relative to a
Poisson distribution may be achieved acting on the density of modes.
Periodic dielectric structures in the form of photonic crystals can alter
the emission properties ( John, 1987; Yablonovitch, 1987). The existence
of forbidden electromagnetic frequencies inside the structure may inhibit
the emission of radiation which is, on the contrary, enhanced at the frequen-
cies where the density of modes is maximum.
Enhanced spontaneous emission by quantum semiconductor boxes in a
monolithic optical microcavity, that can be considered the ancestor of quan-
tum dots experiments, was studied, for example, by Gerard et al. (1996,
1998 and; Gerard & Gayral, 1999) (see also Bulu, Caglayan, & Ozbay,
2003). A general review of 1D photonic structures is, for example,
Bertolotti (2006). In semiconductor systems, enhanced and inhibited spon-
taneous emission from GaAs quantum well excitons was demonstrated using
a planar microcavity (Yamamoto, Machida, Horikoshi, & Igeta, 1991;
Yokoyama et al., 1990). The properties of band gap structures have been
largely studied (see, for example, Yang, Fleischhauer, & Zhu, 2003;
Yang & Zhu, 2000). The effect of nanotubes on the spontaneous emission
of a single InAs quantum dot was considered by Bleuse et al. (2011).
where o is the Rabi frequency and t is the interaction time. So, instead of
simply emitting a photon and going on its way, an excited atom in a resonant
cavity oscillates back and forth between its excited and unexcited states. The
emitted photon remains in the cavity in the vicinity of the atom and is
promptly reabsorbed. The atomcavity system oscillates between two states,
one consisting of an excited atom and no photon and the other a deexcited
atom and a photon trapped in the cavity.
The basic idea of many schemes is that measurement of the atom after its
interaction with the field in a cavity provides information about the field.
The collapse of the entangled state of the atomfield system makes the field
part of the wave function to jump into a different state from the one before
the measurement. Based on the interaction between the radiation field and
atoms, many theoretical schemes were therefore proposed for the generation
of Fock states. We consider here, without pretending to be exhaustive, a few
of them. Some of them, with suitable changes, were later experimentally
realized (see Section 8.2). Eventually, as will be shown later, the real sources
of single photons have been implemented on a much simpler and direct way
choosing with skill the emitter in solid state.
Quantum State Engineering 15
Q=f (41)
on the atoms after having passed through the cavity, the technique is prob-
abilistic and attains a successful engineered quantum state only for a specific
series of measurement outcomes.
Injection of atoms with their Zeeman electronic levels initially prepared
in specific superpositions interacting adiabatically with a cavity was proposed
(Parkins, Marte, Zoller, Carnal, & Kimble, 1995; Parkins, Marte, Zoller, &
Kimble, 1993).
A scheme based upon the same mechanism described by Vogel et al.
(1993) but employing two-photon interactions in a micromaser has been
described by Garraway, Sherman, Moya-Cessa, Knight, and Kurizki
(1994) who proposed a scheme that allows the generation of a variety of
nonclassical field states with controllable (predetermined) photon number
and phase distribution. It was based on the two-photon-resonant interaction
of a single electromagnetic field mode in a high-Q cavity with initially
excited atoms crossing the cavity sequentially (one at the time). The
sequence duration should be much shorter than the cavity lifetime. The
states are generated by selecting only those sequences wherein each atom
is measured to be in the excited state after the interaction. The field distri-
bution resulting from a sequence of N such measurements is peaked about
2N positions in the phase plane. When these peaks are chosen to overlap,
part of the field can be made to interfere giving rise to decimation of the
photon number distribution and Fock states may be generated.
At variance with the practice of measuring atoms once they came out of
the cavity, Law and Eberly (1996) discussed a method in which quantum
states of the atom (source) are manipulated during the atomfield interaction
process, i.e., inside the cavity. They presented a cavity model which allows
the vacuum state to evolve to an arbitrarily prescribed superposition of Fock
states.
Later, Law and Kimble (1997) discussed the possibility to produce a pre-
scribed sequence of single-photon pulses controlling the emission of an atom
with a -type three-level structure in an optical cavity. In 2003, the Caltech
group (Duan, Kuzmich, & Kimble, 2003) extended the study of the inter-
action of an atom trapped in a cavity with hot atoms with an inhomogeneous
distribution in position and/or a time-varying location, using an adiabatic
passage technique.
The previously considered schemes find at least two types of difficulties.
First, the quantum states of the sources of radiation (the atomic states) have
to be manipulated in an arbitrarily controllable manner. Second, the source
must teach the field to evolve toward the desirable quantum state.
Quantum State Engineering 19
Ie,0>
Energy of the atomic bare states
Atomcavity
coupling g
)
AP
TIR
Trigger
pulse Emission of a
r (S
WT single photon
sfe
tran
tic
Ig,1> Ig,0>
aba
Adi
Repumping pulse
Iu,0>
Figure 3 Scheme of the atomic levels coupled by the trigger pulse, the cavity, and a
possible repumping pulse. From Kuhn et al. (1999).
20 Mario Bertolotti et al.
Increasing the pulsed laser rate could help, but in this way also dark
counts will increase and the ratio of detected photons to dark counts
decreases with hni (see Gisin et al., 2002 for more details).
Weak chaotic attenuated sources are unsatisfactory also because if the sin-
gle photon is radiated into a large solid angle, it is quite difficult to collect it
into the optical system with good efficiency. The loss of photons increases
also if filtering in polarization is needed. However, this kind of source is
often used. As an example see the experiment for a daylight quantum key
distribution over a 1.6-km path in atmosphere, performed by Buttler
et al. (2000). At least one system exists based on this type of source
(Bovino, 2013a).
n(t)
500
400
300
200
25 50 75 100
Delay t in nanoseconds
Figure 4 The number of recorded pulse pairs n() as a function of the time delay in
nanoseconds. The growth of n() from 0 shows antibunching. The bars on one point
indicate statistical uncertainties corresponding to one standard deviation. The broken
line just outlines the trend. From Kimble et al. (1977).
Photon emission was triggered with a short excitation pulse (De Martini,
Di Giuseppe, & Marrocco, 1996) in a molecular solution of oxazine in eth-
ylene glycol.
Kitson, Jonsson, Rarity, and Tapster (1998) studied the spontaneous
emission from a thin layer of a dilute solution of fluorescent dye molecules
within an optical microcavity. They identified photon antibunching. They
however reported that true single-photon emission is limited by photo-
bleaching and Brownian motion of the molecules in and out of the excita-
tion volume. The paper discussed the disexcitation of the dye which was
modeled as a singlet and a triplet state.
Hoffges, Baldauf, Lange, and Walther (1997) measured the correlation
function g(2)(t) of fluorescence light from a single 24Mg+ ion confined in a radio
frequency trap, as a function of detuning and intensities of the incident field.
The results clearly show the antibunching effect (Figure 5). For increas-
ing intensity of the exciting laser, the function displays, as should be
expected, oscillations at the Rabi frequency (the Mollow triplet).
26 Mario Bertolotti et al.
Figure 5 Intensity correlation for three different detunings and intensities of the inci-
dent field (a, higher intensity; c, lower intensity). From Hoffges et al. (1997).
Figure 6 (a) The pertinent molecule levels, (b) chemical structures of terrylene (upper)
and p-terphenyl (lower), (c) confocal fluorescence image (10 10 m) of single terrylene
molecules excited at 1.5 W, and (d) schematic of the confocal system. From Lounis and
Moerner (2000).
28 Mario Bertolotti et al.
They used a periodic pulsed laser source (pulse width 35 ps, repetition rate
6.25 MHz). The system worked for hours without photobleaching.
The exciting source was a frequency-doubled (532 nm) mode-locked
picosecond Nd:YAG laser with a maximum average power of 0.2 mW
which pumps the four-level scheme of the molecule from the ground singlet
state to a vibriationaly excited level of the first electronic excited singlet state.
After fast (ps) relaxation to the lowest electronic excited state, the molecule
subsequently emits a single photon (at around 579 nm). Because the mole-
cule can be pumped at high intensity, the probability of preparing the emit-
ting state can approach unity. This scheme has the further advantage of
spectrally separating the laser excitation and the fluorescence emission. Sin-
gle photons were generated at predetermined times, within the accuracy of
the emission lifetime of a few nanoseconds. The pump pulse width was very
short compared to the fluorescence lifetime.
Terrylene molecules embedded in a highly transparent polymer film of
polymethyl methacrylate (PMMA) pumped with the 514 nm line of a cw
argon-ion laser were used also by Treussart, Clouqueur, Grossman, and
Roch (2001) and studied with scanning confocal microscope at room tem-
perature. The group was among the first to publish (Figure 7) the g(2)() cor-
relation function as a function of time showing the deep at its center
( 0).
Indistinguishability through a HongOuMandel experiment was dem-
onstrated by Kiraz et al. (2005) using a single terrylene diimide molecule at
1.4 K.
Treussart et al. (2002) realized also a pulsed saturation of a new single
molecule (cyanine dye DiIC18) dispersed at a concentration of about one
Figure 7 Correlation function (left scale) and counts integrated over about 100s
(right scale). From Treussart et al. (2001).
Quantum State Engineering 29
The Imamoglu and Yamamoto (1994) proposal was later realized (Kim,
Benson, Kan, & Yamamoto, 1999) obtaining single as well as multiple pho-
ton emission events with a repetition rate of 10 MHz at 50 mK. In the
device, single-electron and hole-charging energies must be large compared
to the thermal background energy to ensure single-photon emission. There-
fore, the device can only be operated at ultra-low temperatures (T < 1 K),
and sufficient collection efficiency to measure the photon correlation func-
tion was not obtained. Imamoglu considered also a photon blockade effect
in the context of cavity electromagnetically induced transparency
(Werner & Imamoglu, 1999).
Semiconductor quantum dots show promise as they can be embedded in
devices enabling electrical injection. Of course devices of this kind are much
more easy to handle than optically pumped systems. The realization of an
electrically driven source of single photons based on integrating InAs quan-
tum dots into a conventional semiconductor light-emitting diode structure
at 5 K was studied, obtaining emission at 1.3942 eV. Correlation measure-
ments demonstrated a large suppression of two-photon emission with the
lowest injection current (2 A) with g(2)(0) 0.34 (Yuan et al., 2002).
In GaAs-based LEDs, the high refractive index leads to a low critical
angle and only a small fraction of the emitted photons escape the front sur-
face. To enhance by a factor of 10 the efficiency with which the photons are
collected from an electrically driven single InAs/GaAs quantum dot,
Bennett et al. (2005a) used a planar semiconductor cavity. The device struc-
ture is shown in Figure 8.
3l cavity
Contact
12 period metal
n+ DBR
Substrate/buffer
around 34% were reported by Heindel et al. (2010). Pulsed electrical exci-
tation was used to excite single InP/Ga0.51In0.49P QDs.
Nowak et al. (2014) described a micropillar connected to a 25 25 m2
wide frame through one-dimensional wires. Combining with a diode struc-
ture, they demonstrated a deterministic and electrically tunable single-
photon source with an extraction efficiency around 53%. However, the
device worked at cryogenic temperatures.
The use of a quantum dot embedded in a photonic nanowire and elec-
trically pumped was discussed, claiming that output efficiencies of more than
80% could be achieved (Gregersen, Nielsen, Mrk, Claudon, & Gerard,
2010). Two-photon interference from two separate, independent quantum
dots electrically driven and tuned to the same energy applying a large bias
electric field was obtained by Patel et al. (2010).
Also a proposal for a method of producing single-photon states utilizing
the acoustoelectric effect was made (Foden, Talyanskii, Milburn,
Leadbeater, & Pepper, 2000).
Rubidium oven
Stark shifting
plates Auxilliary
detector
Reference laser
Velocity selecting
UV laser
State selective
field ionization of
Rydberg atoms
Figure 9 The atoms leaving the rubidium oven are excited into the 63P3/2 Rydberg state
by means of a UV laser and enter the microcavity. The cavity is tuned to the 21.456 GHz
transition from the 63P3/2 to the 61D5/2 state, which is the lower or ground state of the
maser transition. After the cavity they are detected by a state-selective field ionization.
From Brattke et al. (2001a).
2001a). Trapped states are typical of strongly coupled systems and occur
when the atomfield coupling, , and the interaction time, tint, are chosen
such that in a cavity field with n photons each atom undergoes an integer
number k of Rabi cycles or
tint n + 1 k (45)
When Equation (45) is fulfilled, the cavity photon number is left unchanged
after the interaction of an atom and hence the photon number is trapped,
regardless of the atomic pump rate.
Quantum State Engineering 35
where Pe(tint) and Pg(tint) are the probability of finding an excited state and a
ground state atom, respectively, for a particular interaction time tint. The
operation of a cw micromaser leads to Fock states, with the micromaser field
evolving into a number state, provided the atomic velocity is conveniently
tuned (Weidinger et al., 1999).
A Fock state shows ideal sub-Poissonian statistics. Consequently, under
the influence of a trapping state, the atomic statistics should also be sub-
Poissonian. As an atom in the ground state must have emitted a photon into
the cavity, the lower state atom statistics is strongly related to cavity photon
statistics. When the maser is under the influence of a trapping state, an atom
cannot emit into the cavity until a photon decays from the mode. When this
occurs, the next excited state atom entering the cavity will emit a photon
with a high probability, returning the cavity to the trapping state. This results
in a regular spacing of ground state atoms and a reduction of fluctuations in
the count rate. The FanoMandel Q function
nt 2 hnti2
Q t (47)
hnti 1
where n(t) is the number of counts in an interval of length t, which may be
used to assess the kind of statistics. If Q(t) 0 the beam is Poissonian; sub-
Poissonian statistics produces Q(t) < 0, and super-Poissonian statistics has
Q(t) > 0. Indications of one-photon and two-photon states were clearly
obtained (Weidinger et al., 1999).
Using the one-atom maser, the first experimental evidence for the oper-
ation of a reliable and robust source of photon Fock states was presented by
Brattke, Varcoe, and Walther (2001b). In the experiment, the micromaser
setup of Figure 9 was used with a beam of Rubidium atoms in Rydberg
states entering the microcavity cooled at 300 mK. The group claimed a suc-
cess rate of Fock state production of 85% (see also Brattke et al., 2003).
Using the same setup of Figure 9, Varcoe, Brattke, Weidinger, and
Walther (2000) observed the buildup of number states in a high-Q cavity
by investigating the interaction dynamics of a probe atom with the field.
With a train of atoms, arbitrary multiphoton Fock states can in principle
36 Mario Bertolotti et al.
be generated in a cavity by post selecting the events in which each atom has
emitted one photon. One- and two-photon states were generated in this
way (Varcoe et al., 2000). The states were individuated looking at the
behavior of the inversion probability (46). The number of Rabi oscillations
permitted to determine the state number. In this way, single-photon states
were obtained in a cavity in which pulsed excitation of atoms was used so
that the number of atoms passing through the cavity could be pre-
determined. The group produced number states with a purity of 99% for
the n 1 state and 95% for the n 2 state (Varcoe et al., 2000).
Also an adiabatic passage in sequence in an optical cavity relying on
STIRAP was made, using 85Rb atoms at 10 K (Hennrich, Legero,
Kuhn, & Rempe, 2000).
The work toward a single-photon source in the visible range using a sin-
gle trapped ion placed in a cavity was described by Pinkse, Fischer, Maunz,
and Rempe (2000) who reported high-finesse optical cavity experiments in
which the change in transmission induced by a single slow atom approaching
the cavity triggers an external feedback switch which traps the atom in a light
field containing about one photon on average. The source was modified
(Kuhn, Hennrich, & Rempe, 2002) still employing the emission of single
photons from a single three-level atom (Rb) strongly coupled to a high-
finesse cavity. The photons are generated by an STIRAP (see also
Hennrich et al., 2000) between two atomic ground states, with the vacuum
field of the cavity stimulating one branch of the transition, and the laser
pulses deterministically driving the other branch. Figure 10 shows the
scheme of the experiment and the pertinent 85Rb levels. A single Rb atom
is prepared in state jui and is pumped with a laser whose frequency is close to
the jui ! jei transition. Provided the pump pulse rises slowly and a large
atomcavity coupling is present, a smooth transition from state u with no
photon ju,0i to state g with one photon jg,1i is realized. This state decays
emitting a photon which leaves the cavity through the mirror. The final state
jg,0i of the coupled system is reached and a recycling laser pulse is applied
between consecutive pump pulses resonant with the jgi ! jei transition so
pumping the atom to state jei from which it decays spontaneously to the
initial state jui. The system dynamics was substantially the same as described
previously (Kuhn et al., 1999). The correlation function g(2)() was measured
finding antibunching (Figure 11). It also oscillated with the same periodicity
as the sequence of pump pulses. This indicated that photons are emitted only
during the pump pulses, and no emission occurs when recycling pulses are
Quantum State Engineering 37
Magneto-optical trap
85
Rb Correlator
e,0
2
g ()
5P3/2
85Rb atoms
Recycling laser
Pump Atom Avalanche photo diodes
laser cavity
coupling
Cavity
u,0
5S1/2
g,1 g,0 Beam splitter
Photon
emission Recycling laser
Pump laser
Figure 10 On the left the relevant energy levels of Rb. The atomic states labeled jui, jei,
and jgi are involved in the Raman process, and the states j0i and j1i denote the photon
number in the cavity. On the right side the setup. The figure is substantially the same as
from Kuhn et al. (2002) but is from Hennrich, Legero, Kuhn, and Rempe (2004).
Figure 11 The correlation function g2(t) averaged over 15,000 experimental cycles
showing antibunching. The hatched area represents correlations between photons
and detector-noise counts. From Kuhn et al. (2002).
Figure 13 Velocity-selected Rb atoms from the oven on the extreme left of the figure
are excited from lasers, interact with the superconducting cavity, and are detected in
the field ionization counter at the extreme right. From Raimond et al. (2005).
Quantum State Engineering 41
Figure 14 (a) A single Cs atom is trapped in a cavity made by two mirrors and is pumped
by external fields; (b) the relevant atomic levels; and (c) the timing sequence of pumping
pulses. From McKeever et al. (2004).
(2005). Only one atom at a time was stored in the trap; if a second atom
enters the trap, both are immediately ejected. In this regime, the atom sta-
tistics is sub-Poissonian and the trap contains either one or zero atoms. The
trapped atom was excited with a 4 ns pulse of polarized laser light at
780.2 nm at a repetition rate of 5 MHz. Fluorescence photons were pro-
duced by spontaneous emission from the upper state of the transition which
has a lifetime of 26 ns. In this way, an efficient (0.6 efficiency) triggered
source of single photons with a well-defined polarization was realized.
Photon blockade was proposed, discussed, and realized by several
authors, initially considering electrically driven devices (Imamoglu et al.,
1997; Kim et al., 1999; Smolyaninov, Zayats, Gungor, & Davis, 2002;
Werner & Imamoglu, 1999). It is possible to obtain the effect also through
photonphoton interactions in a nonlinear optical cavity using, for example,
electromagnetically induced transparency (see also Lukin & Imamoglu,
2001). Observations of photon blockade for the light transmitted by an opti-
cal cavity containing one trapped atom in the regime of strong atomcavity
coupling were reported by Birnbaum et al. (2005). Excitation of the atom
cavity system by a first photon blocks the transmission of a second photon.
Kang et al. (2011) demonstrated a high-speed controlled generation of
single photons in a coupled atomcavity system using a single 85Rb atom
pumped with a nanosecond laser. The rather complex geometry involved
dropping the Rb atoms from a magneto-optical trap into the cavity in such
a way to select a single atom and then exciting it and extract the photon from
the cavity. Rates up to 10 MHz were obtained.
42 Mario Bertolotti et al.
Figure 15 The source utilizing Rydberg atoms. (a) Sketch of the apparatus; (b) relevant
87
Rb energy levels. From Dudin and Kuzmich (2012).
The emission process is spontaneous and in most cases takes place after a
rapid optical excitation of the emitter. The main difficulty with quantum
dots is that they interact with a solid-state environment, necessitating cryo-
genic operation temperatures, and yet environment-induced decoherence is
still a problem. However, these difficulties are offset by advantages such as
being fixed in place, having large dipole moments, and allowing the possi-
bility of integration into monolithic optical microcavity structures or in pho-
tonic crystals, thereby changing the density of states of the electromagnetic
field, shortening lifetime, increasing efficiency, and allowing a better cou-
pling with the external world. In this way, quantum dots integrated in an
optical microcavity are efficient solid-state sources of single photons. At
cryogenic temperatures, they display spectrally narrow emission lines com-
patible with the control of spontaneous emission based on cavity quantum
electrodynamics and the use of the Purcell effect.
Single-photon emission from semiconductor quantum dots has been
reported for many different material systems covering different wavelengths
and maximum operation temperatures. The most studied systems have been
InAs in near infrared, InP in the red region, CdSe in the green, and GaN in
the ultraviolet.
A number of review papers have been published on the subject (see, for
example, Shields, 2007), the most recent being Buckley, Rivoire, and
Vuckovic (2012). Another excellent discussion has been done by Lodahl,
Mahmoodian, and Stobbe (2013).
First experiments were performed by Michler, Imamoglu, et al. (2000)
with CdSe/ZnS quantum dots (microcrystals) that were shown to generate
single photons in the green, at room temperature. Optical pumping was per-
formed using the circularly polarized light of the 488 nm line of a continuous
Ar+ laser. Blinking was observed and the value of the correlation function at
zero time delay was rather high: g(2)(0) 0.47.
IIIV semiconductor quantum dots looked a more attractive material to
investigate for single-photon emitters, having the advantage of being free
from photobleaching, blinking, spectral diffusion effects, or broad spectral
emission, as in the case with most other systems such as molecules or semi-
conductor microcrystals.
Therefore, Michler produced a first epitaxial self-assembled semicon-
ductor InAs QD demonstrating a single-photon turnstile device that gener-
ated a train of single-photon pulses using pulsed laser excitation of a single
quantum dot, embedded in a high-quality factor microcavity structure
(Michler, Kiraz, et al., 2000). The sample was grown by molecular beam
44 Mario Bertolotti et al.
Figure 18 (a) An SEM image of a typical device, showing the tapered etch and contra-
sting AlAs and GaAs distributed Bragg reflectors. (b) A photoluminescence spectrum
(solid line) and the spectral dependence of spontaneous emission lifetime at 4 K. From
Solomon et al. (2001).
(a) (c)
XX*
GaAs/AlAs } x 12 XX
DBR
QDs
X*
X
GaAs/AlAs
} x 30 (Emitted
DBR
photon)
Empty
(b)
1 m
Figure 19 (a) Schematic diagram of single-photon device, (b) scanning electron micro-
scope image of actual pillar structures, and (c) optical excitation scheme. From Santori
et al. (2004).
cavity with a distributed Bragg structure. The optical microcavity was used
to modify the spontaneous emission properties of the QD through the
Purcell effects. When a radiative transition of the QD is on resonance with
a cavity mode, if the QD couples much more strongly with this mode than
with the background leaky modes, the spontaneous emission rate can
increase substantially and light is emitted mainly into the cavity mode.
The resulting microcavities, exhibiting three-dimensional photon con-
finement, had quality factors of approximately 1.000 and measured
spontaneous-emission rate enhancement factors as high as 5. Many pillars
with only one or two quantum dots on resonance with a fundamental cavity
mode were found. The sample was cooled to 37 K in a cryostat. To gen-
erate single photons, 3 ps pulses from a Ti:sapphire laser every 13 ns were
focused onto the pillars from a steep angle. The quantum dot emission
was collected and a single polarization was selected and then spectrally
filtered.
The operation scheme is shown in Figure 19c. A short (23 ps) optical
pulse generated by a tunable Ti:sapphire laser raises the quantum dot into an
excited state containing one electronhole pair. The QD then quickly
relaxes (with a timescale of the order of 10 ps) to a lowest excited state. This
state then decays through a much slower spontaneous emission process
(100300 ps) to emit a single photon.
Quantum State Engineering 49
(a)
300
Counts
200
100
0
0 50 100 150
(b)
1000
Counts
500
0
0 50 100 150
(c)
600
Counts
400
200
0
0 50 100 150
t = t1 t2 (ns)
Figure 20 The figure shows photon correlation g(2)() measured with (a) above-band
excitation at 750 and 0.4 nm collection bandwidth; (b) resonant excitation at 909
and 0.4 nm collection bandwidth; and (c) resonant excitation at 904 and <0.1 nm col-
lection bandwidth. The narrower peaks in (c) are mainly due to the improved time res-
olution of the setup used for that measurement. The g(2)(0) value in (c) was 0.05. From
Santori et al. (2004).
50 Mario Bertolotti et al.
photons were detected in the same pulse, and thus the first goal in develop-
ing a single-photon source is to make the area of this peak as small as possible.
The peaks at times nTrep, where Trep 13 ns is the laser repetition
period, correspond to events in which one photon was detected from each
of two different pulses.
In Figure 20a, the photon correlation data obtained from a dot under
pulsed, above-band, excitation at 750 nm are shown. The light was collected
from a bright emission line at 932 nm. Figure 20b and c is obtained for res-
onant excitations at 909 and 904 nm, respectively. The performance was
considerably improved.
The two-photon suppression relative to an equivalent Poisson source
was estimated by comparing the area of the central peak with the areas of
the more distant side peaks. In this way, the probability of emitting two pho-
tons in the same pulse was approximately g(2)(0) 0.3 times that for a
Poisson-distributed source of the same intensity.
Most quantum information applications of single-photon sources require
pure photon states because they employ two-photon interference in their
schemes. The simplest example of two-photon interference is the Hong
OuMandel (Hong, Ou, & Mandel, 1987) experiment; if two independent
photons with identical wave packets collide at a 50:50 beamsplitter, they
always exit together, randomly choosing a side. They never exit in opposite
directions. However, if the photons do not have identical wave packets, they
can behave independently, and the two-photon interference effect is
reduced. The indistinguishability of the photons produced by the micro-
pillars was studied by Santori, Fattal, Vuckovic, et al. (2002). Figure 21
shows the main elements of the setup.
The single-photon source (InAs quantum dot) was excited twice every
13 ns by a pair of equally intense pulses with 2 ns separation. Two pulses,
each containing zero or one photons, emerge and are split into two arms
by a beam splitter, with one arm (2ns + t) longer than the other. The beams
then recombine at a different place on the same beam splitter and are col-
lected by two photon counters t1 and t2 as shown in Figure 21a. The
resulting histogram is shown in Figure 21b. Five peaks appear for every pair
of photons. For peaks 1 and 5 at 4 ns, the first photon follows the short
arm and the second photon the long arm and one photon goes to each coun-
ter. For peaks 2 and 4 at 2 ns, both photons follow the same arm. Peak
labeled 3 at 0 corresponds to the first photon traveling through the long
arm and the second through the short arm, so that the two photons collide
upon their second pass through the beam splitter. Only in this case can
Quantum State Engineering 51
Figure 21 (a) The main elements of the setup. (b) Histogram of the photon counts. From
Santori, Fattal, Vuckovic, et al. (2002).
20
DBRs
AIAs (79 nm)
Self-assembled
InGaAs/ GaAs
GaAs (130 nm)
GaAs-substrate
(a) (b)
WL
Cavity
PL intensity (a.u.)
s=0
s
r a 0.10a
y 0.12a
500 nm
0.15a
x 900 930 960 990
z
Wavelength (nm)
(c) (d)
Resonant wavelength (nm) 1000
980
960
940
920
900
0.00 0.05 0.10 0.15 0.20
1.0 0 +1.0 s (a)
Figure 24 (a) Scanning electron micrograph of the fabricated nanocavity with
a 300 nm, r 0:31a, and s 0:1a. (b) PL spectra for a series of nanocavities with
s 00:15a under Pex 5 kW/cm2. (c) The calculated electric field profile (Ey) of the shal-
low donor-type cavity mode confined in the L3-defect cavity with s 0:1a. The color
scale is normalized to maxjEyj. (d) The calculated (solid line) and the measured (solid
squares) resonant wavelengths of the cavity mode as a function of s. The inset in
(d) shows the calculated far-field radiation pattern of the cavity mode. From Chang,
Chen, et al. (2006).
Figure 26 An InAs quantum dot is embedded in the central part of a GaAs photonic
nanowire. The far-field emission collection is optimized with an integrated modal mirror
and a smooth tapering of the wire tip. From Claudon et al. (2010).
58 Mario Bertolotti et al.
Other centers of interest for photon emission are nickel related, silicon,
carbon, chromium color centers, and some other unknown defects.
Brouri, Beveratos, Poizat, and Grangier (2000b) reported on the use of
single color centers in diamond as single quantum emitters. They used bulk
diamond at room temperature with nonresonant excitation from an argon-
ion laser at 514 nm with a typical power of 10 mW. The used color center
was the NV defect with a zero-phonon line at 637 nm. Impurities were
introduced irradiating with 2 MeV electrons and annealing. The irradiation
was chosen such that the density of vacancies was of the order of 1 m3.
A limitation of the system is the existence of shelving in a metastable singlet
state. Shelving leads to the observation of photon bleaching for a timescale
longer than the lifetime of the center and to a decrease of the fluorescence
rate owing to the time spent by the center in this long-lived states. This com-
plex model was discussed by Brouri, Beveratos, Poizat, and Grangier
(2000a). A simplified level structure is a three-level scheme with fast non-
radiative decay within the two upper states and within two lower states
(Figure 27). The fluorescence emission is between the 2 and 1 level. Level
3 is the shelving level. The model is an oversimplification of what is the real
physical system. Shelving in some cases may be avoided.
Figure 28a shows a confocal microscopy raster scan (5 m 5 m) of the
sample made about 10 m below the diamond surface. Figure 28b is a line
scan along the dashed line in (a). Individual bright spots that correspond to
NV centers appear clearly.
Figure 29 shows the normalized correlation function g(2)(t) corrected for
the random coincidences from the background. The data are for the NV
center circled in Figure 28a.
The total absence of photobleaching allowed Brouri to lock the laser
exciting beam onto a single center during several hours. The evaluated over-
all detection efficiency was low 0.0014.
|2 rd
bG |3
r G GM
|1
Figure 27 A three-level model for an NV center. Fluorescence occurs from j2i to j1i. The
j3i level is a metastable level. From Brouri et al. (2000a).
Quantum State Engineering 63
5
(a)
4
m
2
0
0 1 2 3 4 5
m
(b)
7000
Count rate (s1)
6000
5000
4000
0 1 2 3 4 5
Position (m)
Figure 28 (a) Confocal microscope scanning; (b) counts of the scanning along the dot-
ted line shown in (a). From Brouri et al. (2000b).
c(t)
g ( 2 )(t )
450
2
400
1
0 350
20 0 20 40
t (ns)
Figure 29 Normalized correlation function corrected for the random coincidences from
the background. The data are for the center circled in Figure 28a. From Brouri et al.
(2000b).
64 Mario Bertolotti et al.
1000
800
600
400
200
0
200 0 200 400 600
t (ns)
Figure 30 The autocorrelation function of the emission from a single NV center. From
Beveratos, Brouri, et al. (2002).
Quantum State Engineering 65
1.50
1.00
g (2)( T )
g (2)( T )
1.00
0.50
0.50
0.00 0.00
100 50 0 50 100 100 50 0 50 100
Delay (ns) Delay (ns)
(c) (d)
3.00 0.12
2.50
2.00 0.08
1/T (ns1)
g (2)( T )
1.50
1.00 0.04
0.50
0.00 0
100 50 0 50 100 0 50 100
Delay (ns) Pump power (mW)
Figure 31 Measured correlation function for different excitation powers: (a) 11 W;
(b) 190 W; (c) 1.6 mW; and (d) fluorescent time as a function of pump power. Red
curves (grey in the print version) are fits. From Babinec et al. (2010).
Figure 33 Structure of the NE8 center consisting of one Ni atom surrounded by four
N atoms. From Gaebel et al. (2004).
1
G 12
k k 1/T1 2
g 12
0
Figure 34 Energy level structure of the NE8 center. From Gaebel et al. (2004).
Figure 35 Measured correlation function with cw excitation. The red (gray in the print
version) line is the theoretical curve. From Gaebel et al. (2004).
9. ENTANGLED STATES
A beam splitter is the simplest quantum optical device in which two
incident light beams interfere to produce two output beams. Quantum
properties of the beam splitter are manifested in the ability to generate an
entangled output state from a nonclassical but unentangled input (Gerry &
Knight, 2005; Kim et al., 2002).
Entangled states may thus be introduced by considering radiation inter-
acting with a beam splitter. Classically, if a field E1 enters from the left
(Figure 36), it is alternatively reflected (E2) or transmitted (E3), with reflec-
tion r and transmission t coefficients defined through
E2 rE1 and E3 tE1 (48)
Quantum State Engineering 71
(o)
t
t
E1 (1) r E3 (3)
r
E2 (2)
Figure 36 A beam splitter.
where the coefficients with primes refer to the properties of the beam splitter
for a field coming from the right.
It is easily seen that the commutation relations for a^2 and a^3 hold, if
jrj jr 0 j, jtj jt 0 j, jrj2 + jtj2 1 and r*t 0 + r 0 t* 0, and r*t + r 0 t 0 * 0.
We may now ask for a given input state to the beam splitter, what is the
output state. Remembering that all photon number states jni (and therefore
any superposition) may be constructed by the action of n powers of the
creation operator on the vacuum, we may use Equation (49) to construct
output states from the action of the transformed creation operators on the
vacuum states of the output modes.
Consider now, for example, a single-photon input state j0ioj1i1 which
we may write as a1j0ioj0i1. For a 50/50 beam splitter we have
~a{ 1 i~a{ 2 + ~a{ 3 =2 (50)
Thus,
j0io j1i1 ! 1=2 i~a{ 2 + ~a{ 3 j0i2 j0i3
1=2 ij1i2 j0i3 + j0i2 j1i3 (51)
72 Mario Bertolotti et al.
This result says that a single-photon incident at one of the input ports of
the beam splitter, the other part containing only vacuum, will be either
transmitted or reflected with equal probability. The output state (51) is an
entangled state: it cannot be written as a simple product of states of the indi-
vidual modes 2 and 3.
A first example of entanglement was observed experimentally at a time
before the quantum theory of matter had even be formalized. In the Franck
Hertz scattering, an incoming electron scatters from an electron bound in an
atom, say hydrogen for simplicity, and an electron is observed coming out.
There is no way to tell which electron comes out, and the quantum state
describing the two is an entangled state.
Entanglement can loosely be described as a kind of correlation. But it is
quite a bit more than that and, besides, classical systems can also exhibit cor-
relations. A correlation between two systems is simply the statement that if a
measurement of one system yields the result A, then a measurement on the
second system will yield the result B, with some probability. Perfect corre-
lation occurs when the second result is certain, given the outcome of the
first. Entangled states have some correlation property more. Let us consider
the case in which two photons are entangled in polarization
1=2 j0i1 j0i2 + j1i1 j1i2 (52)
If one only looks at one of the two particles, one finds it with equal prob-
ability in state j0i or in state j1i. The state shows classical two-particle cor-
relations in that when we measure a one (0) in channel 1, this immediately
implies a one (0) in channel 2. The quantum state however shows strong
correlation for any arbitrary superposition. For instance, if we consider a
polarization-entangled state, and put j0i for H (horizontal polarization)
and j1i for V (vertical polarization), a measurement in any polarization
direction in channel 1 will be 100% correlated in channel 2. No matter
which axis is chosen for the measurement of the polarization of one of
the photons, that choice completely determines what happens when we
measure the polarization state of the other photon (Aspect, Grangier, &
Roger, 1981; see also Grangier & Abram, 2003).
Two-photon states are the most popular entangled states in fundamental
and applied quantum optics. In a two-photon system, correlated or
entangled photon states can be generated by an atomic cascade process
(Grangier, Roger, & Aspect, 1986) or by a parametric down-conversion
process (Hong & Mandel, 1986; Rarity, Tapster, & Jakeman, 1987), where
Quantum State Engineering 73
incident photons are converted into photon pairs (signal and idler). The two
photons may propagate away from each other in different directions, and the
quantum states of each single photon may differ considerably. Despite the
distance between the subsystems, the pair will keep its correlation.
The subtle difference between classical correlation and entanglement
may be appreciated looking at the radiation emitted in a parametric gener-
ation process. In such processes, the interaction of a laser beam in a nonlinear
crystal produces simultaneously two beams dubbed signal and idler
(Figure 37). In what is described as type II down-conversion,2 the signal
and idler photons have orthogonal polarizations. In the process, energy
and momentum must be conserved, and birefringence effects cause the pho-
tons to be emitted along two intersecting cones, one of the ordinary (o) ray
and the other of the extraordinary (e) ray (Figure 38).
If one photon of the pair is detected at any time, then its partner is known
to be present.
If we measure the polarization of a photon in a point of the extraordinary
cone of Figure 38, we always measure a vertical polarization and we know
that in the corresponding point for which phase matching is fulfilled, the
polarization of the incoming photon is horizontal, and if we repeat the
experiment, the two polarizations will always be vertical on the extraordi-
nary cone and horizontal on the ordinary one.
Momentum conservation
Spontaneous
parametric ks ki
downconversion
kPUMP
s (signal)
Pump
Energy conservation
ws
Nonlinear i (idler) w PUMP
c(2) crystal wi
j PUMP = j s + j i
Figure 37 Scheme of spontaneous parametric down-conversion and momentum and
energy conservation.
2
The down-conversion process is called type I if the signal and idler photons have identical polarizations.
With type II phase matching, the signal and idler photons have orthogonal polarizations. If the angle
between the optical axis of the crystal and the pump beam direction is different from zero, the two
cones tilt toward the pump and intersect along two rays.
74 Mario Bertolotti et al.
Figure 38 (a) The spontaneous down-conversion cones present with type II phase
matching. (b) A photograph of the down-conversion photons, through an interference
filter at 702 nm. From Kwiat et al. (1995).
There are however two points clearly visible in the front photo of
Figure 38b in which the two polarizations are contemporarily present.
A measure in one of these points will yield vertical or horizontal polariza-
tion, indifferently. In these points, if we measure a vertical polarization in
one point, we know that in the other point polarization will be horizontal;
but we could also have measured a horizontal polarization instead (with a
50:50 probability) and in the other point polarization would have been ver-
tical. The wave function of the total system of the two produced photons
Quantum State Engineering 75
will collapse in one of the two situations just by virtue of the measurement,
and before it, any of the two polarizations will equally be probable.
The structure of Figure 38b repeats itself for different colors as shown in
Figure 39.
In the case of the signal-idler pair of photons emitted in spontaneous
parametric down-conversion, the two photons may be entangled in wave-
length, momentum, angular momentum, and frequency, as well as polari-
zation. Entanglement received precocious attention in a famous Einstein
paper (Einstein, Podolky, & Rosen, 1935) in which the possibility that a
measurement on a particle in some place at some time made another particle
far away at the same time to collapse is some state was taken as an indication
that quantum theory was not complete. This nonlocal property is today rec-
ognized as the most characteristics aspect of quantum mechanics.
Entangled pairs of photons may be used to have single photons. In fact,
when we detect in a given direction one photon in the pair we know, there
is a second one in the direction fulfilling the momentum and energy con-
servation rules due to the phase-matching conditions in the generating crys-
tal. Entanglement is further discussed, for example, in Strekalov, Kim, and
Shih (1999) and Bovino and Degiovanni (2008). The coupling of the down-
converted photons into an optical fiber received attention (see, for example,
Castelletto, Degiovanni, Migdall, & Ware, 2004; references therein).
Probably, Clauser (1974) was among the first to use twin pairs of photons
and study coincidences between them, although his attention was directed
to a study of the photoelectric effect and not to single-photon behavior. In
Figure 39 The intersecting beams at different colors (different gray shades in the print
version). The credit of this photo is to Paul Kwiat and Michael Reck from the University of
Innsbruck. Reprinted with permission. Copyright Faculty of Physics, University of Vienna.
76 Mario Bertolotti et al.
his experiment, the source contained 202Hg atoms which were excited by
electron bombardment. Light was produced at 1 567.6 nm and
2 435.8 nm by the cascade 91P1 ! 73S1 ! 63P1. The light emitted on
opposite sides of an assembly of excited atoms was focused separately into
two beams by means of a beam splitter. Each resulting beam was directed
through interference filter so that the wavelengths 1 and 2 were selected,
onto an appropriate photomultiplier and coincidences were studied. The
presence of one wavelength was used to know whether the other wave-
length was present. Clauser utilized two photons of different wavelengths,
produced in a cascade process in which one photon was produced after the
other. It was not an entangled pair. However, the presence of one photon at
one wavelength assured the existence of the other at the accompanying
wavelength.
In parametric interactions each time a pump photon creates an idler pho-
ton, it must also create a signal photon. Thus, if a photon is detected at the
idler mode, one can expect the presence of a corresponding photon at the
signal mode. This was verified experimentally in parametric fluorescence
(Hong & Mandel, 1986). In this way, single photons are created and anti-
bunching should occur. Such scheme of generating antibunched light was
studied theoretical by Stoler and Yurke (1986).
For these sources, the creation of photon pairs is probabilistic, rather than
deterministic. However, because the photons are created in pairs, one pho-
ton (the heralding photon) can be used to herald the creation of the other
photon. The second detector must be activated only whenever the first one
has detected a photon and not whenever a pump pulse has been emitted,
therefore circumventing the problem of empty pulses. Phase matching
allows one to choose the wavelength and determine the bandwidth of the
down-converted photons. The latter is in general rather large and varies
from a few nanometers up to some tens of nanometers. For the non-
degenerate case, one typically gets a bandwidth of 510 nm, whereas in
the degenerate case (where the central frequency of both photons is equal)
the bandwidth can be as large as 70 nm.
This photon-pair creation is very inefficient; typically, it takes some 1010
pump photons to create one pair in a given mode. The number of photon
pairs per mode is thermally distributed within the coherence time of the
photons and follows a Poissonian distribution for larger time windows
(Walls & Milburn, 1995).
With a pump power of 1 mW about 106 pairs per second can be col-
lected in single-mode fibers. Accordingly, in a time window of roughly
Quantum State Engineering 77
A quantum dot biexciton (two electrons and two holes, 2X) decays
radiatively through two intermediate optically active exciton states. First,
it decays to one of two optically active single-exciton states (one electron
and one hole, 1X) by emitting one photon and then to the empty-dot state
by emitting a second photon (Moreau, Robert, Manin, et al., 2001;
Regelman et al., 2001). The proposal that the biexciton-radiative cascade
could provide a source of event-ready polarization-entangled photon pairs
was made by Benson, Santori, Pelton, and Yamamoto (2000).
A first experimental study of the polarization correlation properties of
photon pairs emitted through biexciton decay in a single InAs quantum
dot was carried out by Santori, Fattal, Pelton, et al. (2002) with no positive
result, suggesting that quantum-dot asymmetry is an obstacle to realizing an
entangled photon source. Entanglement requires two decay paths with dif-
ferent polarizations, but indistinguishable otherwise. This is the case if the
intermediate exciton states are energetically degenerate and if the final state
of the QD is independent of the decay path. We may see how to achieve this
following Akopian et al. (2006). Let us consider Figure 40. In Figure 40b,
the biexciton-radiative cascade is shown with its two collinearly polarized
photons (either H or V) and distinguishable paths. As shown in
Figure 40b, the polarizations of the two photons are determined by the
intermediate exciton states (X) which are typically split into energy due
to electronhole exchange interaction and QD confining potential anisot-
ropy. Therefore, the two decay paths are spectrally distinguishable and
therefore the polarization states cannot be entangled. Akopian et al.
(2006) showed that the which path information can be erased by spectrally
selecting only the overlap of the two polarization components, and so pho-
ton entanglement can be achieved. If the splitting is smaller than the radiative
linewidth, the two decay paths are indistinguishable, producing two verti-
cally (horizontally) polarized entangled photons. The experiment was per-
formed using InAs QDs at 20 K. In a further paper, Akopian et al. (2007)
measured the density matrix of the polarization state of the photon pair emit-
ted from a continuously pumped quantum dot and showed that by applying
a temporal window, the quantum dot becomes a source of entangled light.
Young et al. demonstrated a similar system using InAs and carefully
selecting the quantum dots below 10 K. They succeeded to obtain more
than 70% of the detected photons to be entangled at 10 K, eliminating
the polarization splitting by a suitable design of the QD structure (Young
et al., 2006; Young, Stevenson, & Shields, 2007).
80 Mario Bertolotti et al.
(a)
XX0 20
X1
15
20 K X0 10
5
0 2
1.2803 1.2805
PL energy (eV)
Figure 40 (a) Photoluminescence spectrum of a single QD; (b) schematic description of
the biexciton cascade in natural QDs with its two collinearly polarized photons (either
H or V) and energetically distinguishable paths. (c, d) High-resolution polarization-
sensitive photoluminescence spectra of the biexciton XX and exciton X lines, respec-
tively. From Akopian et al. (2006).
Figure 41 Sketch of the source: two identical pillar microcavities with diameter D are
coupled. The center to center distance is labeled CC0 . A single quantum dot is inserted in
one of the pillars, k is the photon wavevector. From Dousse et al. (2010).
Figure 42 (a) Michelson interferometer set. (b) Mollow spectrum obtained after filtering
out the central peak. (c) High-resolution spectrum of the central Rayleigh line filtered
from the Mollow sidebands. (d) Photon autocorrelation of the central Mollow triplet
peak. Inset resonance fluorescence emission spectrum of the full Mollow triplet.
(e) Autocorrelation on the T line of the Mollow spectrum. From Ulhaq et al. (2012).
Figure 43 (a) Sketched geometry composed of a GaAs SPP channel (magenta; dark gray
in the print version) on a gold (yellow; light gray in the print version) substrate with a
dielectric nanogroove (blue; gray in the print version). The QD is assumed to be linearly
polarized along the z-direction; (b), (c1), and (c2) define physical quantities. From Gan
et al. (2012).
Figure 45 (a) Sketch of the Hanbury BrownTwiss experiment; (b) the fluorescence
image of a single quantum emitter; and (c) g(2)(t) of a single NV emitter, showing the
lowered value at t 0. From Kolesov et al. (2009).
86 Mario Bertolotti et al.
photon thus plays a role of a catalyst: it is explicitly present in both the input
and the output channels of the interaction and yet facilitates generation of
nonclassical states of light. By changing the amplitude of the target coherent
state, a gradual transition between a highly classical and a highly nonclassical
state of light occurs. The experiment was implemented by means of linear
optics and can be viewed as a step toward experimental realization of linear
optical quantum computation (Lvovsky & Mlynek, 2002).
Zavatta, Viciani, and Bellini (2004) reported the experimental genera-
tion of single-photon-added coherent states and their complete characteri-
zation by quantum tomography.
Cruz-Ramirez et al. (2012) obtained nondiffracting single photons, char-
acterized by a single-photon transverse intensity distribution which remains
essentially unchanged over a significant propagation distance (about 25 cm),
by properly shaping the pumping beam in a down-conversion process.
In analogy with Coulomb blockade for electrons (Fulton & Dolan, 1987;
Kastner, 1992; Likharev, 1999), photonphoton interactions in a nonlinear
optical cavity were proposed to realize photon blockade (Imamoglu et al.,
1997) for which a first photon within an optical system blocks the transmission
of a second photon, leading to an orderly output of photons one by one. The
proposal was extended by Dayan et al. (2008) using a microscopic optical res-
onator demonstrating a mechanism for the regulated transport of photons one
by one. Other proposals and experiments on single-photon blockade sources
were performed in the last 10 years (Bajcsy, Majumdar, Rundquist, &
Vuckovic, 2013; Bamba, Imamoglu, Carusotto, & Ciuti, 2011; Birnbaum
et al., 2005; Carusotto, Volz, & Imamoglu, 2010; Chang, Srensen,
Demler, & Lukin, 2007; Faraon et al., 2008; Gerace, T ureci, Imamoglu,
Giovannetti, & Fazio, 2009; He, Wei, et al., 2013; Liew & Savona, 2010;
Majumdar, Bajcsy, Rundquist, & Vuckovic, 2012; Reinhard et al., 2012;
Verger, Ciuti, & Carusotto, 2006). Eventually, Majumdar and Gerace
(2013) speculated about using a doubly resonant nanocavity filled with a
(2) material to realize a new scheme of photon blockade source.
Sotier et al. (2009) operating a single CdSe/ZnSe quantum dot on fem-
tosecond timescale, in a nonlinear regime, demonstrated the ability to
change the number of quanta in a femtosecond light pulse by exactly 1.
McCutcheon, Chang, Zhang, Lukin, and Loncar (2009) discussed an
approach to generate single photons with controllable wavelength and
bandwidth making use of an integrated nonlinear optical cavity in which
optical emission is directly frequency shifted into the desired domain using
intracavity nonlinear optical processes. Some examples were discussed with
90 Mario Bertolotti et al.
a GaAs photonic crystal and a GaP-based design for direct coupling between
an NV center in diamond or an InAs/GaAs quantum dot.
Upconversion, in principle, can be used to transduce one photon at a
given wavelength to a photon at another wavelength. Rakher, Ma,
Slattery, Tang, and Srinivasan (2010) proposed using it to couple different
quantum systems at nominally different energies. He demonstrated efficient
upconversion of single photons generated by a quantum dot at 1.3 m to
710 nm. In particular, they measured the second-order intensity correlation
of the upconverted 710 nm signal and showed that the field was dominantly
composed of single photons. The simultaneous wavelength translation and
amplitude modulation of single photons generated by a quantum dot emit-
ting near 1300 nm was further discussed (Rakher et al., 2011).
Control of single-photon packets is an important resource for developing
hybrid quantum systems which are composed of different physical systems
interacting via photons. Rakher and Srinivasan (2011) extended this control
to triggered photons emitted by a quantum dot, temporally shaping single-
photon wave packets on timescales fast compared to their radiative decay by
electro-optic modulation. In particular, telecommunicationsband single
photons resulting from the recombination of an exciton in a quantum dot
with exponentially decaying wave packets were synchronously modulated
to create Gaussian-shaped single photon wave packets.
Shen et al. (2011) discussed several proposals to achieve, at the single-
photon level, near complete optical isolation, by coupling a quantum impu-
rity to a passive, linear waveguide which has a locally planar, circular
polarization.
Aoki et al. (2009) realized a cavity quantum electrodynamic system in the
optical domain with efficient inputoutput coupling, while still maintaining
high internal efficiency for coupling to a single atom.
In quantum information processing, a promising approach is to create an
on-chip quantum network in which, for example, NVs interact optically via
nanophotonic interconnects. Coupling NVs to optical cavities plays a crucial
role in this implementation, by enhancing the NV emission into a well-
defined optical mode, which can be efficiently coupled to waveguides
and routed on-chip. Efforts to efficiently couple NVs in nanocrystalline dia-
mond to nanophotonic structures have been made by Englund, Faraon,
Zhang, Yamamoto, and Vuckovic (2007), Faraon, Barclay, Santori, Fu,
and Beausoleil, (2011), Barclay, Fu, Santori, Faraon, and Beausoleil
(2011), Englund et al. (2010), Wolters et al. (2010), and Van der Sar
et al. (2011, 2009).
Quantum State Engineering 91
Figure 47 (a) SEM image of a hybrid GaP-diamond whispering gallery mode cavity,
(b) top view, and (c) cross-section of the dominant electric field component. From
Barclay et al. (2011).
Figure 48 (a) Schematic, (b) SEM image, and (c) generated electric field intensity
(calculated). From Takemoto et al. (2007).
in which a qubit was transferred from an atomic carrier to a field one, and
then to another atom. The initial atom was either in one of its two energy
eigenstates or in the superposition of them. The mediating field was pre-
pared either in a 0 or 1 photon number state or in a superposition of the
two (Matre et al., 1997).
The principle of the quantum information transfer relied on the Rabi
precession at frequency /2 of an atom between two energy eigenstates
e and g in the cavity vacuum j0i. If the atom starts in the upper level e
and the effective resonant atomcavity interaction time t is such that t ,
the combined system evolves from the je,0i into the jg,1i state: the atomic
excitation is transferred to the field. If the atom is initially in level g, the sys-
tem starts in the jg,0i state and no evolution occurs. If the atom is initially in a
superposition jei + jgi, the linearity of quantum mechanics implies that
the combined system evolves into the state (j1i + j0i)jgi. The interaction
has transferred the quantum superposition from the atom to the field, leaving
the former in g. This information can then be transferred to a second atom
initially in g and crossing the cavity after a delay, in a process reverse to the
one experienced by the first atom.
94 Mario Bertolotti et al.
12. CONCLUSIONS
Although at present none of the sources discussed in this review could
be considered the best, a great improvement has occurred since the first
experiments. On the other hand, a perfect turnstile device can never be
made in practice. A real device suffers of unavoidable intrinsic loss and
the generation of multiple photons can never be excluded with absolute cer-
titude. At present, entangled state sources seem to be the best way to have
heralded single photons, but a real on demand source does not exist, yet.
However, for example, quantum cryptography is now matured, giving
the realization of quantum cryptographic devices, such as the
Q-KeyMaker by Bovino and Giardina (2011). Security aspects of quantum
key distribution with sub-Poisson light were examined by Waks, Santori,
and Yamamoto (2002).
The recent developments in the field of quantum technologies allow to
think to the development of quantum computers, where integrated photon-
ics sources are still one of the key elements. This review has given a pano-
rama of the different efforts to obtain single-photon sources and improve
their characteristics so as to be implemented into practical applications.
We are midway to a real satisfactory situation.
ADDENDUM
After the submission of the manuscript, a number of papers have been
published which we think deserve mention. The use of semiconductor
quantum dots has been excellently revised by Beveratos, Abram, Gerard,
and Robert-Philip (2014) discussing the tailoring of the emission from single
96 Mario Bertolotti et al.
REFERENCES
Acharya, R., & Sudarshan, E. C. G. (1960). Front description in relativistic quantum
mechanics. Journal of Mathematical Physics, 1, 532.
Aharonovich, I., Castelletto, S., Johnson, B. C., McCallum, J. C., Simpson, D. A.,
Greentree, A. D., et al. (2010). Chromium single-photon emitters in diamond fabricated
by ion implantation. Physics Review B, 81, 121201(R).
Aharonovich, I., Castelletto, S., Simpson, D. A., Su, C. H., Greentree, A. D., & Prawer, S.
(2011). Diamond-based single-photon emitters. Reports on Progress in Physics, 74, 076501.
Quantum State Engineering 97
Aharonovich, I., Zhou, C. Y., Stacey, A., Orwa, J., Castelletto, S., Simpson, D., et al. (2009).
Enhanced single-photon emission in the near infrared from a diamond color center.
Physics Review B, 79, 235316.
Aichele, T., Zwiller, V., & Benson, O. (2004). Visible single-photon generation from semi-
conductor quantum dots. New Journal of Physics, 6, 90.
Aichele, T., Zwiller, V., Benson, O., Akimov, I., & Henneberger, F. (2003). Single CdSe
quantum dots for high-bandwidth single-photon generation. Journal of the Optical Society
of America, B20, 2189.
Akimov, A. V., Mukherjee, A. Yu., Chang, D. E., Zibrov, A. S., Hemmer, P. R., Park, H.,
et al. (2007). Generation of single optical plasmons in metallic nanowires coupled to
quantum dots. Nature, 450, 402.
Akopian, N., Lindner, N. H., Poem, E., Berlatzky, Y., Avron, J., Gershoni, D., et al. (2006).
Entangled photon pairs from semiconductor quantum dots. Physical Review Letters, 96,
130501.
Akopian, N., Lindner, N. H., Poem, E., Berlatzky, Y., Avron, J., Gershoni, D., et al. (2007).
Correlated and entangled pairs of single photons from semiconductor quantum dots.
Journal of Applied Physics, 101, 081712.
Akselrod, G. M., Argyropoulos, C., Hoang, T. B., Ciraci, C., Fang, C., Huang, J., et al.
(2014). Probing the mechanisms of large Purcell enhancement in plasmonic nano-
antennas. Nature Photonics, 8, 835.
Alleaume, R., Treussart, F., Messin, G., Dumeige, Y., Roch, J. F., Beveratos, A., et al.
(2004). Experimental open-air quantum key distribution with a single-photon source.
New Journal of Physics, 6, 92.
Altewischer, E., Van Exter, M. P., & Woerdman, J. P. (2002). Plasmon-assisted transmission
of entangled photons. Nature, 418, 304.
Aoki, T., Parkins, A. S., Alton, D. J., Regal, C. A., Dayan, B., Ostby, E., et al. (2009). Effi-
cient routing of single photons by one atom and a microtoroidal cavity. Physical Review
Letters, 102, 083601.
Arcari, M., Sollner, I., Javadi, A., Lindskov, H. S., Mahmoodian, S., & Liu, J. (2014). Near-
unity coupling efficiency of a quantum emitter to a photonic-crystal waveguide. Physical
Review Letters, 113, 093603.
Aspect, A., Grangier, P., & Roger, G. (1981). Experimental tests of realistic local theories via
Bells theorem. Physical Review Letters, 47, 460.
Babinec, T. M., Hausmann, B., Choy, J., Khan, M., Hemmer, P., & Loncar, M. (2011).
Quantum photonics with diamond. IEEE Photonics Society Newsletter, 24, 13.
Babinec, T. M., Hausmann, B. J., Khan, M., Zhang, Y., Maze, J. R., Hemmer, P. R., et al.
(2010). A diamond nanowire single-photon source. Nature Nanotechnology, 5, 195.
Baier, M. H., Pelucchi, E., Kapon, E., Varoutsis, S., Gallart, M., et al. (2004). Single photon
emission from site-controlled pyramidal quantum dots. Applied Physics Letters, 84, 648.
Bajcsy, M., Majumdar, A., Rundquist, A., & Vuckovic, J. (2013). Photon blockade with a
four-level quantum emitter coupled to a photonic-crystal nanocavity. New Journal of
Physics, 15, 025014.
Bamba, M., Imamoglu, A., Carusotto, I., & Ciuti, C. (2011). Origin of strong photon anti-
bunching in weakly nonlinear photonic molecules. Physical Review A, 83, 021802(R).
Barclay, P. E., Fu K-M, C., Santori, C., & Beausoleil, R. G. (2009). Hybrid photonic crystal
cavity and waveguide for coupling to diamond NV-centers. Optics Express, 17, 9588.
Barclay, P. E., Fu K-M, C., Santori, C., Faraon, A., & Beausoleil, R. G. (2011). Hybrid
nanocavity resonant enhancement of color center emission in diamond. Physical Review
X, 1, 011007.
Barnes, W. L., Bjork, G., Gerard, J. M., Jonsson, P., Wasey, J. A. E., Worthing, P. T., et al.
(2002). Solid-state single photon sources: Light collection strategies. European Physical
Journal D: Atomic, Molecular, Optical and Plasma Physics, 18, 197.
98 Mario Bertolotti et al.
Barros, H. G., Stute, A., Northup, T. E., Russo, C., Schmidt, P. O., & Blatt, R. (2009).
Deterministic single-photon source from a single ion. New Journal of Physics, 11, 103004.
Basche, Th., Moerner, W. E., Orrit, M., & Talon, H. (1992). Photon antibunching in the
fluorescence of a single dye molecule trapped in a solid. Physical Review Letters, 69, 1516.
Becher, C., Kiraz, A., Michler, P., Schoenfeld, W. V., Petroff, P. M., Zhang, L., et al. (2002).
A quantum dot single-photon source. Physica E, 13, 412.
Bennett, A. J., Atkinson, P., See, P., Ward, M. B., Stevenson, R. M., Yuan, Z. L., et al.
(2006). Single-photon-emitting diodes: A review. Physica Status Solidi (b), 243, 3730.
Bennett, A. J., Patel, R. B., Nicoll, C. A., Ritchie, D. A., & Shields, A. J. (2009). Interference
of dissimilar photon sources. Nature Physics, 5, 715.
Bennett, A. J., Unitt, D. C., Atkinson, P., Ritchie, D. A., & Shields, A. J. (2005). High per-
formance single photon sources from photolithographically defined pillar microcavities.
Optics Express, 13, 50, and 7772.
Bennett, A. J., Unitt, D. C., See, P., Shields, A. J., Atkinson, P., Cooper, K., et al. (2005a).
Microcavity single-photon-emitting diode. Applied Physics Letters, 86, 181102.
Bennett, A. J., Unitt, D. C., See, P., Shields, A. J., Atkinson, P., Cooper, K., et al. (2005b).
Electrical control of the uncertainty in the time of single photon emission events. Physics
Review B, 72, 033316.
Benson, O., Santori, C., Pelton, M., & Yamamoto, Y. (2000). Regulated and entangled pho-
tons from a single quantum dot. Physical Review Letters, 84, 2513.
Benyoucef, M., Ulrich, S. M., Michler, P., Wiersig, J., Jahnke, F., & Forchel, A. (2004).
Enhanced correlated photon pair emission from a pillar microcavity. New Journal of
Physics, 6, 91.
Berry, D. W., Scheel, S., Myers, C. R., Sanders, B. C., Knight, P. L., & Laflamme, R. (2004).
Postprocessing with linear optics for improving the quality of single-photon sources. New
Journal of Physics, 6, 93.
Bertet, P., Osnaghi, S., Milman, P., Auffeves, A., Maioli, P., Brune, M., et al. (2002). Gen-
erating and probing a two-photon Fock state with a single atom in a cavity. Physical
Review Letters, 88, 143601.
Bertolotti, M. (2006). Wave interactions in photonic band structures: An overview. Journal of
Optics A: Pure and Applied Optics, 8, 81.
Beveratos, A., Abram, I., Gerard, J. M., & Robert-Philip, I. (2014). Quantum optics with
quantum dots. European Physical Journal D: Atomic, Molecular, Optical and Plasma Physics,
68, 377.
Beveratos, A., Brouri, R., Gacoin, T., Poizat, J.-P., & Grangier, P. (2001). Nonclassical radi-
ation from diamond nanocrystals. Physical Review A, 64, 061802(R).
Beveratos, A., Brouri, R., Gacoin, T., Villing, A., Poizat, J.-P., & Grangier, P. (2002). Single
photon quantum cryptography. Physical Review Letters, 89, 187901.
Beveratos, A., Kuhn, S., Brouri, R., Gacoin, T., Poizat, J.-P., & Grangier, P. (2002). Room
temperature stable single-photon source. The European Physical Journal D, 18, 191.
Bialynicki-Birula, I. (1996). Photon wave function. Progress in Optics, 36, 248.
Birnbaum, K. M., Boca, A., Miller, R., Boozer, A. D., Northup, T. E., & Kimble, H. J.
(2005). Photon blockade in an optical cavity with one trapped atom. Nature, 436, 87.
Bleuse, J., Claudon, J., Creasey, M., Malik, N. S., Gerard, J. M., Maksymov, I., et al. (2011).
Inhibition, enhancement, and control of spontaneous emission in photonic nanowires.
Physical Review Letters, 106, 103601.
Blinov, B. B., Moehring, D. L., Duan, L. M., & Monroe, C. (2004). Observation of entan-
glement between a single trapped atom and a single photon. Nature, 428, 153.
Bommer, M., Schulz, W. M., Rossbach, R., Jetter, M., Michler, P., Thomay, T., et al.
(2011). Triggered single-photon emission in the red spectral range from optically excited
InP/(Al, Ga)InP quantum dots embedded in micropillars up to 100 K. Journal of Applied
Physics, 110, 063108.
Quantum State Engineering 99
Borgstrom, M. T., Zwiller, V., Muller, E., & Imamoglu, A. (2005). Optically bright quan-
tum dots in single nanowires. Nano Letters, 5, 1439.
Bovino, F. A. (2013a). System for distributing cryptographic keys. Patent
EP2555466OS20130208894.
Bovino, F. A. (2013b). Interferometer and method for controlling the coalescence of a pair of
photons. Patent EP2543957USP20130176573.
Bovino, F. A., & Degiovanni, I. P. (2008). Quantum correlation bounds for quantum infor-
mation experiments optimization: The Wigner inequality case. Physical Review A, 77,
053822.
Bovino, F. A., & Giardina, M. (2011). Practical quantum cryptography on metropolitan area
network: The Q-KeyMaker. Polaris Innovation Journal, 062011.
Bovino, F. A., Varisco, P., Colla, A. M., Castagnoli, G., Di Giuseppe, G., & Sergienko, A. V.
(2003). Effective fiber-coupling of entangled photons for quantum communication.
Optics Communications, 227, 343.
Bovino, F. A., Varisco, P., Martinoli, A., De Nicolo, P., Bruzzo, S., Colla, A. M., et al.
(2005). Practical quantum key distribution using polarization entangled states. The Inter-
national Journal of Quantum Information, 3, 1.
Braig, C., Zarda, P., Kurtsiefer, C., & Weinfurter, H. (2003). Experimental demonstration of
complementarity with single photons. Applied Physics B: Lasers and Optics, 76, 113.
Branczyk, A., Ralph, T. C., Helwig, W., & Silberhorn, C. (2010). Optimized generation of
heralded Fock states using parametric down-conversion. New Journal of Physics, 12,
063001.
Brattke, S., Guthohrlein, G. R., Keller, M., Lange, W., Varcoe, B., & Walther, H. (2003).
Generation of photon number states on demand. Journal of Modern Optics, 50, 1103.
Brattke, S., Varcoe, B. T. H., & Walther, H. (2001a). Preparing Fock states in the micro-
maser. Optics Express, 8, 131.
Brattke, S., Varcoe, B. T. H., & Walther, H. (2001b). Generation of photon number states on
demand via cavity quantum electrodynamics. Physical Review Letters, 86, 3534.
Brida, G., Degiovanni, I. P., Genovese, M., Piacentini, F., Traina, P., Della Frera, A., et al.
(2012). An extremely low-noise heralded single-photon source: A breakthrough for
quantum technologies. Applied Physics Letters, 101, 221112.
Brokmann, X., Giacobino, E., Dahan, M., & Hermier, J. P. (2004). Highly efficient triggered
emission of single photons by colloidal CdSe/ZnS nanocrystals. Applied Physics Letters,
85, 712.
Brokmann, X., Messin, G., Desbiolles, P., Giacobino, E., Dahan, M., & Hermier, J. P.
(2004). Colloidal CdSe/ZnS quantum dots as single-photon sources. New Journal of
Physics, 6, 99.
Brouri, R., Beveratos, A., Poizat, J.-P., & Grangier, P. (2000a). Single-photon generation by
pulsed excitation of a single dipole. Physics Review A, 62, 063817.
Brouri, R., Beveratos, A., Poizat, J.-P., & Grangier, P. (2000b). Photon antibunching in the
fluorescence of individual color centers in diamond. Optics Letters, 25, 1294.
Brune, M., Haroche, S., Lefevre, V., Raimond, J. M., & Zagury, N. (1990). Quantum non-
demolition measurements of small photon numbers by Rydberg-atom phase-sensitive
detection. Physical Review Letters, 65, 976.
Brune, M., Haroche, S., Lefevre, V., Raimond, J. M., & Zagury, N. (1992). Manipulation of
photons in a cavity by dispersive atom-field coupling: Quantum nondemolition mea-
surements and generation of Schr odinger cat states. Physics Review A, 45, 5193.
Brunel, C., Lounis, B., Tamarat, P., & Orrit, M. (1999). Triggered source of single photons
based on controlled single molecule fluorescence. Physical Review Letters, 83, 2722.
Brunel, C., Tamarat, P., Lounis, B., Plantard, J., & Orrit, M. (1998). Driving the Bloch vector
of a single molecule: Towards a triggered single photon source. Comptes-Rendus de
lAcademie des Sciences, Paris, 326, 911.
100 Mario Bertolotti et al.
Buckley, S., Rivoire, K., & Vuckovic, J. (2012). Engineered quantum dot single-photon
sources. Reports on Progress in Physics, 75, 126503.
Bulu, I., Babinec, T., Hausmann, B., Choy, J. T., & Loncar, M. (2011). Plasmonic resonators
for enhanced diamond NV-center single photon sources. Optics Express, 19, 5268.
Bulu, I., Caglayan, H., & Ozbay, E. (2003). Radiation properties of sources inside photonic
crystals. Physics Review B, 67, 205103.
Busch, K., Vats, N., John, S., & Sanders, B. C. (2000). Radiating dipoles in photonic crystals.
Physics Review E, 62, 4251.
Buttler, W. T., Hughes, R. J., Lamoreaux, S. K., Morgan, G. L., Nordholt, J. E., &
Peterson, C. G. (2000). Daylight quantum key distribution over 1.6 km. Physical Review
Letters, 84, 5652.
Carmichael, H. J., & Walls, D. F. (1976). Proposal for the measurement of the resonant Stark
effect by photon correlation techniques. Journal of Physics B, 9, L43.
Carusotto, I., Volz, T. B., & Imamoglu, A. (2010). Feshbach blockade: Single-photon
nonlinear optics using resonantly enhanced cavity polariton scattering from biexciton
states. Europhysics Letters, 90, 37001.
Castelletto, S., Degiovanni, I. P., Migdall, A., & Ware, M. (2004). On the measurement of
two-photon single-mode coupling efficiency in parametric downconversion photon
sources. New Journal of Physics, 6, 87.
Castelletto, S., Johnson, B. C., Ivady, V., Stavrias, N., Umeda, T., Gali, A., et al. (2014).
A silicon carbide room-temperature single-photon source. Nature Materials, 13, 151.
Centini, M., Perina, J., Jr., Sciscione, L., Sibilia, C., Scalora, M., Bloemer, M. J., et al.
(2005). Entangled photon pair generation by spontaneous parametric down
conversion in finite-length one-dimensional photonic crystals. Physics Review A, 72,
033806.
Cerf, N. J., & Flurasek, J. (2006). Optical quantum cloning. Progress in Optics, 49, 455.
Chance, R. R., Prock, A., & Silbey, R. (1974). Lifetime of an emitting molecule near a par-
tially reflecting surface. The Journal of Chemical Physics, 60, 2749.
Chaneliere, T., Matsukevich, D. N., Jenkins, S. D., Lan, S.-Y., Kennedy, T. A. B., &
Kuzmich, A. (2005). Storage and retrieval of single photons transmitted between remote
quantum memories. Nature, 438, 833.
Chang, W. H., Chen, W. Y., Chang, H.-S., Hsieh, T. P., Chyi, J.-I., & Hsu, T.-M. (2006).
Efficient single-photon sources based on low-density quantum dots in photonic-crystal
nanocavities. Physical Review Letters, 96, 117401.
Chang, D. E., Srensen, A. S., Demler, E. A., & Lukin, M. D. (2007). A single-photon tran-
sistor using nanoscale surface plasmons. Nature Physics, 3, 807.
Chang, D. E., Srensen, A. S., Hemmer, P. R., & Lukin, M. D. (2006). Quantum optics with
surface plasmons. Physical Review Letters, 97, 053002.
Chen, S., Chen, Yu.-A., Strassel, T., Yuan, Zhen-Sheng, Zhao, Bo, Schmiedmayer, J., et al.
(2006). Deterministic and storable single-photon source based on a quantum memory.
Physical Review Letters, 97, 173004.
Choi, K. S., Deng, H., Laurat, J., & Kimble, H. J. (2008). Mapping photonic entanglement
into and out of a quantum memory. Nature, 452, 67.
Choy, J. T., Hausmann, B. J. M., Babinec, T. M., Bulu, I., Khan, M., Maletinsky, P., et al.
(2011). Enhanced single-photon emission from a diamond-silver aperture. Nature
Photonics, 5, 738.
Claudon, J., Bleuse, J., Malik, N. S., Bazin, M., Jaffrennou, P., Gregersen, N., et al. (2010).
A highly efficient single-photon source based on a quantum dot in a photonic nanowire.
Nature Photonics, 4, 174.
Clauser, J. F. (1974). Experimental distinction between the quantum and classical field-
theoretic predictions for the photoelectric effect. Physics Review D, 9, 853.
Quantum State Engineering 101
Einstein, A. (1905). Concerning an heuristic point of view toward the emission and trans-
formation of light. Annalen der Physik, 17, 132.
Einstein, A., Podolky, B., & Rosen, N. (1935). Can quantum-mechanical description of
physical reality be considered complete? Physics Review, 47, 777.
Eisaman, M. D., Andre, A., Massou, F., Fleischhauer, M., Zibrov, A. S., & Lukin, M. D.
(2005). Electromagnetically induced transparency with tunable single-photon pulses.
Nature, 438, 837.
Eisaman, M. D., Fan, J., Mugdall, A., & Polyakov, S. V. (2011). Single photon sources and
detectors. Review of Scientific Instruments, 82, 071101.
Ellis, D. J. P., Bennett, A., Dewhurst, S., Nicoll, C. A., Ritchie, D. A., & Shields, A. J.
(2008). Cavity enhanced radiative emission rate in a single-photon emitting diode
operating at 0.5GHz. New Journal of Physics, 10, 043035.
Ellis, D. J. P., Bennett, A. J., Shields, A. J., Atkinson, P., & Ritchie, D. A. (2006). Electrically
addressing a single self-assembled quantum dot. Applied Physics Letters, 88, 133509.
Englund, D., Faraon, A., Zhang, B., Yamamoto, Y., & Vuckovic, J. (2007). Generation and
transfer of single photons on a photonic crystal chip. Optics Express, 15, 5550.
Englund, D., Fattal, D., Waks, E., Solomon, G., Zhang, B., Nakaoka, T., et al. (2005). Con-
trolling the spontaneous emission rate of single quantum dots in a two-dimensional pho-
tonic crystal. Physical Review Letters, 95, 013904.
Englund, D., Shields, B., Rivoire, K., Hatami, F., Vuckovic, J., Park, H., et al. (2010). Deter-
ministic coupling of a single nitrogen vacancy center to a photonic crystal cavity. Nano
Letters, 10, 3922.
Faddeev, L. D., Khalfin, L. A., & Komarov, I. V. (2004). Selected works V.A. Fock. London:
Chapman & Hall/CRC.
Fan, B., Duan, Z., Zhou, L., Yuan, C., Ou, Z. Y., & Zhanget, W. (2009). Generation of a single-
photon source via a four-wave mixing process in a cavity. Physics Review A, 80, 063809.
Fan, J., Eisaman, M. D., & Migdall, A. (2007). Bright phase-stable broadband fiber-based
source of polarization-entangled photon pair. Physics Review A, 76, 043836.
Fan, J., & Migdall, A. (2007). A broadband high spectral brightness fiber-based two-photon
source. Optics Express, 15, 2915.
Faraon, A., Barclay, P. E., Santori, C., Fu K-M, C., & Beausoleil, R. G. (2011). Resonant
enhancement of the zero-phonon emission from a color center in a diamond cavity.
Nature Photonics, 5, 301.
Faraon, A., Fushman, I., Englund, D., Stoltz, N., Petroff, P., & Vucokovic, J. (2008). Coher-
ent generation of non-classical light on a chip via photon-induced tunnelling and block-
ade. Nature Physics, 4, 859.
Fasel, S., Robin, F., Moreno, E., Erni, D., Gisin, N., & Zbinden, H. (2005). Energy-time
entanglement preservation in plasmon-assisted light transmission. Physical Review Letters,
94, 110501.
Ferretti, S., & Gerace, D. (2012). Single-photon nonlinear optics with Kerr-type nanostruc-
tured materials. Physics Review B, 85, 033303.
Filip, R., & Zapletal, P. (2014). Squeezed-state generation from single-photon sources. Phys-
ics Review A, 90, 043854.
Filipowicz, P., Javanainen, J., & Meystre, P. (1986a). Quantum and semiclassical steady states
of a kicked cavity mode. Journal of the Optical Society of America, B3, 906.
Filipowicz, P., Javanainen, J., & Meystre, P. (1986b). Theory of a microscopic maser. Physics
Review A, 34, 3077.
Flagg, E. B., Muller, A., Polyakov, S. V., Ling, A., Migdall, A., et al. (2010). Interference of
single photons from two separate semiconductor quantum dots. Physical Review Letters,
104, 137401.
Flagg, E. B., Muller, A., Robertson, J. W., Founta, S., Deppe, G. D., Xiao, M., et al. (2009).
Resonantly driven coherent oscillations in a solid-state quantum emitter. Nature Physics,
5, 203.
Quantum State Engineering 103
Fleury, L., Segura, J.-M., Zumofen, G., Hecht, B., & Wild, U. P. (2000). Nonclassical pho-
ton statistics in single-molecule fluorescence at room temperature. Physical Review Letters,
84, 1148.
Florescu, M., Scheel, S., Haffner, H., Lee, H., Strekalov, D., Knight, P. L., et al. (2005).
Single photons on demand from 3D photonic band-gap structures. Europhysics Letters,
69, 945.
Fock, V. (1932). Konfigurazionsraum und zweite quantellung. Zeitschrift f ur Physik, 75, 622.
Foden, C. L., Talyanskii, V. I., Milburn, G. J., Leadbeater, M. L., & Pepper, M. (2000).
High-frequency acousto-electric single-photon source. Physics Review A, 62,
011803R.
Friedler, I., Sauvan, C., Hugonin, J. P., Lalanne, P., Claudon, J., & Gerard, J. M. (2009).
Solid-state single photon sources: The nanowire antenna. Optics Express, 17, 2095.
Fu, K. M. C., Santori, C., Barclay, P. E., Aharonovich, I., Prawer, S., Meyer, N., et al.
(2008). Coupling of nitrogen-vacancy centers in diamond to a GaP waveguide. Applied
Physics Letters, 93, 234107.
Fulton, T. A., & Dolan, G. J. (1987). Observation of single-electron charging effects in small
tunnel junctions. Physical Review Letters, 59, 109.
Gabrielse, G., & Dehmelt, H. (1985). Observation of inhibited spontaneous emission. Phys-
ical Review Letters, 55, 67.
Gaebel, T., Popa, I., Gruber, A., Domhan, M., Jelezko, F., & Wrachtrup, J. (2004). Stable
single-photon source in the near infrared. New Journal of Physics, 6, 98.
Gan, C. H., Hugonin, J. P., & Lalanne, P. (2012). Proposal for compact solid-state III-V
single-plasmon sources. Physics Review X, 2, 021008.
Garraway, B. M., Sherman, B., Moya-Cessa, H., Knight, P. L., & Kurizki, G. (1994). Gen-
eration and detection of nonclassical field states by conditional measurements following
two-photon resonant interactions. Physics Review A, 49, 535.
Gazzano, O., Almeida, M. P., Nowak, A. K., Portalupi, S. L., Lematre, A., Sagnes, I., et al.
(2013). Entangling quantum-logic gate operated with an ultrabright semiconductor
single-photon source. Physical Review Letters, 110, 250501.
Geerligs, L. S., Anderegg, V. F., Holweg, P. A. M., Mooij, J. E., Pothier, H., Esteve, D., et al.
(1990). Frequency-locked turnstile device for single electrons. Physical Review Letters, 64,
2691.
Gerace, D., T ureci, H. E., Imamoglu, A., Giovannetti, V., & Fazio, R. (2009). The
quantum-optical Josephson interferometer. Nature Physics, 5, 281.
Gerard, J. M., Barrier, D., Marzin, J. Y., Kuszelewicz, R., Manin, L., & Costard, E. (1996).
Quantum boxes as active probes for photonic microstructures: The pillar microcavity
case. Applied Physics Letters, 69, 449.
Gerard, J. M., & Gayral, B. (1999). Strong Purcell effect for InAs quantum boxes in three-
dimensional solid-state microcavities. Journal of Lightwave Technology, 17, 2089.
Gerard, J. M., Sermage, B., Gayral, B., Legrand, B., Costard, E., & Thierry-Mieg, V. (1998).
Enhanced spontaneous emission by quantum boxes in a monolithic optical microcavity.
Physical Review Letters, 81, 1110.
Gerry, C., & Knight, P. (2005). Introductory quantum optics. Cambridge: Cambridge
University Press.
Gevaux, D. G., Bennett, A. J., Stevenson, R. M., Shields, A. J., Atkinson, P., & Griffiths, J.
(2006). Enhancement and suppression of spontaneous emission by temperature tuning
InAs quantum dots to photonic crystal cavities. Applied Physics Letters, 88, 131101.
Ghirardi, G. C., & Weber, T. (1983). Quantum mechanics and faster-than-light communi-
cation: Methodological considerations. Nuovo Cimento, B78, 9.
Gisin, N., Ribordy, G., Tittel, W., & Zbinden, H. (2002). Quantum cryptography. Reviews
of Modern Physics, 74, 145.
Gisin, N., & Thew, R. (2007). Quantum communication. Nature Photonics, 1, 165.
Glauber, R. J. (1963a). The quantum theory of optical coherence. Physics Review, 130, 2529.
104 Mario Bertolotti et al.
Glauber, R. J. (1963b). Coherent and incoherent states of the radiation field. Physics Review,
131, 2766.
Gogyan, A., Sisakyan, N., Akhmedzhanov, R., & Malakyan, Y. (2014). Heralded generation
of single photons entangled in multiple temporal modes with controllable waveforms.
Laser Physics, 24, 115204.
Goldschmidt, E. A., Eisaman, M. D., Fan, J., Polyakov, S. V., & Migdall, A. (2008).
Spectrally bright and broad fiber-based heralded single-photon source. Physics Review
A, 78, 013844.
Goy, P., Raimond, J. M., Gross, M., & Haroche, S. (1983). Observation of cavity-enhanced
single-atom spontaneous emission. Physical Review Letters, 50, 1903.
Grangier, P., & Abram, I. (2003). Single photons on demand. Physics World, 27, 31.
Grangier, P., Roger, G., & Aspect, A. (1986). Experimental evidence for a photon antic-
orrelation effect on a beam splitter: A new light on single-photon interferences.
Europhysics Letters, 1, 173.
Gregersen, N., Nielsen, T. R., Mrk, J., Claudon, J., & Gerard, J.-M. (2010). Designs for
high-efficiency electrically pumped photonic nanowire single-photon sources. Optics
Express, 18, 21204.
Gruber, A., Drabenstedt, A., Tietz, C., Fleury, L., Wrachtrup, J., & von Borczyskowski, C.
(1997). Scanning confocal optical microscopy and magnetic resonance on single defect
centers. Science, 276, 2012.
Guerreiro, T., Martin, A., Sanguinetti, B., Pelc, J. S., Langrock, C., Fejer, M. M., et al.
(2014). Nonlinear interaction between single photons. Physical Review Letters, 113,
173601.
Hafenbrak, R., Ulrich, S. M., Michler, P., Wang, L., Rastelli, A., & Schmidt, O. G. (2007).
Triggered polarization-entangled photon pairs from a single quantum dot up to 30 K.
New Journal of Physics, 9, 315.
Han, B., Ji, Z., & Zheng, Y. (2009). Photon counting statistics of single molecule in solid
matrix. The Journal of Chemical Physics, 130, 244502.
Hanbury Brown, R., & Twiss, R. Q. (1956a). Correlation between photons in two coherent
beams of light. Nature, 177, 27.
Hanbury Brown, R., & Twiss, R. Q. (1956b). Interferometry of the intensity fluctuations in
light. I. Basic theory: The correlation between photons in coherent beams of radiation.
Proceedings of the Royal Society A, 242, 300.
Hanbury Brown, R., & Twiss, R. Q. (1958). Interferometry of the intensity fluctuations in
light. II. An experimental test of the theory for partially coherent light. Proceedings of the
Royal Society A, 243, 291.
Hanzawa, H., Umemura, N., Nisida, Y., Kanda, H., Okada, M., & Kobayashi, M.
(1996). Disorder effects of nitrogen impurities, irradiation-induced defects, and
C13 isotope composition on the Raman spectrum in synthetic Ib diamond. Physics
Review B, 54, 3793.
Harder, G., Ansari, V., Brecht, B., Dirmeier, T., Marquardt, C., & Silberhorn, C. (2013). An
optimized photon pair source for quantum circuits. Optics Express, 21, 13975.
Haroche, S., Raimond, J. M., & Brune, M. (2002). Experimental quantum computation and
information. In F. de Martini & C. Monroe (Eds.), Proceedings of the international school of
physics Enrico Fermi, course CXLVIII, Amsterdam 2002 (pp. 3766).
Harris, N. C., Grassani, D., Simbula, A., Pant, M., Galli, M., Baehr-Jones, T., et al. (2014).
Integrated source of spectrally filtered correlated photons for large-scale quantum
photonic systems. Physics Review X, 4, 041047.
Hartschuh, A., Pedrosa, H. N., Peterson, J., Huang, L., Anger, P., Qian, H., et al. (2005).
Single carbon nanotube optical spectroscopy. Chemphyschem: A European Journal of
Chemical Physics and Physical Chemistry, 6, 1.
Quantum State Engineering 105
Hausmann, B. J. M., Babinec, T. M., Choy, J. T., Hodges, J. S., Hong, S., Bulu, I., et al.
(2011). Single-color centers implanted in diamond nanostructures. New Journal of Physics,
13, 045004.
He, Y., & Barkai, E. (2006). Theory of single photon on demand from a single molecule
source. Physical Chemistry Chemical Physics: PCCP, 8, 5056.
He, Y., He, Y.-M., Wei, Y.-J., Jiang, X., Chen, M.-C., Xiong, F.-L., et al. (2013). Indis-
tinguishable tunable single photons emitted by spin-flip Raman transitions in InGaAs
quantum dots. Physical Review Letters, 111, 237403.
He, Y., Wei, Y.-J., Wu, D., Atat ure, M., Schneider, C., H ofling, S., et al. (2013). On-
demand semiconductor single-photon source with near-unity indistinguishability.
Nature Nanotechnology, 8, 213.
Heindel, T., Schneider, C., Lermer, M., Kwon, S. H., Braun, T., & Reitzenstein, S. (2010).
Electrically driven quantum dot-micropillar single photon source with 34% overall effi-
ciency. Applied Physics Letters, 96, 011107.
Heinrich, J., Huggenberger, A., Heindel, T., Reitzenstein, S., H ofling, S., Worschech, L.,
et al. (2010). Single photon emission from positioned GaAs/AlGaAs photonic nano-
wires. Applied Physics Letters, 96, 211117.
Hennrich, M., Legero, T., Kuhn, A., & Rempe, G. (2000). Vacuum-stimulated Raman scat-
tering based on adiabatic passage in a high-finesse optical cavity. Physical Review Letters,
85, 4872.
Hennrich, M., Legero, T., Kuhn, A., & Rempe, G. (2004). Photon statistics of a non-
stationary periodically driven single-photon source. New Journal of Physics, 6, 86.
Hijlkema, M., Weber, B., Specht, H. P., Webster, S. C., Kuhn, A., & Rempe, G. (2007).
A single-photon server with just one atom. Nature Physics, 3, 253.
Hoffges, J. T., Baldauf, H. W., Lange, W., & Walther, H. (1997). Heterodyne measurement
of the resonance fluorescence of a single ion. Journal of Modern Optics, 44, 1999.
Hogele, A., Galland, C., Winger, M., & Imamoglu, A. (2008). Photon antibunching in the
photoluminescence spectra of a single carbon nanotube. Physical Review Letters, 100, 217401.
Holland, M. J., Walls, D. F., & Zoller, P. (1991). Quantum nondemolition measurements of
photon number by atomic beam deflection. Physical Review Letters, 67, 1716.
Holmes, M. J., Choi, K., Kako, S., Arita, M., & Arakawa, Y. (2014). Room-temperature
triggered single photon emission from a III-Nitrite site-controlled nanowire quantum
dot. Nano Letters, 14, 982.
Hong, C. K., & Mandel, L. (1986). Experimental realization of a localized one-photon state.
Physical Review Letters, 56, 58.
Hong, C. K., Ou, Z. Y., & Mandel, L. (1987). Measurement of subpicosecond time intervals
between two photons by interference. Physical Review Letters, 59, 2044.
Hours, J., Varoutsis, S., Gallart, M., Bloch, J., Robert-Philip, I., Cavanna, A., et al. (2003). Single
photon emission from individual GaAs quantum dots. Applied Physics Letters, 82, 2206.
Hwang, J., & Hinds, E. A. (2011). Dye molecules as single-photon sources and large optical
nonlinearities on a chip. New Journal of Physics, 13, 085009.
Imamoglu, A., Schmidt, H., Woods, G., & Deutschet, M. (1997). Strongly interacting pho-
tons in a nonlinear cavity. Physical Review Letters, 79, 1467.
Imamoglu, A., & Yamamoto, Y. (1994). Turnstile device for heralded single photons: Cou-
lomb blockade of electron and hole tunneling in quantum confined p-i-n hetero-
junctions. Physical Review Letters, 72, 210.
Jacobs, B. C., Pittman, B., & Franson, J. D. (2006). Single photon source using laser pulses
and two-photon absorption. Physics Review A, 74, 010303.
Jakubczyk, T., Pacuski, W., Smole nski, T., Golnik, A., Florian, M., & Jahnke, F. (2012).
Pronounced Purcell enhancement of spontaneous emission in CdTe/ZnTe quantum
dots embedded in micropillar cavities. Applied Physics Letters, 101, 132105.
106 Mario Bertolotti et al.
Jarjour, A. F., Taylor, R. A., Oliver, R. A., Kappers, M. J., Humphreys, C. J., & Tahraoui, A.
(2007). Cavity-enhanced blue single-photon emission from a single InGaN/GaN quan-
tum dot. Applied Physics Letters, 91, 052101.
Jaynes, E. T., & Cummings, F. W. (1963). Comparison of quantum and semiclassical radi-
ation theories with application to the beam maser. Proceedings of the IEEE, 51, 89.
Jeffrey, E., Peters, N. A., & Kwiat, P. G. (2004). Towards a periodic deterministic source of
arbitrary single-photon states. New Journal of Physics, 6, 100.
Jennewein, T., Simon, C., Weihs, G., Weinfurter, H., & Zeilinger, A. (2000). Quantum
cryptography with entangled photons. Physical Review Letters, 84, 4729.
John, S. (1987). Strong localization of photons in certain disordered dielectric superlattices.
Physical Review Letters, 58, 2426.
Julsgaard, B., Sherson, J., Cirac, J. I., Fiurasek, J., & Polziket, E. S. (2004). Experimental
demonstration of quantum memory for light. Nature, 432, 482.
Kako, S., Santori, C., Hoshino, K., G otzinger, S., Yamamoto, Y., & Arakawa, Y. (2006).
A gallium nitride single-photon source operating at 200 K. Nature Materials, 5, 887.
Kang, S., Lim, S., Hwang, M., Kim, W., Kim, J.-R., & An, K. (2011). Controlled generation
of single photons in a coupled atom-cavity system at a fast repetition-rate. Optics Express,
19, 2440.
Kaniber, M., Kress, A., Laucht, A., Bichler, M., Meyer, R., Amann, M. C., et al. (2007).
Efficient spatial redistribution of quantum dot spontaneous emission from two-
dimensional photonic crystals. Applied Physics Letters, 91, 061106.
Kaniber, M., Laucht, A., H urlimann, T., Bichler, M., Meyer, R., Amann, M. C., et al.
(2008). Highly efficient single-photon emission from single quantum dots within a
two-dimensional photonic band-gap. Physics Review B, 77, 073312.
Kastner, M. A. (1992). The single-electron transistor. Reviews of Modern Physics, 64, 849.
Keller, O. (2007). Historical papers on the particle concept of light. Progress in Optics, 50, 51.
Keller, M., Lange, B., Hayasaka, K., Lange, W., & Walther, H. (2004a). Continuous gen-
eration of single photons with controlled waveform in an ion-trap cavity system. Nature,
431, 1075.
Keller, M., Lange, B., Hayasaka, K., Lange, W., & Walther, H. (2004b). A calcium ion in a
cavity as a controlled single-photon source. New Journal of Physics, 6, 95.
Kilin, S. Ya. (2001). Quanta and information. Progress in Optics, 42, 1.
Kim, J., Benson, O., Kan, H., & Yamamoto, Y. (1999). A single-photon turnstile device.
Nature, 397, 500.
Kim, M. S., Son, W., Buzek, V., & Knight, P. L. (2002). Entanglement by a beam splitter:
Nonclassicality as a prerequisite for entanglement. Physics Review A, 65, 034343.
Kimble, H. J., Dagenais, M., & Mandel, L. (1977). Photon antibunching in resonance fluo-
rescence. Physical Review Letters, 39, 691.
Kimble, H. J., & Mandel, L. (1976). Theory of resonance fluorescence. Physical Review A, 13,
2123.
Kiraz, A., Ehrl, M., Hellerer, Th., M E., Brauchle, C., & Zumbusch, A. (2005).
ustecaploglu, O.
Indistinguishable photons from a single molecule. Physical Review Letters, 94, 223602.
Kiraz, A., Falth, S., Becher, C., Gayral, B., Schoenfeld, W. V., Petroff, P. M., et al. (2002).
Photon correlation spectroscopy of a single quantum dot. Physics Review B, 65, 161303R.
Kitson, S. C., Jonsson, P., Rarity, J. G., & Tapster, P. R. (1998). Intensity fluctuation spec-
troscopy of small numbers of dye molecules in a microcavity. Physics Review A, 58, 620.
Kleppner, D. (1981). Inhibited spontaneous emission. Physical Review Letters, 47, 233.
Knill, E., Laflamme, R., & Milburn, G. J. (2001). A scheme for efficient quantum compu-
tation with linear optics. Nature, 409, 46.
Kok, P. (2010). Quantum optics: Entangled photons report for duty. Nature Photonics, 4, 504.
Kolesov, R., Grotz, B., Balasubramanian, G., St ohr, R. J., Nicolet, A. A. L., Hemmer, P. R., et al.
(2009). Waveparticle duality of single surface plasmon polaritons. Nature Physics, 5, 470.
Quantum State Engineering 107
Kolesov, R., Xia, K., Reuter, R., St ohr, R., Zappe, A., Meijer, J., et al. (2012). Optical
detection of a single rare-earth ion in a crystal. Nature Communications, 3, 1029.
http://dx.doi.org/10.1038/ncomms2034.
Kramers, H. A. (1957). Quantum mechanics. Amsterdam: North-Holland.
Krause, J., Scully, M. O., & Walther, H. (1987). State reduction and jni-state preparation in a
high-Q micromaser. Physical Review A, 36, 4547.
Krause, J., Scully, M. O., Walther, T., & Walther, H. (1989). Preparation of a pure number
state and measurement of the photon statistics in a high-Q micromaser. Physics Review A,
39, 1915.
Kremer, P. E., Dada, A. C., Kumar, P., Ma, Y., Kumar, S., Clarke, E., et al. (2014).
Strain-tunable quantum dot embedded in a nanowire antenna. Physical Review B, 90,
201408(R).
Kremling, S., Tessarek, C., Dartsch, H., Figge, S., H ofling, S., Worschech, L., et al. (2012).
Single photon emission from InGaN/GaN quantum dots up to 50 K. Applied Physics Let-
ters, 100, 061115.
Kress, A., Hofbauer, F., Reinelt, N., Kaniber, M., Krenner, H. J., Meyer, R., et al. (2005).
Manipulation of the spontaneous emission dynamics of quantum dots in two-
dimensional photonic crystals. Physical Review B, 71, 241304R.
Kuhn, H. (1970). Classical aspects of energy transfer in molecular systems. The Journal of
Chemical Physics, 53, 101.
Kuhn, A., Hennrich, M., Bondo, T., & Rempe, G. (1999). Controlled generation of single
photons from a strongly coupled atom-cavity system. Applied Physics B: Lasers and Optics,
69, 373.
Kuhn, A., Hennrich, M., & Rempe, G. (2002). Deterministic single-photon source for dis-
tributed quantum networking. Physical Review Letters, 89, 067901.
Kumar, R., Ong, J. R., Savanier, M., & Mookherjea, S. (2014). Controlling the spectrum of
photons generated on a silicon nanophotonic chip. Nature Communications. http://dx.doi.
org/10.1038/ncomms6489.
Kurtsiefer, C., Mayer, S., Zarda, P., & Weinfurter, H. (2000). Stable solid-state source of
single photons. Physical Review Letters, 85, 290.
Kwiat, P. G., Mattle, K., Weinfurter, H., Zeilinger, A., Sergienko, A. V., & Shih, Y. (1995).
New high-intensity source of polarization-entangled photon pairs. Physical Review
Letters, 75, 4337.
Kwiat, P. G., Waks, E., White, A. G., Appelbaum, I., & Eberhar, P. H. (1999). Ultrabright
source of polarization-entangled photons. Physical Review A, 60, 773(R).
Kyriienko, O., Shelykh, I. A., & Liew, T. C. H. (2014). Tunable single-photon emission
from dipolaritons. Physics Review A, 90, 033807.
Ladd, T. D., Jelezko, F., Laflamme, R., Nakamura, Y., Monroe, C., & OBrien, J. L. (2010).
Quantum computers. Nature, 464, 45.
Landau, L. D., & Peierls, R. (1930). Quantenelektrodynamik im konfigurationsraum.
Zeitschrift f
ur Physik, 62, 188.
Lasota, M., & Thew, R. (2014). Linear optics schemes for entanglement distribution with
realistic single-photon sources. Physics Review A, 90, 033836.
Laucht, A., P utz, S., Gunthner, T., Hauke, N., Saive, R., Frederick, S., et al. (2012).
A waveguide-coupled on-chip single-photon source. Physics Review X, 2, 011014.
Laurent, S., Varoutsis, S., Le Gratiet, L., Lematre, A., Sagnes, I., Raineri, F., et al. (2005).
Indistinguishable single photons from a single-quantum dot in a two-dimensional pho-
tonic crystal cavity. Applied Physics Letters, 87, 163107.
Law, C. K., & Eberly, J. H. (1996). Arbitrary control of a quantum electromagnetic field.
Physical Review Letters, 76, 1055.
Law, C. K., & Kimble, H. J. (1997). Deterministic generation of a bit-stream of single-
photon pulses. Journal of Modern Optics, 44, 2067.
108 Mario Bertolotti et al.
Lee, K. G., Chen, X. W., Eghlidi, H., Kukura, P., Lettow, R., Renn, A., et al. (2011).
A planar dielectric antenna for directional single-photon emission and near-unity collec-
tion efficiency. Nature Photonics, 5, 166.
Leistikow, M. D., Mosk, A. P., Yeganegi, E., Huisman, S. R., Lagendijk, A., & Vos, W. L.
(2011). Inhibited spontaneous emission of quantum dots observed in a 3D photonic band
gap. Physical Review Letters, 107, 193903.
Leonski, W. (1996). Fock states in a Kerr medium with parametric pumping. Physics Review
A, 54, 3369.
Leonski, W., & Tanas, R. (1994). Possibility of producing the one-photon state in a kicked
cavity with a nonlinear Kerr medium. Physical Review A, 49, R20.
Lewis, G. N. (1926). The conservation of photons. Nature, 118, 874.
Liebermeister, L., Petersen, F., Munchow, A., Burchardt, D., Hermelbracht, J., Tashima, T.
(2013). Tapered fiber coupling of single photons emitted by a deterministically posi-
tioned single nitrogen vacancy center. ArXiv, 1309.0421v2.
Liew, T. C. H., & Savona, V. (2010). Single photons from coupled quantum modes. Physical
Review Letters, 104, 183601.
Likharev, K. K. (1999). Single-electron devices and their applications. Proceedings of the IEEE,
87, 4.
Liu, Y-X, Wei, L. F., & Nori, F. (2004). Generation of nonclassical photon states using a
superconducting qubit in a microcavity. Europhysics Letters, 67, 941.
Liu, Y., Chen, G., Song, M., Ci, X., Wu, B., Wu, E., et al. (2013). Fabrication of nitrogen
vacancy color centers by femtosecond pulse laser illumination. Optics Express, 21, 12843.
Lochmann, A., Stock, E., Schulz, O., Hopfer, F., Bimberg, D., Haisler, V. A., et al. (2007).
Electrically driven quantum dot single photon source. Physica Status Solidi (c), 4, 547.
Lodahl, P., Mahmoodian, S., & Stobbe, S. (2013). Interfacing single photons and single quan-
tum dots with photonic nanostructures. arXiv:13121079v1.
Lodahl, P., & Stobbe, S. (2013). Solid-state quantum optics with quantum dots in photonic
nanostructures. Nanophotonics, 2, 39.
Lohrmann, A., Pezzagna, S., Dobrinets, I., Spinicelli, P., Jacques, V., Roch, J.-F., et al.
(2011). Diamond based light-emitting diode for visible single-photon emission at room
temperature. Applied Physics Letters, 99, 251106.
Loudon, R. (2000). The quantum theory of light. Oxford: Oxford University Press.
Lounis, B., Bechtel, H. A., Gerion, D., Alivisatos, P., & Moerner, W. E. (2000). Photon
antibunching in single CdSe/ZnS quantum dot fluorescence. Chemical Physics Letters,
329, 399.
Lounis, B., & Moerner, W. E. (2000). Single photons on demand from a single molecule at
room temperature. Nature, 407, 491.
Lounis, B., & Orrit, M. (2005). Single-photon sources. Reports on Progress in Physics, 68, 1129.
Lukin, M. D., & Imamoglu, A. (2001). Controlling photons using electromagnetically
induced transparency. Nature, 413, 273.
Lukishova, S. G., Bissell, L. J., Menon, V. M., Valappil, N., Hahn, M. A., Evans, C. M., et al.
(2009). Organic photonic bandgap microcavities doped with semiconductor nano-
crystals for room-temperature on-demand single-photon sources. Journal of Modern
Optics, 56, 167.
Lukishova, S. G., Schmid, A. W., Knox, R., Freivald, P., Bissell, L. J., Boyd, R. W., et al.
(2007). Room temperature source of single photons of definite polarization. Journal of
Modern Optics, 54, 417.
Lukishova, S. G., Schmid, A. W., Knox, R. P., Freivald, P., McNamara, A., Boyd, R. W.,
et al. (2006). Single-photon source for quantum information based on single dye mol-
ecule fluorescence in liquid crystal host. Molecular Crystals and Liquid Crystals, 454, 403.
Lukishova, S. G., Schmid, A. W., McNamara, A. J., Boyd, R. W., & Stroud, C. R., Jr.
(2003). Room temperature single-photon source: Single-dye molecule fluorescence
in liquid crystal host. IEEE Journal of Selected Topics in Quantum Electronics, 9, 1512.
Quantum State Engineering 109
Lukishova, S. G., Schmid, A. W., Supranowitz, C. M., Lippa, N., Mcnamara, A. J.,
Boyd, R. W., et al. (2004). Dye-doped cholesteric-liquid-crystal room-temperature
single-photon source. Journal of Modern Optics, 51, 1535.
Lukosz, W., & Kunz, R. E. (1977a). Light emission by magnetic and electric dipoles close
to a plane interface. I. Total radiated power. Journal of the Optical Society of America, 67,
1607.
Lukosz, W., & Kunz, R. E. (1977b). Light emission by magnetic and electric dipoles close
to a plane dielectric interface. II. Radiation patterns of perpendicular oriented dipoles.
Journal of the Optical Society of America, 67, 1615.
Lvovsky, A. I., & Mlynek, J. (2002). Quantum-optical catalysis: Generating nonclassical
states of light by means of linear optics. Physical Review Letters, 88, 250401.
Ma, X.-S., Zotter, S., Kofler, J., Jennewein, T., & Zeilinger, A. (2011). Experimental gen-
eration of single photons via active multiplexing. Physical Review A, 83, 043814.
Matre, X., Hagley, E., Nogues, G., Wunderlich, C., Goy, P., Brune, M., et al. (1997).
Quantum memory with a single photon in a cavity. Physical Review Letters, 79, 769.
Majumdar, A., Bajcsy, M., Rundquist, A., & Vuckovic, J. (2012). Loss-enabled sub-
Poissonian light generation in a bimodal nanocavity. Physical Review Letters, 108, 183601.
Majumdar, A., & Gerace, D. (2013). Single-photon blockade in doubly resonant
nanocavities with second-order nonlinearity. Physical Review B, 87, 235319.
Makhonin, M. N., Dixon, J. E., Coles, R. J., Royall, B., Luxmoore, I. J., Clarke, E., et al.
(2014). Waveguide coupled resonance fluorescence from on-chip quantum emitter.
Nano Letters, 14, 6997.
Maksymov, I. S., Miroshnichenko, A. E., & Kivshar, Y. S. (2012). Plasmonic nanoantennas
for efficient control of polarization-entangled photon pairs. Physical Review A, 86,
011801(R).
Mallek-Zouari, I., Buil, S., Quelin, X., Mahler, B., Dubertret, B., & Hermier, J.-P. (2010).
Plasmon assisted single photon emission of CdSe/CdS nanocrystals deposited on random
gold film. Applied Physics Letters, 97, 053109.
Mandel, L., & Wolf, E. (1995). Optical coherence and quantum optics. Cambridge: Cambridge
University Press.
Matsukevich, D., & Kuzmich, A. (2004). Quantum state transfer between matter and light.
Science, 306, 663.
Maurer, C., Becher, C., Russo, C., Eschner, J., & Blatt, R. (2004). A single-photon source
based on a single Ca + ion. New Journal of Physics, 6, 94.
McCusker, K. T., & Kwiat, P. G. (2009). Efficient optical quantum state engineering. Phys-
ical Review Letters, 103, 163602.
McCutcheon, M. W., Chang, D. E., Zhang, Y., Lukin, M. D., & Loncar, M. (2009). Broad-
band frequency conversion and shaping of single photons emitted from a nonlinear cav-
ity. Optics Express, 17, 22689.
McKeever, J., Boca, A., Boozer, A. D., Miller, R., Buck, J. R., Kuzmich, A., et al. (2004).
Deterministic generation of single photons from one atom trapped in a cavity. Science,
303, 1992.
Meschede, D., Walther, H., & M uller, G. (1985). One-atom maser. Physical Review Letters,
54, 551.
Meystre, P. (1992). Cavity quantum optics and the quantum measurement process. Progress in
Optics, 30, 261.
Meystre, P., Rempe, G., & Walther, H. (1988). Very-low-temperature behavior of a mic-
romaser. Optics Letters, 13, 1078.
Meystre, P., Slosser, J. J., & Wilkens, M. (1990). Cat in a cold niobium box. Optics Commu-
nications, 79, 300.
Michler, P., Imamoglu, A., Mason, M. D., Carson, P. J., Strouse, G. F., & Buratto, S. K.
(2000). Quantum correlation among photons from a single quantum dot at room tem-
perature. Nature, 406, 968.
110 Mario Bertolotti et al.
Michler, P., Kiraz, A., Becher, C., Schoenfeld, W. V., Petroff, P. M., Zhang, L., et al. (2000).
A quantum dot single-photon turnstile device. Science, 290, 2282.
Migdall, A. L., Branning, D., & Castelletto, S. (2002). Tailoring single-photon and
multiphoton probabilities of a single-photon on-demand source. Physics Review A, 66,
053805.
Mirin, R. P. (2004). Photon antibunching at high temperature from a single InGaAs/GaAs
quantum dot. Applied Physics Letters, 84, 1260.
Mizuochi, N., Makino, T., Kato, H., Takeuchi, D., Ogura, M., Okushi, H., et al. (2012).
Electrically driven single-photon source at room temperature in diamond. Nature Pho-
tonics, 6, 299.
Moerner, W. E. (2004). Single-photon sources based on single molecules in solids. New Jour-
nal of Physics, 6, 88.
Moerner, W. E., & Fromm, D. P. (2003). Methods of single-molecule fluorescence spectros-
copy and microscopy. The Review of Scientific Instruments, 74, 3597.
Mollow, B. R. (1969). Power spectrum of light scattered by two-level systems. Physics
Review, 188, 1969.
Moreau, E., Robert, J. M., Gerard, J. M., Abram, L., Manin, L., & Thierry-Mieg, V. (2001).
Single-mode solid-state single photon source based on isolated quantum dots in pillar
microcavities. Applied Physics Letters, 79, 2865.
Moreau, E., Robert, J. M., Manin, L., Thierry-Mieg, V., Gerard, J. M., & Abram, I. (2001).
Quantum cascade of photons in semiconductor quantum dots. Physical Review Letters, 87,
183601.
Mosley, P. J., Lundeen, J. S., Smith, B. J., Wasylezyk, P., URen, A. B., Silberhorn, C., et al.
(2008). Heralded generation of ultrafast single photons in pure quantum states. Physical
Review Letters, 100, 133601.
Muller, A., Fang, W., Lawall, J., & Solomon, G. S. (2009). Creating polarization-entangled
photon pairs from a semiconductor quantum dot using the optical Stark effect. Physical
Review Letters, 103, 217402.
Naik, D. S., Peterson, C. G., White, A. G., Berglund, A. J., & Kwiat, P. G. (2000). Entangled
state quantum cryptography: Eavesdropping on the Ekert protocol. Physical Review Let-
ters, 84, 4733.
Naiki, H., Masuo, S., Machida, S., & Itaya, A. (2011). Single-photon emission behavior of
isolated CdSe/ZnS quantum dots interacting with the localized surface plasmon reso-
nance of silver nanoparticles. Journal of Physical Chemistry C, 115, 23299.
Neu, E., Arend, C., Gross, E., Guldner, F., Hepp, C., Steinmetz, D., et al. (2011). Narrow-
band fluorescent nanodiamonds produced from chemical vapor deposition films. Applied
Physics Letters, 98, 243107.
Neu, E., Guldner, F., Arend, C., Liang, Y., Ghodbane, S., Sternschulte, H., et al. (2013).
Low temperature investigations and surface treatments of colloidal narrowband fluores-
cent nanodiamonds. Journal of Applied Physics, 113, 203507.
Nisbet-Jones, P. B. R., Dilley, J., Ljunggren, D., & Kuhn, A. (2011). Highly efficient source
for indistinguishable single photons of controlled shape. New Journal of Physics, 13,
103036.
Nisbet-Jones, P. B. R., Dillrey, J., Holleczek, A., Barter, O., & Kuhn, A. (2013). Photonic
qubits, qutrits and ququads accurately prepared and delivered on demand. New Journal of
Physics, 15, 053007.
Nowak, A. K., Portalupi, S. L., Giesz, V., Gazzano, O., Dal Savio, C., Braun, P.-F., et al.
(2014). Deterministic and electrically tunable bright single-photon source. Nature Com-
munications, 5, 3240. http://dx.doi.org/10.1038/ncomms4240.
Oppenheimer, J. R. (1931). Note on light quanta and the electromagnetic field. Physics
Review, 38, 725.
Oxborrow, M., & Sinclair, A. G. (2005). Single-photon sources. Contemporary Physics,
46, 173.
Quantum State Engineering 111
Parkins, A. S., Marte, P., Zoller, P., Carnal, O., & Kimble, H. J. (1995). Quantum-
state mapping between multilevel atoms and cavity light fields. Physics Review A,
51, 1578.
Parkins, A. S., Marte, P., Zoller, P., & Kimble, H. J. (1993). Synthesis of arbitrary
quantum states via adiabatic transfer of Zeeman coherence. Physical Review Letters, 71,
3095.
Patel, R. B., Bennett, A. J., Farrer, I., Nicoll, C. A., Ritchie, D. A., & Shields, A. J. (2010).
Two-photon interference of the emission from electrically tunable remote quantum
dots. Nature Photonics, 4, 632.
Pauli, W. (1933). Principles of Wave Mechanics. Handbuch der Physik (Vol. 24).
Pegg, D. T., Phillips, L. S., & Barnett, S. M. (1998). Optical state truncation by projection
synthesis. Physical Review Letters, 81, 1604.
Pelton, M., Santori, C., Vuckovic, J., Zhang, B., Solomon, G. S., Plant, J., et al. (2002).
Efficient source of single photons: A single quantum dot in a micropost microcavity.
Physical Review Letters, 89, 233602.
Perina, J. (1984). Quantum statistics of linear and nonlinear optical phenomena. Dordrecht:
D. Reidel Publishing Company.
Perina, J., Hradil, Z., & Jur, B. (1994). Quantum optics and fundamentals of physics. London:
Kluwer Academic Publishers.
Pinkse, P. W. H., Fischer, T., Maunz, P., & Rempe, G. (2000). Trapping an atom with single
photons. Nature, 404, 365.
Pittman, T. B., Franson, J. D., & Jacobs, B. C. (2007). Investigation of a single-photon source
based on quantum interference. New Journal of Physics, 9, 195.
Pittman, T. B., Jacobs, B. C., & Franson, J. D. (2002). Single photons on pseudodemand
from stored parametric down-conversion. Physics Review A, 66, 042303.
Planck, M. (1900a). On an improvement of Wiens equation for the spectrum. Verhandlungen
der Deutsche Physikalische Gesellschaft, 2, 202.
Planck, M. (1900b). On the theory of the energy distribution law of the normal spectrum.
Verhandlungen der Deutsche Physikalische Gesellschaft, 2, 237.
Poddubny, A. N., Ginzburg, P., Belov, P. A., Zayats, A. V., & Kivshar, Y. S. (2012). Tai-
loring and enhancing spontaneous two-photon emission using resonant plasmonic
nanostructures. Physics Review A, 86, 033826.
Politi, A., Cryan, M. J., Rarity, J. G., Yu, S., & OBrien, J. L. (2008). Silica-on-silicon wave-
guide quantum circuits. Science, 320, 646.
Press, D., Gotzinger, S., Reitzenstein, S., Hofmann, C., L offler, A., Kamp, M., et al. (2007).
Photon antibunching from a single quantum-dot-microcavity system in the strong cou-
pling regime. Physical Review Letters, 98, 117402.
Purcell, E. M. (1946). Spontaneous emission probabilities at radio frequencies. Physics Review,
69, 681.
Quartieri, A., Morello, G., Spinicelli, P., Todaro, M. T., Stomeo, T., Martiradonna, L., et al.
(2009). Nonclassical emission from single colloidal nanocrystals in a microcavity:
A route towards room temperature single photon sources. New Journal of Physics, 11,
033025.
Rabeau, J. R., Chin, Y. L., Prawer, S., Jelezko, F., Gaebel, T., & Wrachtrup, J. (2005). Fab-
rication of single nickel-nitrogen defects in diamond by chemical vapor deposition.
Applied Physics Letters, 86, 131926.
Raimond, J. M., Brune, M., & Haroche, S. (2001). Manipulating quantum entanglement
with atoms and photons in a cavity. Reviews of Modern Physics, 73, 565.
Raimond, J. M., Brune, M., Lepape, J., & Haroche, S. (1989). Rydberg atoms in a cavity:
A new method to generate photon number states. In M. S. Feld, J. E. Thomas, &
A. Mooradian (Eds.), Laser spectroscopy IX (p. 140). New York: Academic Press.
Raimond, J. M., Meunier, T., Bertet, P., Gleyzes, S., Maioli, P., Auffeves, A., et al. (2005).
Probing a quantum field in a photon box. Journal de Physique, B38, S535.
112 Mario Bertolotti et al.
Rakher, M. T., Ma, L., Davanco, M., Slattery, O., Tang, X., & Srinivasan, K. (2011).
Simultaneous wavelength translation and amplitude modulation of single photons from
a quantum dot. Physical Review Letters, 107, 083602.
Rakher, M. T., Ma, L., Slattery, O., Tang, X., & Srinivasan, K. (2010). Quantum transduc-
tion of telecommunications-band single photons from a quantum dot by frequency
upconversion. Nature Photonics, 4, 786.
Rakher, M. T., & Srinivasan, K. (2011). Subnanosecond electro-optic modulation of
triggered single photons from a quantum dot. Applied Physics Letters, 98, 211103.
Rakher, M. T., Stoltz, N. G., Coldren, L. A., Petroff, P. M., & Bouwmeester, D. (2008).
Polarization-switchable single photon source using the Stark effect. Applied Physics
Letters, 93, 091118.
Rarity, J. G., Tapster, P. R., & Jakeman, E. (1987). Observation of sub-Poissonian light in
parametric downconversion. Optics Communications, 62, 201.
Regelman, D. V., Mizrahi, U., Gershoni, D., Ehrenfreund, E., Schoenfeld, W. V., &
Petroff, P. M. (2001). Semiconductor quantum dot: A quantum light source of multi-
color photons with tunable statistics. Physical Review Letters, 87, 257401.
Reimer, M. E., Bulgarini, G., Akopian, N., Hocevar, M., Bavinck, M. B., Verheijen, M. A.,
et al. (2012). Bright single-photon sources in bottom-up tailored nanowires. Nature Com-
munications, 3, 737. http://dx.doi.org/10.1038/1746(2012).
Reinhard, A., Volz, T., Winger, M., Badolato, A., Hennessy, K. J., Hu, E. L., et al. (2012).
Strongly correlated photons on a chip. Nature Photonics, 6, 93.
Reischle, M., Beirne, G. J., Schulz, W.-M., Eichfelder, M., Robach, R., Jetter, M., et al.
(2008). Electrically pumped single-photon emission in the visible spectral range up to
80 K. Optics Express, 16, 12771.
Reischle, M., Kessler, C., Schulz, W.-M., Eichfelder, M., Robach, R., Jetter, M., et al.
(2010). Triggered single-photon emission from electrically excited quantum dots in
the red spectral range. Applied Physics Letters, 97, 143513.
Reiserer, A., & Rempe, G. (2014). Cavity-based quantum networks with single atoms and
optical photons. ArXiv, 1412.2889v1.
Rivas, D., Munoz-Matutano, G., Canet-Ferrer, J., Garca-Calzada, R., Trevisi, G.,
Seravalli, L., et al. (2014). Two-color single-photon emission from InAs quantum dots:
Toward logic information management using quantum light. Nano Letters, 14, 456.
Rogers, L. J., Doherty, M. W., Barson, M. S. J., Onoda, S., Ohshima, T., & Manson, N. B.
(2015). Singlet levels of the NV-centre in diamond. New Journal of Physics, 17, 013048.
Robach, R., Reischle, M., Beirne, G. J., Jetter, M., & Michler, P. (2008). Red single-
photon emission from an InP/GaInP quantum dot embedded in a planar monolithic
microcavity. Applied Physics Letters, 92, 071105.
Sandstrom, R. G., Shimoni, O., Martin, A. A., & Aharonovich, I. (2014). Study of narrowband
single photon emitters in polycrystalline diamond films. Applied Physics Letters, 105, 181104.
Santori, C., Fattal, D., Pelton, M., Solomon, G. S., & Yamamoto, Y. (2002). Polarization-
correlated photon pairs from a single quantum dot. Physics Review B, 66, 045308.
Santori, C., Fattal, D., Vuckovic, J., Solomon, G. S., & Yamamoto, Y. (2002). Indistinguish-
able photons from a single-photon device. Nature, 419, 594.
Santori, C., Fattal, D., Vuckovic, J., Solomon, G. S., & Yamamoto, Y. (2004). Single-
photon generation with InAs quantum dots. New Journal of Physics, 6, 89.
Santori, C., Fattal, D., & Yamamoto, Y. (2010). Single-photon devices and applications. Leipzig:
Wiley.
Santori, C., Pelton, M., Solomon, G. S., Dale, Y., & Yamamoto, Y. (2001). Triggered single
photons from a quantum dot. Physical Review Letters, 86, 1502.
Santori, C., Solomon, G. S., Pelton, M., & Yamamoto, Y. (2002). Time-resolved spectroscopy
of multiexcitonic decay in an InAs quantum Dot. Physics Review B, 65, 073310.
Quantum State Engineering 113
Savage, C. M., Braunstein, S. L., & Walls, D. F. (1990). Macroscopic quantum superpositions
by means of single-atom dispersion. Optics Letters, 15, 628.
Sayrin, C., Dotsenko, I., Gleyzes, S., Brune, M., Raimond, J. M., & Haroche, S. (2012).
Optimal time-resolved photon number distribution reconstruction of a cavity field by
maximum likelihood. New Journal of Physics, 14, 115007.
Scheel, S. (2009). Single-photon sourcesAn introduction. Journal of Modern Optics,
56, 141.
Scheel, S., Florescu, M., Haffner, H., Lee, H., Strekalov, D. V., Knight, P. L., et al. (2007).
Single photons on demand from tunable 3D photonic band-gap structures. Journal of
Modern Optics, 54, 409.
Schietinger, S., Barth, M., Aichele, T., & Benson, O. (2009). Plasmon-enhanced single pho-
ton emission from a nanoassembled metal-diamond hybrid structure at room tempera-
ture. Nano Letters, 9, 1694.
Schmidgall, E. R., Schwartz, I., Gantz, L., Cogan, D., Raindel, S., & Gershoni, D. (2014).
Deterministic generation of a quantum-dot-confined triexciton and its radiative decay
via three-photon cascade. Physical Review B, 90, 241411(R).
Schroder, T., Gadeke, F., Banholzer, M. J., & Benson, O. (2011). Ultrabright and efficient
single-photon generation based on nitrogen-vacancy centres in nanodiamonds on a solid
immersion lens. New Journal of Physics, 13, 055017.
Schwagmann, A., Kalliakos, S., Farrer, I., Griffiths, J. P., Jones, G. A. C., Ritchie, D. A., et al.
(2011). On-chip single photon emission from an integrated semiconductor quantum dot
into a photonic crystal waveguide. Applied Physics Letters, 99, 261108.
Sciscione, L., Centini, M., Sibilia, C., Bertolotti, M., & Scalora, M. (2006). Entangled,
guided photon generation in (1+1)-dimensional photonic crystals. Physics Review A,
74, 013815.
Scully, M. O., & Zubairy, M. S. (1997). Quantum optics. Cambridge: Cambridge University
Press.
Sebald, K., Michler, P., Passow, T., Hommel, D., Bacher, G., & Forchel, A. (2002).
Single-photon emission of CdSe quantum dots at temperatures up to 200K. Applied
Physics Letters, 81, 2920.
Settimi, A., Severini, S., Centini, M., Sibilia, C., Bertolotti, M., Napoli, A., et al. (2005).
Coherent control of stimulated emission inside one-dimensional photonic crystals.
Physics Review E, 71, 066606.
Shan, G.-C., Yin, Z.-Q., Shek, C. H., & Huang, W. (2014). Single photon sources with
single semiconductor quantum dots. Frontiers of Physics, 9, 170.
Shapiro, J. H., & Wong, F. N. (2007). On-demand single-photon generation using a
modular array of parametric downconverters with electro-optic polarization controls.
Optics Letters, 32, 2698.
Shen, Y., Bradford, M., & Shen, J.-T. (2011). Single-photon diode by exploiting the photon
polarization in a waveguide. Physical Review Letters, 107, 173902.
Shields, A. J. (2007). Semiconductor quantum light sources. Nature Photonics, 1, 215.
Short, R., & Mandel, L. (1983). Observation of sub-Poissonian photon statistics. Physical
Review Letters, 51, 384.
Simon, J., Tanji, H., Ghosh, S., & Vuletic, V. (2007). Single-photon bus connecting spin-
wave quantum memories. Nature Physics, 3, 765.
Singh, S. (1983). Antibunching, sub-Poissonian photon statistics and finite bandwidth effects
in resonance fluorescence. Optics Communications, 44, 254.
Slosser, J. J., Meystre, P., & Wright, E. M. (1990). Generation of macroscopic superpositions
in a micromaser. Optics Letters, 15, 233.
Smolyaninov, I. I., Zayats, A. V., Gungor, A., & Davis, C. C. (2002). Single-photon tunnel-
ing via localized surface plasmons. Physical Review Letters, 88, 187402.
114 Mario Bertolotti et al.
Solomon, G. S., Pelton, M., & Yamamoto, Y. (2001). Single-mode spontaneous emission
from a single quantum dot in a three-dimensional microcavity. Physical Review Letters,
86, 3903.
Sotier, F., Thomay, T., Hanke, T., Korger, J., Mahapatra, S., Frey, A., et al. (2009). Fem-
tosecond few-fermion dynamics and deterministic single-photon gain in a quantum dot.
Nature Physics, 5, 352.
Soujaeff, A., Nishioka, T., Hasegawa, T., Takeuchi, S., Tsurumaru, T., Sasaki, K., et al.
(2007). Quantum key distribution at 1550 nm using a pulse heralded single photon
source. Optics Express, 15, 726.
Spring, J. B., Salter, P. S., Metcalf, B. J., Humphreys, P. C., Moore, M., Thomas-Peter, N.,
et al. (2013). On-chip low loss heralded source of pure single photons. Optics Express, 21,
13522.
Srivathsan, B., Gulati, G. K., Chng, B., Maslennikov, G., Matsukevich, D., & Kurtsiefer, C.
(2013). Narrow band source of transform-limited photon pairs via four-wave mixing in a
cold atomic ensemble. Physical Review Letters, 111, 123602.
Steiner, M., Hartschuh, A., Korlacki, R., & Meixner, A. J. (2007). Highly efficient,
tunable single photon source based on single molecules. Applied Physics Letters, 90,
183122.
Stoler, D., & Yurke, B. (1986). Generating antibunched light from the output of a non-
degenerate frequency converter. Physics Review A, 34, 3143.
Strauf, S., Michler, P., Klude, M., Hommel, D., Bacher, G., & Forchel, A. (2002). Quantum
optical studies on individual acceptor bound excitons in a semiconductor. Physical Review
Letters, 89, 177403.
Strauf, S., Stoltz, N. G., Rakher, M. T., Coldren, L. A., Petroff, P. M., & Bouwmeester, D.
(2007). High frequency single photon source with polarization control. Nature Photonics,
1, 704.
Strekalov, D. V., Y-Ho, Kim, & Shih, Y. (1999). Experimental study of a subsystem in an
entangled two-photon state. Physical Review A, 60, 2685.
Takemoto, K., Sakuma, Y., Hirose, S., Usuki, T., Yokoyama, N., Miyazawa, T., et al.
(2004). Non-classical photon emission from a single InAs/InP quantum dot in the
1.3-m optical-fiber band. Japanese Journal of Applied Physics, 43, L993.
Takemoto, K., Takatsu, M., Hirose, S., Yokoyama, N., Sakuma, Y., Usuki, T., et al. (2007).
An optical horn structure for single-photon source using quantum dots at telecommu-
nication wavelengths. Journal of Applied Physics, 101, 081720.
Tame, M. S., Lee, C., Lee, J., Ballester, D., Paternostro, M., Zayats, A. V., et al. (2008). Single-
photon excitation of surface plasmon polaritons. Physical Review Letters, 101, 190504.
Tanji, H., Ghosh, S., Simon, J., Bloom, B., & Vuleti, V. (2009). Heralded single-magnon
quantum memory for photon polarization states. Physical Review Letters, 103, 043601.
Taylor, G. I. (1909). Interference fringes with feeble light. Proceedings of the Cambridge Phil-
osophical Society, 15, 114.
Thompson, R. M., Stevenson, R. M., Shields, A. J., Farrer, I., Lobo, C. J., Ritchie, D. A.,
et al. (2001). Single-photon emission from exciton complexes in individual quantum
dots. Physical Review B, 64, 201302(R).
Thomson, J. J. (1907). On the ionization of gases by ultra-violet light and on the evidence as
to the structure of light afforded by its electrical effects. Proceedings of the Cambridge Phil-
osophical Society, 14, 417.
Tittel, W., Brendel, J., Zbinden, H., & Gisin, N. (2000). Quantum cryptography using
entangled photons in energy-time bell states. Physical Review Letters, 84, 4737.
Torres, J. P., Banaszek, K., & Walmsley, I. A. (2011). Engineering nonlinear optic sources of
photonic entanglement. Progress in Optics, 56, 227.
Quantum State Engineering 115
Treussart, F., Alleaume, R., Le Floch, V., Xiao, L. T., Courty, J.-M., & Roch, J.-F. (2002).
Direct measurement of the photon statistics of a triggered single photon source. Physical
Review Letters, 89, 093601.
Treussart, F., Clouqueur, A., Grossman, C., & Roch, J.-F. (2001). Photon antibunching in
the fluorescence of a single dye molecule embedded in a thin polymer film. Optics Letters,
26, 1504.
Tribu, A., Sallen, G., Aichele, T., Andre, R., Poizat, J.-P., Bougerol, C., et al. (2008).
A high-temperature single-photon source from nanowire quantum dots. Nano Letters,
8, 4326.
URen, A. B., Silberhorn, C., Banaszek, K., & Walmsley, I. A. (2004). Efficient conditional
preparation of high-fidelity single photon states for fiber-optic quantum networks. Phys-
ical Review Letters, 93, 093601.
Ugur, A., Kremling, S., Hatami, F., H ofling, S., Worschech, L., Forchel, A., et al. (2012).
Single-photon emitters based on epitaxial isolated InP/InGaP quantum dots. Applied
Physics Letters, 100, 023116.
Ulhaq, A., Weiler, S., Ulrich, S. M., Robach, R., Jetter, M., & Michler, P. (2012). Cas-
caded single-photon emission from the Mollow triplet sidebands of a quantum dot.
Nature Photonics, 6, 238.
Unitt, D. C., Bennett, A. J., Atkinson, P., Cooper, K., See, P., Gevaux, D., et al. (2005).
Quantum dots as single-photon sources for quantum information processing. Journal
of Optics B: Quantum and Semiclassical Optics, 7, S129.
Unitt, D. C., Bennett, A. J., Atkinson, P., Ritchie, D. A., & Shields, A. J. (2005). Polarization
control of quantum dot single-photon sources via a dipole-dependent Purcell effect.
Physics Review B, 72, 033318.
Vamivakas, A. N., Saleh, B. E. A., Sergienko, A. V., & Teich, M. C. (2004). Theory of spon-
taneous parametric down-conversion from photonic crystals. Physics Review A, 70,
043810.
Vamivakas, A. N., Zhao, Y., Lu, C.-Y., & Atat ure, M. (2009). Spin-resolved quantum-dot
resonance fluorescence. Nature Physics, 5, 198.
Van der Sar, T., Hagemeier, J., Pfaff, W., Heeres, E. C., Thon, S. M., Kim, H., et al. (2011).
Deterministic nanoassembly of a coupled quantum emitter-photonic crystal cavity sys-
tem. Applied Physics Letters, 98, 193103.
Van der Sar, T., Heeres, E. C., Dmochowski, G. M., de Lange, G., Robledo, L.,
Oosterkamp, T. H., et al. (2009). Nanopositioning of a diamond nanocrystal containing
a single nitrogen-vacancy defect center. Applied Physics Letters, 94, 173104.
Varcoe, B. T. H., Brattke, S., Weidinger, M., & Walther, H. (2000). Preparing pure photon
number states of the radiation field. Nature, 403, 743.
Vasilev, G. S., Ljunggren, D., & Kuhn, A. (2010). Single photons made-to-measure. New
Journal of Physics, 12, 063024.
Verger, A., Ciuti, C., & Carusotto, I. (2006). Polariton quantum blockade in a photonic dot.
Physics Review B, 73, 193306.
Versteegh, M. A. M., Reimer, M. E., Joens, K. D., Dalacu, D., Poole, P. J., Gulinatti, A.,
et al. (2014). Observation of strongly entangled photon pairs from a nanowire quantum
dot. Nature Communications. http://dx.doi.org/10.1038/ncomms6298.
Vogel, K., Akulin, V. M., & Schleich, W. P. (1993). Quantum state engineering of the radi-
ation field. Physical Review Letters, 71, 1816.
Vogel, W., & Wellentowitz, S. (2001). In J. Perina (Ed.), Coherence and statistics of photons and
atoms (p. 333). New York: Wiley.
Volz, J., Kurtsiefer, C., & Weinfurter, H. (2001). Compact all-solid-state source of
polarization-entangled photon pairs. Applied Physics Letters, 79, 869.
116 Mario Bertolotti et al.
Vuckovic, J., Fattal, D., Santori, C., Solomon, G. S., & Yamamoto, Y. (2003). Enhanced
single-photon emission from a quantum dot in a micropost microcavity. Applied Physics
Letters, 82, 3596.
Waks, E., Diamanti, E., & Yamamoto, Y. (2006). Generation of photon number states. New
Journal of Physics, 8, 4.
Waks, E., Santori, C., & Yamamoto, Y. (2002). Security aspects of quantum key distribution
with sub-Poisson light. Physics Review A, 66, 042315.
Walls, D. F. (1979). Evidence for the quantum nature of light. Nature, 280, 451.
Walls, D. F., & Milburn, G. J. (1995). Quantum optics. Berlin: Springer.
Walmsley, I. A., & Raymer, M. G. (2005). Toward quantum-information processing with
photons. Science, 307, 1733.
Walther, H. (1998). Single atom experiments in cavities and traps. Proceedings of the Royal Soci-
ety of London, Series A, 454, 431.
Walther, H. (2002). Generation and detection of Fock states of the radiation field. Journal of
Optics B: Quantum and Semiclassical Optics, 4, S418.
Walther, H. (2004). Generation of photon number states on demand. Journal of Modern Optics,
51, 1859.
Wang, X.-H., Gu, B.-Y., Wang, R., & Xu, H.-Q. (2003). Decay kinetic properties of atoms
in photonic crystals with absolute gaps. Physical Review Letters, 91, 113904.
Wang, C., Kurtsiefer, C., Weinfurter, H., & Burchard, B. (2006). Single photon emission
from SiV centres in diamond produced by ion implantation. Journal of Physics B: Atomic,
Molecular and Optical Physics, 39, 37.
Wang, D., Li, M., Zhu, F., Yin, Z.-Q., Chen, W., Han, Z.-F., et al. (2014). Quantum key
distribution with the single-photon-added coherent source. Physics Review A, 90,
062315.
Wang, J., Santamato, A., Jiang, P., Bonneau, D., Engin, E., Silverstone, J. W., et al. (2014).
Gallium arsenide (GaAs) quantum photonic waveguide circuits. Optics Communications,
327, 49.
Ward, M. B., Farrow, T., See, P., Yuan, Z. L., Karimov, O. Z., Bennett, A. J., et al. (2007).
Electrically driven telecommunication wavelength single-photon source. Applied Physics
Letters, 90, 063512.
Weidinger, M., Varcoe, B. T. H., Heerlein, R., & Walther, H. (1999). Trapping states in the
micromaser. Physical Review Letters, 82, 3795.
Weiler, S., Stojanovic, D., Ulrich, S. M., Jetter, M., & Michler, P. (2013). Postselected indis-
tinguishable single-photon emission from the Mollow triplet sidebands of a resonantly
excited quantum dot. Physical Review B, 87, 241302(R).
Weisskopf, V. (1931). Zur theorie der resonanzfluoreszenz. Annalen der Physik (Leipzig), 9, 23.
Werner, M. J., & Imamoglu, A. (1999). Photon-photon interactions in cavity electromag-
netically induced transparency. Physical Review A, 61, 011801(R).
Wilk, T., Webster, S. C., Specht, H. P., Rempe, G., & Kuhn, A. (2007). Polarization-
controlled single photons. Physical Review Letters, 98, 063601.
Wolters, J., Schell, A. W., Kewes, G., N usse, N., Schoengen, M., D oscher, H., et al. (2010).
Enhancement of the zero phonon line emission from a single nitrogen vacancy center in a
nanodiamond via coupling to a photonic crystal cavity. Applied Physics Letters, 97, 141108.
Wootters, W. K., & Zurek, W. H. (1982). A single quantum cannot be cloned. Nature, 299, 802.
Wu, E., Jacques, V., Zeng, H., Grangier, P., Treussart, F., & Roch, J.-F. (2006). Narrow-
band single-photon emission in the near infrared for quantum key distribution. Optics
Express, 14, 1296.
Wu, E., Rabeau, J. R., Roger, G., Treussart, F., Zeng, H., Grangier, P., et al. (2007). Room
temperature triggered single-photon source in the near infrared. New Journal of Physics,
9, 434.
Quantum State Engineering 117
XingSheng, X., Xu, X., Chen, S., & Yamada, T. (2010). Antibunching and blinking from
a single colloidal CdSe quantum dot. Science China Physics, Mechanics & Astronomy,
53, 1619.
Xu, X., Chen, H., Yamada, T., & Otomo, A. (2008). Single-photon source in strong
photon-atom interaction regime in quasiperiodic photonic crystals. Europhysics Letters,
84, 67003.
Yablonovitch, E. (1987). Inhibited spontaneous emission in solid-state physics and electron-
ics. Physical Review Letters, 58, 2059.
Yamamoto, Y., Machida, S., Horikoshi, Y., & Igeta, K. (1991). Enhanced and inhibited
spontaneous emission of the free excitons in GaAs quantum wells in microcavity. Optics
Communications, 80, 337.
Yang, Y., Fleischhauer, M., & Zhu, S.-Y. (2003). Spontaneous emission from a two-level
atom in two-band anisotropic photonic crystals. Physics Review A, 68, 043805.
Yang, Y., & Zhu, S. Y. (2000). Spontaneous emission from a three-level atom in two-band
photonic crystals. Journal of Modern Optics, 47, 1513.
Yokoyama, H., Nishi, K., Anan, T., Yamada, H., Boorson, S. D., & Ippen, P. (1990).
Enhanced spontaneous emission from GaAs quantum wells in monolithic microcavities.
Applied Physics Letters, 57, 2814.
Young, R. J., Stevenson, R. M., Atkinson, P., Cooper, K., Ritchie, D. A., & Shields, A. J.
(2006). Improved fidelity of triggered entangled photons from single quantum dots. New
Journal of Physics, 8, 29.
Young, R. J., Stevenson, R. M., & Shields, A. J. (2007). Entangled photons from the
biexciton cascade of quantum dots. Journal of Applied Physics, 101, 081711.
Yuan, Z., Kardynal, B. E., Stevenson, R. M., Shields, A. J., Lobo, C. J., Cooper, K., et al.
(2002). Electrically driven single-photon source. Science, 295, 102.
Yuen, H. P. (1986). Generation, detection, and application of high-intensity photon
number-eigenstate fields. Physical Review Letters, 56, 2176.
Zaitsev, A. M. (2001). Optical properties of diamond: A data handbook. Berlin, New York:
Springer Science & Business Media.
Zavatta, A., Viciani, S., & Bellini, M. (2004). Quantum-to-classical transition with single
photon-added coherent states of light. Science, 306, 660.
Zhang, Q., Dang, C., Urabe, H., Wang, J., Sun, S., & Nurmikko, A. (2008). Large ordered
arrays of single photon sources based on IIVI semiconductor colloidal quantum dot.
Optics Express, 16, 19592.
Zinoni, C., Alloing, B., Monat, C., Zwiller, V., Li, L. H., Fiore, A., et al. (2006). Time-
resolved and antibunching experiments on single quantum dots at 1300 nm. Applied
Physics Letters, 8, 131102.
Zwiller, V., Blom, H., Jonsson, P., Panev, N., Jeppesen, S., Tsegaye, T., et al. (2001). Single
quantum dots emit single photons at a time: Antibunching experiments. Applied Physics
Letters, 78, 2476.
This page intentionally left blank
CHAPTER TWO
Contents
1. General Introduction 119
2. Part A: Digital Versus Optical Processing 121
2.1 Introduction 121
2.2 From Light to Image Formation 122
2.3 Digital Image Processing 124
2.4 Optical Image Processing 141
3. Part B: OIP 169
3.1 Optical Fourier Transform for Medical Image Processing and Phase Contrast
Imaging 169
3.2 Optical Implementation of Mathematical Operations 180
3.3 Techniques for Increasing Image Quality 187
3.4 Using Polarization for Increasing Image Quality and Edge Detection 208
3.5 Techniques for Image Denoising, Fusion, and Reconstruction 231
3.6 Techniques for Image Transmission: Optical Compression and/or Optical
Encryption Methods 234
3.7 Summary and Future Prospects 245
Acknowledgments 246
References 247
1. GENERAL INTRODUCTION
Historically, research on image processing (IP) techniques brings
together many facets of optics and information theory and presents scientists
and engineers with an extremely important question: How can informa-
tion be encoded in an optical image? In recent years, techniques of optical
image processing (OIP) have demonstrated the utility of their unique prop-
erties in diverse applications such as compression and encryption (Alfalou &
Brosseau, 2009), holography (Nehmetallah & Banerjee, 2012; Testorf &
Lohmann, 2008), and identification and tracking (Alfalou & Brosseau,
2010a, 2010b; Alfalou, Brosseau, & Alam, 2013; Rosen, Katz, &
Brooker, 2009). As a result, researchers and engineers in the United States
and worldwide have invested significant effort in improving optical
processing (OP) of information technology. Breakthroughs have been
achieved by the well-documented correlation technique (Alfalou &
Brosseau, 2010a, 2010b; Alfalou, Brosseau, & Alam, 2013). In spite of the
aforementioned achievements, optical processing techniques continue to
suffer from the point of view of optical implementation.
While images are originally optical, digital processing (DP) is often real-
ized to fully exploit their information content. As the resources required for
all-optical processing come within experimental reach, it is desirable to
develop a toolbox sufficiently versatile to allow the implementation of a
wide class of optical schemes. Within this context, our aim is to review
the recent progress made in the field of OP of information. More specifi-
cally, we will consider techniques allowing us to increase image quality to
render them more useful for correlation and reconstruction applications.
The overall goal of this chapter is to review, from both historical and tech-
nical perspectives, various important facts that may have escaped the attention
of those interested in this area. Another aim of the review is to highlight
aspects in understanding the developments in optical setup that are particularly
relevant to OIP. In view of the magnitude of the area, which spans nearly half
a century of technical work in the field, it was not possible to cover every con-
ceivable subject. There have in the past been attempts to review topics rele-
vant to OIP and digital image processing (DIP), and some of these contain
useful information and perspectives. All however, due to the inexorable evo-
lution of the subject as researchers have struggled toward the ultimate goal of
understanding, and due to the vastness of the task, are of limited scope, and
clearly localized in time. As far as we are aware, there has not been a substantial
review in which correlation applications of OIP are the central topic. Taking
this into account, we apologize beforehand for the omission of important and
interesting developments that are beyond the scope of this chapter. Although
it is not possible to cover every topic in complete detail, sufficient details are
provided to make the review as self-contained and pedagogical as possible.
The structure of the chapter is as follows. In the first part, techniques for
increasing image quality are described in detail. As a point of interest, those
Recent Advances in Optical Image Processing 121
dealing with color encoding are addressed. In addition, methods for den-
oising images are dealt with. The remaining material fleshes out the advan-
tages and limitations of the optical methods by providing the requisite
details. The second part deals with a discussion of some current develop-
ments in OIP and their practical implementation. We then discuss central
theoretical and experimental aspects of polarization encoding methods.
Throughout this review, many examples illustrating the performances of
these techniques are given. These results illustrate the lesson that polarization
optics coupled to OIP can make important contributions to objects detec-
tion and recognition. Finally, we hope to communicate our sense of future
vistas for progress in this rapidly expanded field.
degradations of the linear homogeneous type, i.e., those in which the point
spread function characterizing the degradation is invariant over the image
plane to be examined, and to the case in which the image spectrum is
not band-limited. Within these limitations, the degraded image Id can be
described by the convolution integral
Id x, y I x, ysI x, y I x x0 ,y y0 sI x0 , y0 dx0 dy0 , (1)
1
x +X NNy 1
NNx 1 y +X
Id i, jsi x + 1, j y + 1
I x, yr + 1 I x, yr XX
n nn (3)
ix jy
I p, qr si p + 1, j q + 1
pm qmm
124 Ayman Alfalou and Christian Brosseau
and
x +X
NNx 1 y + X
NNy1
Id i, jsI i x + 1, j y + 1
I x, y1 X X
n nn , (4)
ix jy
sI i p + 1, j q + 1
pm qmm
where (Nx, Ny) is the dimension of the input image, (NNx, NNy) is the
dimension of the input point spread function, m Max f1; i k + 1g,
mm Max f1; j l + 1g, n Minfi; Nx g, nn Minf j; Ny , and r 1
denotes the iteration number. Nevertheless, the convergence issue remains
untreated in the original paper of Richardson (1972). It is noteworthy that
both Harris (1966) and Richardson (1972) pointed out the positive effect
that preprocessing treatments can have on image restoration.
(a) (b)
Before After
(p1 = t) d
(P2 = t+t)
Bright area Bright area
Figure 4 Illustration of the Moir effect by superimposing (a) two patterns made of par-
allel and equidistant lines and (b) example with an image. Goole (2013).
128 Ayman Alfalou and Christian Brosseau
If the lines of the patterns are superimposed at the left of Figure 4, the
shift between the lines increases when going to the right. After a given num-
ber of lines, the patterns are opposed: the lines of the second pattern are
between the lines of the first pattern. If we look from a far distance, we have
the impression of pale zones when the lines are superimposed (blue (dark
gray in the print version) zone in Figure 4) (there is white between the lines),
and of dark zones when the lines are opposed (red zone (gray in the print
version) in Figure 4). The middle of the first dark zone is when the shift
is equal to t/2. The nth line of the second pattern is shifted by nt0 compared
to the nth line of the first network. The middle of the first dark zone thus
corresponds to nt 0 t=2, that is,
t
n : (6)
2t 0
The distance d between the middle of a pale zone and a dark zone is
d nt t2 =2t0 : (7)
Thus, the bigger the step, the bigger the distance between the pale and
dark zones; the bigger the discrepancy t0 , the closer the dark and pale zones;
and a large spacing between dark and pale zones means that the patterns have
very close steps. The suppression of the undesired effects (dark and pale
zones) of this phenomenon can be done as follows (with reference to
Figure 5).
Yaroslavsky (2012) suggests applying an FT of the target image showing
the moire effect (Figure 5a). It is well known that the FT of diffraction grat-
ing results is a series of spikes (Goodman, 1966). These spikes can be filtered
out by an appropriate filter (Figure 5b). Applying the FT1 leads to
Figure 5c. Comparing Figure 5a and c indicates that the moire effect has
been attenuated by the filtering. Figure 5d shows the moire noise.
A final note is in order. The moire effect can also be beneficial; e.g., it can
be useful to estimate the depth of a 3D object. Several optical analyses and
applications of the moire effect have been considered in the literature
(Aggoun, 2006; Andonian, 1982; Bara, Jaroszewicz, Kolodziejczyk, &
Moreno, 1991; Buytaert & Dirckx, 2007; Kato et al., 1997; Khalil &
Fazio, 1973; ;Kim, Park, Jung, Kim, & Lee, 2009; Lebanon & Bruckstein,
2001; Lim, Kim, & Chung, 1989; Rasouli, Ghasemi, Tavassoly, &
Khalesifard, 2011; Simova & Stoev, 1992; Yokozeki & Ohnishi, 1975).
Recent Advances in Optical Image Processing 129
Figure 5 Illustration of the moir effect on a specific image. (a) Original image clearly
displaying the moir perturbation (framed in red (gray in the print version)), (b) the FT
transform showing the spikes due to the moir effect, (c) processed image for which the
characteristic moir frequencies are filtered out (framed in red (gray in the print
version)), and (d) Moir noise. Yaroslavsky (2012).
130 Ayman Alfalou and Christian Brosseau
of Badd, H*
invar is the conjugate FT of hinvar, and jj is the square of the mod-
2
PSD Iunknown
SNR (12a)
PSD Badd
!
1 jHinvar j2
HWiener : (12b)
Hinvar jHinvar j2 1=SNR
decreases.
Two specific cases are of interest. Figure 6 qualitatively illustrates the
effects of denoising of the dogs image (Figure 6a). Figure 6c shows the
Figure 6 Illustration of the Wiener filtering for noise suppression: (a) original noiseless
image. (b) Noisy image. (c) Filtered image. http://upload.wikimedia.org/wikipedia/com
mons/9/9d/Wiener_filter_-_my_dog.JPG.
132 Ayman Alfalou and Christian Brosseau
denoised image. As outlined in the output image shown in Figure 6c, the
different points (black and white pixels) perturbating the visibility of
the image shown in Figure 6b vanish. Noise suppression performances
depend on the type of noise, its density (SNR), and used architecture
(Baker, 1981; Berriel, Bescos, & Santisteban, 1983; Birch et al., 2001;
Hardie & Barnard, 2012; Ikeda, Sato, & Kojima, 1986; Inbar & Marom,
1995; Jiang & Sawchuk, 1986; Khoury, Gianino, & Woods, 2000;
Kondo, Ichioka, & Suzuki, 1977; Kutay & Ozaktas, 1998; Riris, Carlisle,
Warren, & Cooper, 1994; Zalevsky & Mendlovic, 1996).
A second example deals with the deblurring of image. We show in
Figure 7a an example of blurred image. The deblurred image obtained with
Wiener deconvolution is shown in Figure 7b. The comparison of Figure 7b
and c conclusively demonstrates the improvement provided by this method.
To extend our considerations, we briefly present another related archi-
tecture, i.e., the empirical Wiener filter, which can be used for image
deblurring (Yaroslavsky, 2012). Ghael, Sayeed, and Baraniuk (1997)
reported on an image denoising method using wavelets and empirical
Wiener techniques. The main phenomenon is depicted using an example
in Figure 8. Figure 8a depicts an image with Gaussian white noise of variance
set to 10. As Figure 8b illustrates this denoising technique is efficient. To
further validate this as an approach accurately denoising images, we compare
this with the noiseless original image (Figure 8c).
Figure 8 Illustration of denoising via empirical Wiener filtering. (a) Image with a white
Gaussian noise with variance set to 10, (b) denoised image using a wavelet shrinkage
estimate as a means to design a wavelet-domain Wiener filter, and (c) original image
(Ghael et al., 1997).
(a)
Iin
Input image And/or Modulation
Compression Encryption
(or input video sequence) (adaptation)
Transmission channel
(b)
Iout
Output image Decompression Decryption Demodulation
(or output video sequence)
2.3.4.1 Compression
Data compression refers to the process of reducing the amount of data
required to represent, store, and transmit given quantity of information
(image). There exists a large literature concerning compression techniques
(Barlaud & Labit, 2002; Pennebaker & Mitchell, 1993; Richardson, 2003;
Salomon, 2004; Sayood, 2012). Generally speaking, these techniques aim
at reducing the redundancy of information. Three kinds of redundance
can be distinguished: spectral, spatial, and temporal (Barlaud & Labit,
2002; Richardson, 2003; Salomon, 2004; Sayood, 2012). The work in
Pennebaker and Mitchell (1993), Barlaud and Labit (2002), Richardson
Recent Advances in Optical Image Processing 135
(2003), Salomon (2004), and Sayood (2012) and subsequent efforts have
already paved the way for a development of many compression protocols
such as Bitmap, Portable Network Graphics, Graphics Interchange Format,
Tagged Image File Format, Joint Photographic Experts Group, JPEG 2000,
Moving Picture Experts Group, and MPEG-4 AVC (Advanced Video Cod-
ing). As the amount of information to be transmitted becomes larger and
transmission speed becomes faster, data compression is becoming a critical
challenge in video imaging. The objective of image compression is to reduce
irrelevance and redundancy of the image data in order to be able to store or
transmit data in an efficient form. Certain information has relatively less
importance for the quality of image perception. This information is said
to be psychovisually redundant.
Image compression can be classified as lossy or lossless. Lossless compres-
sion, e.g., the Lempel-Ziv-Welch technique, is preferred for archival pur-
poses and is often used for medical imaging. Lossy compression methods,
e.g., JPEG, especially when used at low bit rates, introduce compression
artifacts. Lossy methods are especially suitable for natural images such as
photographs in applications where minor (sometimes imperceptible) loss
of fidelity is acceptable to achieve a substantial reduction in bit rate. In this
work, we focus on lossy data compression. In an attempt to cope with the
problem, several studies have been conducted in the past in academic
institutions and industry, especially for telecommunications, access control,
biometrics, and security systems.
Here, we are interested in the (lossy) two-step compression technique
depicted in Figure 10. The first step consists in applying a well-defined trans-
form to the image, i.e., mapping. This is a reversible process which is not
lossy. This transform is required to prepare and package the information
(redundant or pertinent). The choice of the specific transform is linked to
the application and the compression technique (Salomon, 2004). The sec-
ond step is the quantification during which redundant information is
suppressed. In order to be effective, quantification results from a compro-
mise between compression rate and quality of the reconstructed images. This
process reduces the accuracy and hence psychovisual redundancies of a given
image. This process is irreversible and therefore lossy. This quantification
step is followed by entropic coding when a fixed or variable-length code
is used to represent mapped and quantized data sets (see, e.g., Neji, Jridi,
Alfalou, & Masmoudi, 2013). This is a reversible process. It removes coding
redundancy by assigning the shortest codes to the most frequently occurring
output values.
136 Ayman Alfalou and Christian Brosseau
Constraints :
application, compression method, ...
Iin
Input image Mapping
(or input video sequence)
Entropic
Quantification Bitstream
coding
Constraints :
application, flow rate, compression rate
quality of reconstructed images, ...
Figure 10 Illustration of the compression block.
where SIin is the spectrum of the discrete image Iin of size M N pixels, and
A MN 1=2 is a normalization constant. For clarity, we will frame our
discussion around the specific one-dimensional version of Equation (13)
(Chang & Nguyen, 2008; Cooley & Tukey, 1965; Glenn, 1968; Kolba &
Parks, 1977; Oppenheim, Schafer, & Buck, 1999; Ouerhani, 2012;
Ouerhani, Jridi, & Alfalou, 2011, 2012; Sorensen, Heideman, & Burrus,
1986):
X
M 1 h xui
SIin u A Iin xexp 2j : (14)
x0
M
X X
1 N 1
1 N
2x + 1u 2y + 1v
SDCTIin u, v c uc v Iin x, ycos cos ,
4 0 0
2N 2N
(16)
p
1= 2 if i 0
where c i , SDCTIin is the DCT spectrum of Iin. For most
1 if i 0
images, much of the signal energy lies at low frequencies; these appear in the
upper left corner of the DCT. The lower right values represent higher fre-
quencies and are often small (usually small enough to be neglected with little
visible distortion). Figure 11a illustrates this point schematically.
As for FT, DCT requires a large number of arithmetic operations, i.e.,
typically N 2 N sums and N2 products. To minimize complexity,
number of arithmetic operations, computation time, and energy con-
sumption, a number of groups have proposed a variety of architectures
(Ahmed et al., 1974; Blinn, 1993; Chen, 2005; Chen et al., 1977; Hou,
1987; Meher, 2006; Pennebaker & Mitchell, 1993; Marom & Inbar,
1996; Slawecki & Li, 1992; Strang, 1999; Sungwook & Swartziander,
2001; White, 1989).
In prior work (Ouerhani, 2012), we validated FFT and DCT algorithms
for implementation on FPGA. The FFT architecture can be optimized using
the Radix-4 algorithm and presents the advantages of a higher throughput,
low area-delay product, and lower computational time. The novelty consists
on using a memory sharing and dividing technique along with parallel-in
parallel-out processing elements. The proposed architecture can perform
N-point FFT using only 4N/3 delay elements and involves a latency of
N/4 cycles. Comparison in terms of hardware complexity and area-delay
product with recent work presented in the literature and commercial IPs
was made to show the efficiency of the proposed design.
(a) (b)
Low frequency
Low frequency
Medium
Medium
High frequency
High frequency
Figure 11 Spectral representation of an image using: (a) DCT and (b) FT.
Recent Advances in Optical Image Processing 139
Light f
applications. Here, we will explore OIP techniques of images and show that
novel optical methods can be considered as substitutes, at least partially, to
many time-consuming numerical techniques.
(a)
(b)
Pixel
Figure 13 Illustration of a nematic modulator: (a) Picture of the SLM Seiko produced by
(a) CRL Opto/VGA3 and (b) a microscope view of the SLM Seiko.
700
50
600
500
100
Signal (mV)
Rows
400
150
300
200
200
100
250 0
50 100 150 200 250
Columns
Figure 15 Example of an image obtained by the detector suggested by Rolando et al.
(2013).
2.4.1.2 Imaging Systems: In Vivo Brain Optical Imaging System Using Vertical
Cavity Surface-Emitting Lasers as a Light Source
Another application that has been a recent focus of attention concerns imag-
ing systems. As pointed out by Atchia et al. (2013), optical methods have at
least five potential advantages compared to other methods such as functional
magnetic resonance imaging, positron emission tomography, and computed
tomography: (1) functional simplicity, (2) cost effectiveness, (3) noninvasive
nature, (4) good performances with spatiotemporal resolution, and (5) min-
iaturization for portable application. Two kinds of architecture exist (Atchia
et al., 2013). The first one is termed intrinsic optical signal imaging, while
the second one is based on laser speckle contrast imaging (LSCI). Based on
the latter, Atchia et al. (2013) demonstrated an imaging technique
implementing vertical cavity lasers with extremely low transient times
for a greatly simplified realization of a multiexposure laser speckle contrast
imaging system.
Recent Advances in Optical Image Processing 147
The LSCI technique allows one to realize a map of relative blood veloc-
ity in vivo through the interference of coherent light (Goodman, 1976). The
underlying principle of the technique consists in calculating a speckle pattern
from scattering events with coherent light excitation. The spatial speckle
contrast ratio K is expressed as
K , (17)
hI i
Figure 16 Experimental apparatus proposed by Atchia et al. (2013) for in vivo brain opti-
cal imaging of small animals.
148 Ayman Alfalou and Christian Brosseau
We note in passing that Atchia et al. (2013) also proposed different opti-
mization schemes to overcome the errors due to defocusing, and to deal with
the coherence length which is altered with changes in the VCSEL current.
Lens
q
R
DOF
f
Figure 18 Imaging geometry of a thin lens including the imaging spot. Ryle et al. (2013).
z2
z1
z
Figure 19 Schematic of the experiment reported by Ryle et al. (2013).
distance larger than the DOF cannot be resolved. To remedy this, Ryle et al.
(2013) suggested using the DIHM architecture. One basic advantage of such
system lies in its simplicity. Only standard optical components such as colli-
mated spatially coherent light source, MO, and digital camera are needed
along with a computer to process the resulting holograms.
Figure 19 illustrates schematically the optical setup composed of a HeNe
laser ( 632:8nm; 20 mW), a neutral density filter (NDF), a microscope
objective (20; MO1), a spatial filter (SF with a 15-m diameter pinhole),
a collimating lens (CL; f 20cm), the sample (located away from the focal
plane of the imaging 10(n sin )), another microscope objective
(MO2), and a CCD camera.
DIHM-based systems have been shown to overcome the limitations of
DOFs, i.e., by reconstruction of a number of images over many depths (Ryle
et al., 2013). Each image, reconstructed from a single hologram, has a limited
number of DOFs (Figure 22). Combining or stacking a series of such DOF
limited numerical reconstructions results in an image with an effectively
150 Ayman Alfalou and Christian Brosseau
Figure 20 Illustration of the optical setup as proposed in Mendlovic (2013). ISP means
image signal processing.
Recent Advances in Optical Image Processing 151
Figure 21 Schematic proposed par Mendlovic (2013) for designing a super imaging
camera.
Figure 22 Multiple aperture cameras: (A) is a microlens array, (B) is the rectangular aper-
ture channel array, (C) is a specific subchannel, and (D) is the detection imager.
Mendlovic (2013).
Conventional
Subject
Main lens
Photosensor
Light field
Microlens
array
Subject
Main lens
Photosensor
of the information distribution among the different field rays per each image
pixel. This is usually done by inserting a lens array in front of the sensing
array and, based on the ray model optics, obtaining the various fields in dif-
ferent pixels. Figure 4 shows a conventional imaging system and a light-field
recording system. In this setup, one can increase the pixel count by a factor
of M along each direction (Mendlovic, 2013).
The third application is the multiexposure camera. The basic idea is to
collect data from a few exposures and fuse it to create a better quality image.
The concept we have in mind is illustrated in Figure 24. A necessary step in
all multiexposure/multiaperture computational photography algorithms is
the frame alignment and fusion.
The purpose of this brief section is to consider in some detail the basic
principle of OC.
Figure 25 Synoptic diagram of the VLC. Alfalou, Brosseau, and Alam (2013).
154 Ayman Alfalou and Christian Brosseau
Figure 26 Examples of target image: (a) Barbara, (b) Lena, and (c) and (d) are the ampli-
tude spectra of Lena and Barbara, respectively.
2.4.3.3 Discussion
It is quite clear that successfully using correlation methods requires a good
knowledge of the application and basic requirements such as the kind of tar-
get images, expected levels of decision robustness and discrimination, real-
time processing, lightning conditions, and noise. Our results point out that
this set of parameters will impose the correlation architecture and the most
appropriate type of correlation filter for the application considered. Con-
sider, for example, the target images Lena and Barbara shown in
Figure 26a and b, respectively. Their respective FTs lead to significantly dif-
ferent spectra. While Lenas spectrum is almost centered (Figure 26c),
Barbaras spectrum shows a larger spatial distribution in the spectral plane.
The above-mentioned issues are extremely relevant here because the choice
of the correlation filter will impact correlators performances. For a face rec-
ognition application, the correlation filter should be able to deal with face
rotation with respect to the reference image. Specially designed correlation
filters to deal with color information have already demonstrated outstanding
performances for smart pattern application (Alfalou, Brosseau, &
Alam, 2013).
where k 2= denotes the wave vector and is the wavelength of the inci-
dent wave illuminating the transparency. If we suppose that z is chosen to be
so large that the phase factor P(x,y) 1 over the entire region of the (x,y)
plane in which Uo(x,y) is nonzero, Ui(u,v) is just the two-dimensional FT
of the aperture Uo(x,y), except for a multiplicative phase factor preceding
Recent Advances in Optical Image Processing 157
(a) FT 1
FT
2w 2w
P1 P2 P3
L1 L2
Identification
(b)
Target
Target image Correlation
image FT FT1 plane
Yes End
spectrum
No
Correlation filter
(Fn)
Figure 27 Illustrating the principle of the optical correlator and numerical implemen-
tation of the VLC (N_VLC). Elbouz et al. (2013)
the integral and which does not affect the intensity of the diffracted light, at
frequencies zu , zv .
In this paper, we consider the special case of a circular aperture with
diameter wl given by the pupil function
p!
x2 + y2
P x, y circ , (20)
wl
Figure 28 The face database used for tests. The image framed in red (gray in the print
version) is used as the reference image for the POF. Elbouz et al. (2013).
Figure 29 Correlation results obtained with different sets of parameters (W, L, f ). Elbouz
et al. (2013).
Recent Advances in Optical Image Processing 159
the lens aperture, L is the input plane size, and f denotes the focal length of
the lens. It is easily seen that the correlators performance is significantly
affected by the values of these parameters.
As a first case, we choose W 6.25 mm, f 0.125 m, M 250, and
L 25 mm. The 4f setup (Figure 30a) is used to reconstruct the input plane
with this set of parameters. Figure 30b shows the reconstructed image at the
output using NO_FT (Equation 19). The face is limited by the aperture 2W
which takes a smaller value than the size of the input plane. Hence, the
reconstructed image do contains only partly the input image.
Application of this scheme leads to erroneous correlation results. This
can be seen in Figure 31 for several target and reference images. The auto-
correlation plane is presented in Figure 31a, while the correlation results
corresponding to a target image and its variant rotated of 45 contain a cen-
tral peak which looks like a correlation peak. However, the peak height
shown in Figure 31b is still large. This is counterintuitive since the target
and reference images are significantly different. In like fashion, this peak
f f f f
FT 1
FT
2w 2w
P1 P2 P3
L1 L2
(a) (b)
2w
L L
Figure 31 Correlation results between target and reference images when L 2W.
Elbouz et al. (2013).
is present in Figure 31c and d. It should be noted that this peak is related to
the circular aperture (Figure 30b).
As a second case, we set W 17.975 mm, f 0.125 m, M 250, and
L 25 mm. To overcome the aperture issue, we assume now that
L 2W as shown in Figure 32; a same requirement is made for the other
lens. It should be noted that the length of the diagonal of the input plane
is equal to the pupils aperture. Application of the proposed reconstruction
scheme leads to disappearance of the circular pattern shown in Figure 31c.
The corresponding correlation results are now shown in Figure 33.
Figure 33a shows the autocorrelation with a numerical implementation of
Recent Advances in Optical Image Processing 161
f f f f
FT 1
FT
2w 2w
P1 P2 P3
L1 L2
(a)
2W
(b) (c)
L L
Figure 32 Reconstruction of an input plane case when L 2W. Elbouz et al. (2013).
Figure 33 Correlation results for two subjects of PHPID. W 17.975 mm, L 25 mm,
f 0.125 m, and M 250. Elbouz et al. (2013).
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
FPR FPR
Figure 34 Comparison between correlation results obtained with FFT and the
algorithm (NO_FT). Elbouz et al. (2013).
Zhao et al. (2013) is motivated by the fact that knowledge of the LCT allows
one to know important transforms used in IP, i.e., FT, fractional FT, (Ding,
2001) and Fresnel transform (Hennelly & Sheridan, 2005; Stern, 2006). The
unitary property of the LCT is of fundamental importance. The principle of
the LCT (Collins, 1970) is as follows. Consider
1 +1 j Ax2 2xy + Dy2
F y LM f x y p f xexp dx, (21)
jB 1 B
where f(x) is the input signal in the space domain, x, and F(y) is the output
signal in the LCT domain, y. Zhao et al. (2013) restrict the parameters of
A B
the ABCD matrix, M to be real numbers with det M 1.
C D
The FT can be calculated from Equation (21) by setting fA, B, C, Dg
f0,1, 1,0g in Equation (21). Thus,
s
1 +1
F y f xexpj 2xy dx: (22)
j 1
where Tx and Ty denote the periodic sampling intervals in the x space and y
LCT domains, respectively, n,m 2 P, + P , the number of sampling
points N is defined as N 2P + 1, Tx Lx =N , Lx is the extent of the input
signal f(x), and the elements of the kernel are
s
1 j n 2 o
WM n, m exp AnTx 2nTx mTy +D mTy
2
: (25)
jB B
where f(x,y) denotes the distribution function, is the Dirac delta function, x
and y are Cartesian coordinates, and r and describe a line at distance r from
the center at angle . The optimized setup is displayed in Figure 36.
As described in Paz et al. (2013), the system contains a telecentric lens, to
inspect the whole depth of the lattice simultaneously. The camera is
connected to the lens. An extended depth of field (EDOF) element can
be combined inside the lens. A qualitative comparison of Figure 37a and b
indicates the good performances of this approach. This technique allows
capturing information from different perspectives, so information can be
reconstructed using the RT. However, modulation of the high frequencies
with the extended DOF is low since with this system it is difficult to enhance
the signal with a deconvolution filter, as the intensity is close to the noise
level, so the accepted deconvolved pictures are a bit noisy.
Modified wavefront
Focus channel
Incoming
wavefront
Camera
Figure 36 Synoptic diagram of the experimental setup used for extending the depth
of focus.
Figure 37 DOF measurements with the resolution target (a) with the EDOF element at
locations 0 mm and (b) without the EDOF element at locations 0 mm.
Figure 38 Illustration of the trade-off between coregistration error and pixel count
(Skauli, 2013). The figure illustrates the pixel footprint for two bands, shown as red (light
gray in the print version) and blue (dark gray in the print version) rectangles, for two
cases. The left image considers small pixels and a spatial coregistration error that is a
large fraction of the pixel size. The right image deals with larger but fewer pixels.
N LtAw , (27)
hc
N LtAw : (28)
hc
Ncoreg
s N , (29)
6000
5000
es0 = 0.07
Lower capacity bound (bits/frame)
4000
3000
es0 = 0.12
2000
0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Binning factor b
Figure 39 Information capacity for one band of a spectral imager with varying sizes of
the detector pixels: b 1, 0.15, and N 20,000. Skauli (2013).
Recent Advances in Optical Image Processing 169
r
1 N
C N , s logN s : (30)
2 2
Using Equation (30), we show the resulting information capacity esti-
mate for one spectral band ( 0.15) in Figure 39. This graph illustrates
how the information capacity reaches an optimum for a particular pixel
count if the distortions of the optics are held fixed. It illustrates also that
coregistration error can lead to a large loss of information.
3. PART B: OIP
There has been palpable progress in the development of optical tech-
niques and tools to aid picture interpretation, improve image quality,
enhance image contours, and correct geometric distortions. In this second
part, we highlight an eclectic mix of important topics in OIP that we feel
have not been given sufficient air time. It is hoped that this collection
of excursions, though idiosyncratic, will help sharpen collective understand-
ing of the core issues to be tackled on the way to a complete and credible
understanding of the basic principles of OIP.
Fourier plane:
Spatial or phase
Filter
Collimated Amplitude Spatial Filtered/
Processed
optical or phase frequency modified
image
source object spectrum spectrum
Optical Inverse
Fourier Fourier
transform transform
Figure 40 4f OFT processing scheme. Yelleswarapu et al. (2008).
170 Ayman Alfalou and Christian Brosseau
CL1
L1 L2 NDF
532 nm CCD
laser camera
frequencies (Figure 41). For the purpose of filtering out the undesired low
spatial frequencies, Yelleswarapu et al. (2008) suggest adapting the standard
4f setup by introducing various spatial filters (Figure 41). As usual, the two
lenses correspond, respectively, to the FT and FT1 processing.
To exemplify the benefit of using the optical setup shown in Figure 37,
the processed image captured by the CCD camera displaying micro-
calcifications corresponding to the high spatial frequencies is displayed in
Figure 42b. This picture is in sharp contrast with the one shown in
Figure 42a where the dense breast tissue background is present. However,
as recognized by Yelleswarapu et al. (2008), the main drawback with this
filtering technique is that the spatial filter has to be modified when the mam-
mogram is changed because the spatial frequency spectrum at the Fourier
plane will be different. Thus, it cannot be real-time processing as the filter
is not all-optical and continually controllable. To remedy this, nonlinear
filtering techniques for self-adaptative and real-time computing such as those
developed by several workers (Babkina & Voloshinov, 2001; Kato &
Goodman, 1975) can be employed.
Figure 42 Processing of mammograms using spatial filtering (Figure 41): (a) original
mammogram. (b) With spatial filtering with a circular filter of diameter adapted to
the mammogram. Yelleswarapu et al. (2008).
Based on this optical setup, Yelleswarapu et al. (2008) were able to show
the correspondence between spatial frequencyintensitypolarization and
each spatial frequency band is encoded with a unique polarization. When
the analyzer is at right angles to the input beam polarization, the low-
frequency components that experience no polarization rotation are blocked
by the analyzer. On the other hand, high spatial frequency components
corresponding to edges of the object experience polarization rotation due
to their low intensity.
Yelleswarapu et al. (2008) provided an example to illustrate the above
discussion. As an illustrative example, Figure 44a shows the original mam-
mogram where microcalcification areas are not visible. Figure 44b is the
processed image clearly displaying these microcalcification areas (inside
the white rectangle) when the low spatial frequencies are filtered out.
Figure 44 Results obtained with the optical setup shown in Figure 39 when the target
image is a mammography image. Yelleswarapu et al. (2008).
174 Ayman Alfalou and Christian Brosseau
Wo x, y, t ao x, yexpio x, y expit,
(31)
WR x, y, t aR x, yexpiR x, y expit;
where a and are the amplitude and the phase of the wave, respectively.
The intensity on the CCD sensor reads
I x, y, t Wo x, y, t + WR x, y, t 2
(32)
Ii x, y + C x, yconso x, y R x, y ;
where Ii(x, y) is the mean intensity, c(x, y) defines the contrast, and o x, y
R x, y is the net phase. Since the second part of the light passes through
a half-wave and a quarter-wave, one can adjust the fast and slow axes at
angle
I x, y, t Ii x, y + C x, yconso x, y R x, y + t : (33)
The central aspect of the PSI method is to capture multiple interfero-
grams I() that each indicate the optical phase relationships ( 0, /2, )
between light returned from all sampled points on the illuminated surface
and a controlled reference beam of light that is collinear to the object beam.
Recent Advances in Optical Image Processing 175
From the set of these I()s, holograms are computed that contain information
defining the shape of the object. Knowledge of these intensity patterns allows
us to determine the actual phase of the wave. The four interference patterns
corresponding to phase shifts of 0, /2, and for object reconstruction
are calculated from
I x, y; 0 Ii x, y + C x, yconso x, y R x, y + 0
h i
I x, y; Ii x, y + C x, ycons o x, y R x, y + (34)
2 2
I x, y; Ii x, y + C x, yconso x, y R x, y + :
Using the PSI algorithm (Yamaguchi & Zhang, 1997), the wavefront at
the plane of the camera can be calculated from the recorded interference
patterns as
1 i hn o n oi
Wo x, y I x, y; 0 I x, y; + i I x, y; I x, y; ,
4jWR j 2 2
(35)
where WR is the amplitude of the reference wave (for a plane normally inci-
dent reference wave WR 1). Equation (35) is called a PSI hologram.
Figure 46 Standard 4f configuration for the study of the edge enhancement using
transient Fourier hologram. M, mirror, BS, beam splitter; L, lens. Kothapalli et al. (2004).
Figure 48 Experimental setup suggested by Kothapalli et al. (2005) for medical image
processing; L1, L2, and L3 are convergent lenses, NBF is a narrow band filter to block
568 nm at the CCD plane, CL1 and CL2 are collimation lenses, and NDF1 and NDF2
are neutral density filters to control the intensity of the beams.
Figure 50 Example showing the edge enhancement of a transparent object: (a) Bright-
field image and (b) Fourier phase contrast image of onion cells. Yelleswarapu et al.
(2008).
where Fb and Rb denote the binary images of the gray interesting region of Fg
and Rg, T is a threshold, Si21, 2, 3 denotes the addition operation for i 1, the
subtraction operation for i 2, and the union operation for i 3. We will
not dwell on the technical aspects of the derivation of Equation (36), but pass
on at once to its intuitive understanding. In practice, Equation (36) can real-
ize the union, the addition, the subtraction, the dilatation, the erosion, the
opening, the closing, the hit-or-miss, the edge detection, and the thinning
(Huang et al., 1996). In what follows, we consider the specific example of the
dilatation of a gray image Fg with a gray structure element Rg. The dilation of
the gray image Fg by a structure element can be written as
Fg Rg O Fg , Rg ,Fb , Rb , T 0, S1 Fg Rb 0 + Fb Rg , (37)
where denotes the dilation operation, T 0 denotes a given threshold, i.e.,
a high-pass filter, and S1 is the subtraction operation. In Figure 51, we show an
example illustrating how Huang et al. (1996) implement dilation.
Let us focus on the optical setup displayed in Figure 51 which is com-
posed of an incoherent source and a first convergent lens L1. The output
light is incident on a liquid crystal TV (LCTV1) used to monitor in real time
the binary images Rb (or Fb). Next, a second liquid crystal TV (LCTV2) is
used to monitor the gray image Fg (or Rg) obtained with the camera CCD1.
182 Ayman Alfalou and Christian Brosseau
Figure 51 Optical setup proposed by Huang et al. (1996) to realize the single operation
gray image processing.
Behind the convergent lens L2, a CCD camera collects the convolution
superposition images Fg Rb and Fb Rg into a computer for thresholding.
Finally, a monitor is used to display the processing results. An illustration of
this method is shown in Figure 52 which implements dilation, i.e.,
Equation (37), thanks to the setup of Figure 51. The first step of this method
consists in encoding Fg and Rg and their binary versions Fb and Rb (line 2).
After thresholding, the results Fg , Rb T 0 and Fb , Rg T 0 are shown in
the third line of Figure 52. Adding the two images leads to the results shown
in Figure 52.
Figure 52 Dilation of a gray image using the single fundamental operation. Huang
et al. (1996).
184 Ayman Alfalou and Christian Brosseau
where W denotes the universal image. Yuan et al. (1996) performed a variety
of tests and demonstrated that the NOBIA achieves better performances for
edge detection than the binary image algebra. The architecture implementing
the optical neighborhood operation binary image processor (ONOBIP) for
executing the parallel algorithms of NOBIA is shown in Figure 53.
Figure 53 contains two parts. The block framed in black is the control
unit which includes a clock, a program counter, a test and feedback control-
ler, and an instruction memory. The block framed in red is the parallel
processing unit which includes an image memory device for storing multiple
images, three frame buffers, and a CPU for executing the basic operation
Fn{.} of the NOBIA (Figure 54). The three buffers X, R, and Y are, respec-
tively, used to store the three input images for the central processing unit.
The images stored in the image memory device can be, respectively, sent
to the three frame buffers. The CPU can be controlled by the controlling
unit to perform proper functions. The ONOBIP operates as follows:
Clock Program counter
Image input
Memory Frame buffer R Fn {.}
Image output
(a)
IP LCLV A M2 OP
M1 BS
S
L
G
P
CB
(b)
P1 ML P2 P
a d
b
(c)
X Y OP
Collimated
beam
f+z
D
n
Figure 54 (a) Optical setup suggested in Yuan et al. (1996) for optically implementing
the optical neighborhood operation binary image processor (ONOBIP). (b) The input
plane P is illuminated by a uniform incoherent light. A mask M is put close in front
of an imaging lens L with a focal length f. The distance between the plane P and the
mask M is D. The distance between the mask M and the lens L is very small and can
be neglected. The mask M is used to display the structuring element image R. Plane
P is the imaging plane in which a clear image of the input plane P1 can be obtained.
In the output plane P2, we can get the convolution result. (c) Optical circuit for
performing the intersection operation.
186 Ayman Alfalou and Christian Brosseau
(1) the corresponding images are sent to the frame buffers X, R, and Y from
the image memory device; (2) the basic operation Fn{.} is executed by
selecting the proper formats of the nonlinear function; (3) the result is out-
put, tested for program control, or sent back to the image memory device.
Then the system can be fed back to step (1). Based on Figure 53, Yuan et al.
(1996) suggested a scheme (Figure 54a) which realizes Fn{.} optically. This
setup is composed of three blocks: an optical convoluter (Figure 54b), an
optical spatial nonlinear device with two operation modes such as a liquid
crystal light valve, and an optical implementation of parallel intersection
(Figure 54c).
The diffuse screen G is put just before the input plane IP. The incoherent
uniform light (S) is used to illuminate it. The mask M of the structuring ele-
ment R is put close in front of the lens L. Plane IP, mask Mi, lens L, and
LCLV construct an incoherent optical convoluter. When the image shown
in the input plane IP passes through mask M and lens L, the convolution
resultant image V(X,R) is formed in the writing plane of LCLV.
A collimated HeNe laser beam (CB) is used as the reading light to read
the written-in image of LCLV, where P is a polarizer and A is an analyzer.
Selecting different nonlinear characteristics of LCLV to simulate the
corresponding formats of the nonlinear function, the corresponding binary
image is read out from LCLV. The output binary image passes through the
mask M2 coded by binary image Y and executes the intersection operation
with image Y. Then in the output plane OP, the operation result of Fn(X, R,
NPn, Y) can be obtained. Using different nonlinear characteristic curves of
LCLV to perform different type nonlinear filterings, the different formats of
the basic operation are implemented experimentally. Figure 55 gives the
experimental results obtained by Yuan et al. (1996).
Figure 55 Results obtained by Yuan et al. (1996) using the optical setup displayed in
Figure 50a.
Recent Advances in Optical Image Processing 187
Figure 55a is an input binary image X. Figure 55b shows the structuring
element R with 3 3 image points, and Figure 55c gives the image Y.
Figures 51d and 55e give the experimental results of F1{X, R, NP1, Y}
and F2{X, R, NP2, Y}. The most remarkable implication of these results
is the validation of the ONOBIP by optical means. The suggested ONOBIP
is of a high parallelism, simple structure, and potential fast processing ability
and can be used to efficiently execute the algorithm of NOBIA. But it
remains difficult to realize in practice. We postpone a discussion of the edge
detection using this method for a few sections.
images of the same area. Figure 1 illustrates the well-known fact that the
underwater medium can significantly degrade the quality of the image. That
is, small-angle forward scattering of light and nonuniform lightning condi-
tions can cause images to blur and result in intensity differences between the
images of the same area.
Light attenuation due to absorption and scattering phenomena has been
studied before (Duntley, 1963; Hou, Gray, Weidemann, Fournier, &
Forand, 2007; Jaffe, 1990; Leonard, 2012; Wells, 1969). Following earlier
studies (Duntley, 1963; Hou et al., 2007; Jaffe, 1990; Wells, 1969), Leonard
(2012) considers attenuation as a linear combination between absorption11
and scattering.12 The idea here is to consider that the light intensity trans-
mitted at a distance d in a sample can be written as
I d I0 expa + bd , (43)
where I0 denotes the intensity of the source, a is the absorption coefficient of
the underwater medium, and b is the total scattering coefficient. Equa-
tion (43) shows that I(d) depends on the absorption of the medium.
Figure 56 illustrates the estimation of the propagation distance of light in
clear water (Sogandares & Fry, 1997) as a function of its wavelength. In addi-
tion, I(d) depends also on scattering phenomena in the medium. The latter
depends on the particle size, light incidence, and the distance between the
particles and the camera. The perturbating noise that affects the overall
image is called sea snow. Leonard (2012) considers that this noise depends
on the forward scattering of light, i.e., blurring the image, and also on back-
scattering of light, i.e., the image is affected by a veil.
It is worth observing that this issue has been comprehensively reviewed
by Leonard (2012), Leonard, Alfalou, Alam, et al. (2012), and Dubreuil et al.
(2013).
11
The light absorption is related to the complex refraction index of the underwater medium and also to
the presence of pigmented particles.
12
The light scattering is due to the presence of transparent biological organisms and also due to
suspending particles.
Recent Advances in Optical Image Processing 189
0
17 m 12 m
49 m
100
The light propagation (m) 92 m
200
300
293 m
640 nm
400 371 m 600 nm
580 nm
530 nm
500 490 nm
470 nm
529 m 400 nm
600
Figure 56 Illustrating the propagation of light in clear water; i.e., the red (dark gray in
the print version) component of light is completely absorbed in water beyond 12 m.
Sogandares and Fry (1997).
13
As you go deeper under water, first red light is filtered and the scenery around you has an orange tint.
Orange light is filtered after going deeper and the scenery around you has a yellowish tint. Yellow light
is filtered going deeper than that at this point the scenery around you looks greenish-blue. Green light
is filtered after going deeper in which the scenery around you looks a dark pure blue. And then finally
blue light is filtered along with indigo and violet which makes everything pitch black. These color tints
can only be viewed by scenery at the same depth as yourself so if you look down deeper into the water,
then objects start becoming bluer even though you yourself may be at the orange depth.
Recent Advances in Optical Image Processing 191
Saturation &
Contrast Transform RGB
Input image intensity Output image
stretching RGB to HSI
stretching HSI
Figure 57 Synoptic diagram suggested in Iqbal et al. (2007) for increasing the quality of
underwater images.
Figure 59 Underwater image by using the method suggested in Iqbal et al. (2007):
(a) without processing and (b) with processing.
in the HSI14 color space. Then, they apply a saturation operation and realize
an intensity stretching on this HSI color model. These authors considered an
underwater scene containing a groundfish (Figure 59a) characterized by an
image of poor quality. Applying their method, they obtained the image
shown in Figure 59b which is clearer and of better quality than the original
one. In addition, one advantage of applying two stretching models is that it
helps to equalize the color contrast in the images and it also addresses the
problem of enlightening.
Alternate protocols for correcting underwater images have been
suggested. For example, Torres-Mendez and Dudek (2005) considered
the problem of color restoration using statistical inference. This is applied
to color recovery for underwater images, using an energy minimization for-
mulation. For aquatic robot tasks, the quality of the images is crucial and
14
HSV, hue, saturation, and value color space.
Recent Advances in Optical Image Processing 193
needed in real time. Their method enhances the color of the images by using
a Markov random field (MRF) to represent the relationship between color-
depleted and color images. The parameters of the MRF model are learned
from the training data, and then the most probable color assignment for each
pixel in the given color-depleted image is inferred by using belief propaga-
tion. This allows the system to adapt the color restoration algorithm to the
current environmental conditions and also to the task requirements. Exper-
imental results on a variety of underwater scenes demonstrate the feasibility
of the method. But this kind of method remains quite difficult to realize in
practice (Granier, Refregier, & Figue, 1996; Gurov & Sheynihovich, 2000).
Figure 60 Schematic diagram used in Leonard (2009) and Leonard, Alfalou, Alam, et al.
(2012) for taking into account the sun enlightening of the underwater scene, is the
angle between the normal N at the surface and the light source SL (SL, sun light), and
denotes the angle between the reflected beam R and the position of the sensor S (CCD
camera).
194 Ayman Alfalou and Christian Brosseau
reflection model (Phong, 1975) used to correct the scene considered. This
approach is an empirical model of the local illumination of points on a sur-
face. In 3D computer graphics, it is sometimes ambiguously referred to as
Phong shading, in particular if the model is used in combination with the
interpolation method of the same name and in the context of pixel shaders
or other places where a lighting calculation can be referred to as shading.
Basically, this model considers that the light reflected by an object IT with
modification before the CCD camera, i.e., ICCD. Following Leonard (2012)
and Phong (1975), the reflected intensity by the target object can be written
as the sum of three contributions as shown in Figure 60
IT IA + ID + IS , (46)
where IA represents the light intensity at the scene. The second term, ID,
characterizes the diffuse reflection, that is, the reflection of light from a sur-
face such that an incident ray is reflected at many angles rather than at just
one angle as in the case of specular reflection. Lambertian reflection is often
used as a model for diffuse reflection. The reflection is calculated by taking
the dot product of the surface normal, N, and the light direction pointing
from the surface to the light source (Figure 60),
ID cos I0 , (47)
where is the angle between the two directions. The intensity will be the
highest if the surface is perpendicular to the direction of the light, and the
lowest if the surface runs parallel with the direction of the light. The third
term, IS, is the specular reflection. Its intensity depends on the cosine of the
angle (see Figure 60) at power which characterizes the brightness of the
target objects material,
IS fcos g I0 : (48)
Once IT is estimated using Equation (46), Leonard (2012) suggests to
make use of the Beer-Lambert law to quantify the received intensity by
the CCD camera ICCD. This intensity is written as
@ 2 I1 x, y
I 2a2 + b2 2a2 I1 x, y a2 , (51)
@x2
A1
DF A0
A1
y L1 L2
L0 LCD M CCD
LED
z
P0
f0 f1 f1 f2 f2 Dz
S0 S
Figure 62 Optical setup used by Flores et al. (2011) for implementing their method of
enhancement of partial second-order derivatives of an image. The proposed system uti-
lizes a liquid crystal display, onto which images to be derived are displayed. The method
is based on the capacity of these devices to generate simultaneously a contrast-reverted
replica of the image displayed on it. A positive replica of original image is obtained
when the LCD is between crossed polarizers, while the negative one is obtained with
the LCD in between parallel polarizers. Since the LCD is a diffraction element, both polar-
izer arrangements may be simultaneously implemented, for example, in a 4f optical pro-
cessor using polarization filtering of the different diffraction orders. When three images
(two replicas of original image displaced in the opposite direction and the other one
with contrast reverted and centered at the origin) are imagined across a slightly
defocused plane, one obtains an image with enhanced second derivatives in a given
direction.
Figure 63 Illustration of results obtained by Flores et al. (2011) for suppressing and/or
enhancing specific zones of the target images. The upper left picture shows the original
Picasso's dove; the other pictures show optical processed images.
8
>
> B KL + a
>
>
< K 0 arg min jjKL Bjj2 + jjKjj2 b
k 2 2
2 2 (52)
>
>
>
>
: L 0 arg min L jjKL Bjj22 + log pL c
2
where B is the blurred image, K is the blur kernel, L is the latent image,
denotes the additive noise in the blurred image, and and are, respectively,
the weights of the likelihood (fidelity) term and the prior (regularization)
term related to the additive noise variance and the function log p(L) is
defined in Lee et al. (2012).
Figure 65 (Lee et al., 2012) clearly shows that one needs two versions of a
given image at the input of the system (top left) in order to restore it: (1) the
blurred image and (2) the noisy image. These two images are used to
Figure 64 Synoptic diagram describing the algorithm suggested by par Lee et al.
(2012).
Proposed
method
Input images
Output image
Figure 65 Illustrating the deblurring results using the algorithm shown in Figure 60. Lee
et al. (2012).
200 Ayman Alfalou and Christian Brosseau
construct the edge map (green (light gray in the print version) blocks in
Figure 64). Next, this map is used to determine the kernel estimation (purple
(light gray in the print version) blocks in Figure 64). Finally, the two input
images, i.e., the edge map and the kernel estimation, are used to get the
restored image (blue (light gray in the print version) block). Based on
the technique described schematically in Figure 64 and the two (input)
images displayed in Figure 65, the restored (output) image is presented in
Figure 65.
These simulation results demonstrate the good performances of this
approach in terms of image quality at the output. In order to understand
quantitatively the above features, it is useful to remark that the peak SNR
of restored images has an average rise of 11%, compared to other algorithms.
Yet, no optical setup permitting an all-optical implementation has been
suggested.
15
A CCD is most simply described as an electronic photon detector. The typical CCD camera used for
object recognition applications consists of a 2D array of photon detectors in a layer of semiconducting
material, in this case silicon, which is placed at the focal plane of a camera in order to collect an image.
Each individual detector in the array is referred to as a pixel and can vary in size from a few to several
tens of microns. The smaller the size of the pixels, the higher the resolution of the image produced for
a given camera. Each individual pixel is capable of collecting photons and storing the produced elec-
trons, which can be read out from the CCD array to a computer to produce a digital image of the
varying intensities of light detected by the CCD. In this way, digital storage, reproduction, and
processing are possible, with the digital representation of the image field containing information that
is extremely accurate and easily reproduced. CCD imagers utilize the photoelectric effect. When pho-
tons hit the detector surface of the CCD sensor, electrons are liberated and stored in the detector ele-
ments, or pixels. Each pixel in the CCD is thus an electron well and accumulates electrons in
proportion to the number of photons received. The electrons in the atoms that make up the silicon
crystal exist in discreet energy bands. The lower energy band is referred to as the valence band, while
the higher energy band is called the conduction band. The electrons, typically occupying the valence
band, can be raised into the higher energy conduction band (excited) either by heat or by absorbing a
photon. Electrons that have been excited into the conductive band are able to move freely in the
silicon crystal lattice. When this occurs, a hole is left in the valence band, which then behaves as a
positively charged carrier. The CCD introduces an electric field in order to keep the electrons from
dropping back into the valance band and filling the hole, a process known as recombination. Recom-
bination will occur if there is no electric field present. Unfortunately, electrons can be excited into the
conduction band by the heat produced by the system (thermal noise) and these electrons cannot be
distinguished from those generated by photons (the signal).
Recent Advances in Optical Image Processing 201
readout of the device. This additive type of noise, described in the number of
electrons per pixel, is independent of exposure time. This noise is produced
during the conversion from an analog to a digital number. In addition, the
electronics of the system will introduce spurious electrons during the readout
process, adding to the fluctuations in output. Read noise can be isolated and
removed from a CCD image via the subtraction of a bias frame. A bias frame
is an image of zero exposure time where the CCD is read out without having
been exposed to light. In this manner, thermal noise produced by the heat
generated by the devices electronics and contributions of light to the expo-
sure are at a minimum, thereby isolating the effect of read noise. Since the
read noise varies from pixel to pixel and readout to readout, a number of such
bias images (nine or more) are recorded and then averaged together.
This is accomplished using the technique of median combining to pro-
duce a master bias frame. As described previously, in the photoelectric effect,
electrons generated by the heat produced by the system (thermal noise) can-
not be distinguished from electrons generated by photons (signal). There-
fore, there are always some number of electrons stored in the pixels that
are not the result of photons hitting the detector. The electrons generated
by thermal noise exist even when light does not hit the detector surface,
and are referred to as dark current. Dark current is a multiplicative form
of noise, the level of which is proportional to the length of the exposure.
Given a long enough exposure time, the detector could be fully saturated
with electrons due in large part to thermal noise. Overall this noise will
degrade the quality of the image. Other kind of noise is due to the moire
effect. This section is devoted to the strategy of optical correction of the
moire effect by making use of the 4f optical setup. As recalled in
Section 2.3.2, the moire effect manifests itself as an additive noise character-
ized by a periodic structure which is superimposed on the image. Conse-
quently, this noise can be written as
nmoir t cos wt : (53)
Fourier transforming Equation (43) leads to a sum of two Dirac delta
functions at frequencies w and w
1 1
Nmoir w + w : (54)
2 2
According to Equation (54), the two Dirac delta functions need to be
suppressed to deal with the moire issue. For that purpose, the 4f optical setup
composed of two convergent lenses L1 and L2 (Figure 66). The first one is
202 Ayman Alfalou and Christian Brosseau
Input image
L1
Fourier plane :
Filter
L2
Output plane
Figure 66 Optical setup used to suppress the moir effect.
used to realize the FT. Next, the spectrum of the noisy image is multiplied
with the appropriate filter for suppressing the moire peaks. Then, a second
convergent lens is used to get the denoised image.
A single example will illustrate the benefit of using this optical setup.
Consider the image in Figure 67a which is pertubated by a moire effect.
Fourier transforming this image leads to the spectrum shown in
Figure 67b. This spectrum displays the characteristic moire peaks (circled
in red). Choosing an appropriate filter to eliminate these peaks and realizing
the inverse FT, one gets the denoised image shown in Figure 67c.
For a more comprehensive description of the moire effect, the methods
employed to suppress it, and the applications which exploit it we refer to
Khalil and Fazio (1973), Yokozeki and Ohnishi (1975), Andonian (1982),
Lim et al. (1989), Bara et al. (1991), Simova and Stoev (1992), Kato et al.
(1997), Aggoun (2006), Buytaert and Dirckx (2007), Kim et al. (2009),
Rasouli et al. (2011), and Leonard, Alfalou, Alam, et al. (2012).
(a) (b)
(c)
time over conventional electronic algorithms (i.e., real time). Huang et al.
(1996) presented and validated a formal procedure for edge detection which
can be implemented optically by making use of the setup displayed in
Figure 51. This approach makes use of the single fundamental operation
O{.} defined earlier by Equation (36). The edge detection of gray image
ED(Fg) can be quantified by the following governing equation:
ED Fg O Fg , Rg1 ,Fb ,Rb1 , T 0,S1 O Fg , Rg2 , Fb ,Rb2 , T 0,S1
Fg Rg1 Fg Rg2 ;
(55)
where O{.} denotes the single fundamental operation, Fg and Fb are, respec-
tively, the target gray image and its binary version, (Rg1, Rg2) and (Rb1, Rb2)
are, respectively, the two structure elements considered and their binary
16
The raw mine images contained in this publication were provided by the GESMA (Groupe dEtudes
Sous Marines de lAtlantique) within the TOPVISION project coordinated by Thales Underwater
Systems SAS. This project is related to Techno-Vision Program launched by the French Ministry
of Defense. More details can be found on the Web site: http://topvision.gesma.fr.
204 Ayman Alfalou and Christian Brosseau
Figure 68 Example, provided by Huang et al. (1996), of the extraction of the whole pro-
file of a given gray image.
Figure 68b. The result shown in Figure 68c indicates that this approach
allows us to optically detect the edge of an image with a good quality.
3.3.5.2 Edge Detection and Image Reconstruction from Its Spectral Phase
Several authors (Alfalou et al., 2010; Alfalou, Mansour, et al., 2011) empha-
sized the importance of the spectral phase of an image to address the issue of
optical compression of images. Oppenheim and Lim (1981) demonstrated
that the spectral phase of a target image contains the information required
for edge detection and its reconstruction. Quatieri and Oppenheim
(1981) proposed an iterative algorithm enabling reconstruction of a target
image from the knowledge of the spectrums phase.17 In addition, according
the root-mean-square (RMS) duration criterion (Papoulis, 1962), only a
part of the image spectrum is required.
Edge detection
In this section, a simple and efficient method of edge detection is presented.
This method relies on earlier work by Oppenheim and Lim (1981) and the 4f
optical setup. The synoptic diagram is illustrated in Figure 69. This approach
is based on Fourier transforming the target image with a single convergent
lens. Solely the phase information in the Fourier plane is considered. Finally,
the FT1 allows reconstructing the phase or edge image (Figure 69).
This protocol is used to deal with pure phase filter (POF) correlation
(Horner & Gianino, 1984).
Image reconstruction
We now turn to describe the principle of iterative reconstruction using only
the real spectral phase of the image such as reported by Quatieri and
17
The same is also true from the knowledge of its amplitude.
206 Ayman Alfalou and Christian Brosseau
(a) (b)
FT Reconstructed
image (2N,2N) pixels:
FT
A exp(if)
Image spectrum
A exp(if)
Phase only
(f) A
Figure 70 Iterative algorithm principle: (a) only phase section and (b) iterative
reconstruction image.
spectrum of the target image with a filter which has the effect of suppressing
its amplitude. Before doing that, the size of the target image (N N pixels)
is increased by zero-padding the FT; i.e., the size becomes 2N 2N pixels.
The second step uses an iterative algorithm as is illustrated in Figure 70b.
First, the image spectrum is reconstructed by taking its amplitude A set to 1.
Then, a set of FT and FT1 is realized until the output image converges
to the original image. To do this, while conserving the phase obtained at
the beginning, we change for each cycle the amplitude of the image spec-
trum through the use of the image reconstructed on N N pixels. To
show the good performances of this reconstruction algorithm, we calculate
the MSE defined as
1 XN X
N
MSE 2
jI0 i, j In i, jj2 (61)
N
for each iteration. This criterion evaluates the difference between the ampli-
tude of the target image I0 and the reconstructed image with the considered
algorithm In, where n denotes the number of iterations and N is the number
of pixels of the target image I0.
In Figure 71, simulations were conducted with the Lena image I0. By
comparing the phase image obtained with a single iteration (top left image
in the panel of Figure 71) with the target image I0, we find that the former
has a very low quality as evidenced by the high value of the MSE
211. If
we apply our algorithm with 10 iterations, we observe a significant increase
of the image quality. This can be seen by visual inspection and a lower value
of the MSE
75. After 500 iterations, the reconstructed image shown at the
0.25
0.2
0.15
MSE
0.05
0
0 50 100 150 200 250 300 350 400 450 500
Iteration number
Ten iterations Five hundred iterations
Figure 71 Influence of the iteration number on the quality of the reconstructed image.
208 Ayman Alfalou and Christian Brosseau
Partially
depolarizing
target
Backscattering
Light source
Cross-polarizer
Camera
Linear polarization
Figure 72 Illustrating the backscattering and depolarization of light by an underwater
target. Leonard, Alfalou, and Brosseau (2013).
Recent Advances in Optical Image Processing 211
Fornetti, et al., 2007; Boffety, Galland, & Allais, 2012; Cariou, Le Jeune,
Lotrian, & Guern, 1990; Clemenceau, Dogariu, & Stryewski, 2000;
Gorman, 2011; Jaffe, 1990; Kattawar & Rakovic, 1999; Kim & Moscoso,
2002; Kohlgraf-Owens & Dogariu, 2009; Kouzoubov, Brennan, &
Thomas, 1998; Miller & Caprari, 1999; Miller & Dereniak, 2012; Mullen,
Cochenour, Rabinovich, Mahon, & Muth, 2009; Piederrie`re et al., 2005;
Treibitz & Schechner, 2009; Tyo, Goldstein, Chenault, & Shaw, 2006;
Voss & Fry, 1984; Walker, Chang, & Hopcraft, 2000; You et al., 2011 for
reviews and many references).
Combining polarization filtering and correlation techniques is justified
because it is simple to implement, and it enables simultaneous detection
and identification of target objects. Schechner and Karpel (2005) suggested
an approach to increase the visibility in water by exploiting linearly polarized
images. However, this requires an a priori knowledge on the properties of the
embedding medium, which can be difficult to obtain, and of the nature of
the target object. Hence, the degree of depolarization of the wave by the
target does not allow us to select the optimal polarization state for detection.
In order to address these shortcomings, Treibitz and Schechner (2009)
introduced a technique which is based on using linear parallel and crossed
linearly polarized images for evaluating backscattering. In the following,
we will restrict ourselves to the method suggested in Schechner and
Karpel (2005).
POL0 R P
Block
Laser A q
Target
L1 L2 L3
Block
B
CCD
L4
Figure 73 Optical setup for polarization measurement. Details have been previously
discussed in Leonard, Alfalou, and Brosseau (2013) and Dubreuil et al. (2013).
Figure 74 Speckle image obtained after scattering of light by a flat surface: (a) image
obtained without R; (b) image with R showing that the speckle is significantly
attenuated.
Several things are worth noting about Equation (64). It is evident that the
DOP ranges between 0, i.e., unpolarized light, and 1, i.e., totally polarized
p
S2 + S2
light. The linear degree of polarization is DOPL S2 0 3 , and the circular
degree of polarization is DOPC SS30 . We wish to show several examples that
illustrate the benefit of using polarimetric techniques to improve the image
quality. Another purpose is to show that PO coupled to OC techniques
provides a powerful yet simple method to detect and recognize objects.
We begin by considering target images in air. Details have been previously
discussed in Dubreuil, Alfalou, and Brosseau (2012), Leonard, Alfalou, and
Brosseau (2013), Dubreuil et al. (2013), and Leonard, Alfalou, Zallat,
Lallement, and Brosseau (2013).
Recent Advances in Optical Image Processing 215
(a) POL0 R P
Block A
L1 L2 L3
POL1
0
Laser
q
Target
90
QWP1
QWP2
POL2
Block B
CCD
L4
(b)
Laser
Rotating diffuser
Pinhole
CCD camera
Optics components
Lens (before)
Optics components
(after)
Target to be analyzed
Figure 75 (a) Polarization imaging optical setup for measurement in air; (b) a picture of
the optical setup. Leonard, Alfalou, and Brosseau (2013) and Dubreuil et al. (2013).
216 Ayman Alfalou and Christian Brosseau
01
Sexp0
B Sexp1 C
If the output Stokes vector Sexp B C
@ Sexp2 A is measured, then we can
Sexp3
determine the Stokes vector after reflection/scattering by the target in the
0 1
Starg0
B Starg1 C
scene to be analyzed Starg B C
@ Starg2 A. The two vectors are related to each
Starg3
other by a matrix multiplication
Sexp MStarg
(65)
M M Pol M 1=4 M 2=4 :
For example,
Sexp0 m00 Starg0 + m01 Starg1 + m02 Starg 2 + m03 Starg 3 (67)
can be measured by the CCD camera. Each Stokes image corresponds to a
specific orientation of the quarter-wave plates (Figure 76) according to some
specific values of 1 and 2.
QWP1 QWP2
POL1 Pol2
0 90
q2 q1
Target
with a high DOP in sharp contrast with the plastic piece which is charac-
terized by a small DOP. Two kinds of materials with similar reflectance
but otherwise different depolarization properties will appear identical in
standard (intensity) imaging, but PO offers a mechanism of image contrast
as illustrated in Figure 77b.
In addition, PO makes it possible to get away from the nonuniformity
of the lighting of the scene, i.e., compare Figure 78a and b. It also renders
easier the application of correlation techniques. As a quantitative check,
we show the correlation planes, respectively, with or without taking into
account polarization (respectively, Figure 78c and d) along with the values
of the PCE which is the basic descriptor typically used to characterize the
correlation.
On the experimental side, we conclude that correlation associated with
polarization imaging is well adapted when the lighting condition is weak and
nonuniform, e.g., Figure 78c. PO can be implemented in a complete form
using the Mueller matrix formalism to characterize each pixel of an image.
The correlation peak is more or less intense, depending on the degree of
(a) (b)
(c) (d)
(a)
POL1
POL2
Target
(b) POL2
POL1
Target
Figure 79 Linear polarimeter with either (a) crossed polarizers or (b) parallel polarizers.
220 Ayman Alfalou and Christian Brosseau
Schechner, 2009; Tyo et al., 2006; Voss & Fry, 1984; Walker et al., 2000;
You et al., 2011), the DOPL can be written as
Ik I?
DOPL : (70)
Ik + I?
Figure 80a and b shows, respectively, I? and Ik for a metal coin in air.
Then, by making use of Equation (70), we get the DOPL image shown
in Figure 80c. This permits enhancement of the external edges of the coin.
These DOPL measurements are consistent with the low depolarizing feature
of metal.
Figure 80 Experimental results obtained with a metal coin: (a) intensity I? (crossed
polarizers), (b) intensity Ik (parallel polarizers), and (c) corresponding DOPL.
(a) QWP1
POL1 POL2
Target QWp2
(b) QWP1
QWp2
POL2
POL1
Target
Figure 81 Circular polarimeter used for obtaining: (a) I45 and (b) I + 45
Recent Advances in Optical Image Processing 221
Figure 82 Experimental results with a metal coin: (a) intensity I45 (parallel polarizers
and quarter-wave plates placed with 90), (b) intensity I + 45 (parallel polarizers and par-
allel quarter-wave plates), and (c) corresponding DOPc.
222 Ayman Alfalou and Christian Brosseau
Figure 83 Images of a metal coin placed behind a scattering medium (Scotch piece):
(a) intensity image and (b) DOP.
0.0018
DOP
0.0016
I0
0.0014
0.0012
0.001
PCE
0.0008
0.0006
0.0004
0.0002
0
0 2 4 6 8 10 12
Scotch pieces
Figure 84 A comparison of correlation tests of the DOP and intensity image as a func-
tion of the number of Scotch pieces added to the target.
Figure 85 (a) Scene with different types of disk-shaped materials ((c) is a metal coin,
(d) is piece of cardboard, (e) is a piece of absorbing paper, and (f ) is a piece of white
paper) and (b) corresponding DOP images.
Figure 86 Image of an object on a paper sheet: (a) in clear water and (b) in water
with milk.
conducted in a glass tank filled either with a scattering medium (milk dis-
persed in water) or with seawater.
imaging because it filters the major part of the backscatter and only a
small part of the target signal.
For polarization-maintaining objects, performing only one measurement
is insufficient to improve target detection because an important part of the
signal is lost. However, whatever the target is, performing two measure-
ments with two orthogonal polarization states improves recovery of the tar-
get signal by suppressing the backscatter that veils the object. The contours
of the target are thus well defined which enhances correlation performance.
Experiments performed in different scattering media and different targets
showed that there is always an improvement in target detection when target
estimation is considered. In this study, no difference in decision performance
was found between linear and circular polarized lights.
The panels of Figure 87 show images of a plastic plug for skimmed milk
dispersed in water (20 mL milk to 10 L water) giving s 0.08 cm1.
Typically, the distance that light travels before reaching the target is
15 cm, corresponding to an optical thickness of 1.2. Images taken with
I and CRL modes, linear estimation, are compared to an image taken in pure
water. Differences between these images merit discussion.
Figure 87 Images of the plastic plug. Experimentally, 20 mL of skimmed milk has been
added to 10 L of tap water. (a) Intensity image. The red (gray in the print version) square
indicates the area over which the degree of polarization is calculated, (b) cross-linear
image, (c) signal estimation image (linear), and (d) pure water image. Dubreuil et al.
(2013).
Recent Advances in Optical Image Processing 225
(a) (b)
PCE = 6.4
PCE = 2.2
1
1 0.8
0.6
0.5 0.4
0.2
0 0
Figure 88 Correlation planes corresponding to the images shown in Figure 83. Dubreuil
et al. (2013).
226 Ayman Alfalou and Christian Brosseau
Figure 89 (a) PCE versus optical thickness 0 for a plastic plug and the different imaging
modes shown in the inset. (b) As in (a) for painted metal. (c) As in (a) for rusted metal.
The scattering medium is composed of skimmed milk diluted in tap water. Dubreuil et al.
(2013).
looks better in the CRL image (Figure 88b) and even better for the target
estimation (Figure 88c). The PCE parameter provides valuable information
about images and can be useful for automatic target recognition.
In order to validate the advantage supplied by polarization imaging, a
series of experiments were conducted with varying milk concentration
and for different targets (plastic plug, painted and rusted metal). Figure 89
illustrates that as the optical thickness increases in skimmed milk, the
PCE tends to decrease for I, COL and CRL, COC and CRC imaging
modes, and target estimations for linear (ESTL) and circular (ESTC) polar-
izations. The panels in Figure 90 show the corresponding results for exper-
iments in semi-skimmed milk. It should be noted that the PCE values are
normalized by the highest value for one set of experiment (corresponding
to a specific target). Hence, these values can safely be compared for different
polarizations states.
We find that all curves have the same behavior. For depolarizing media
(plastic plug and painted metal), the PCE values (and thus the detection
Recent Advances in Optical Image Processing 227
Figure 90 As in Figure 5 for semi-skimmed milk diluted in tap water. Dubreuil et al.
(2013).
efficiency) are lower for COL and CRC than for I. Imaging with such polar-
ization modes does not allow for backscatter suppression. Moreover, a part
of the target signal is filtered with these modes. The PCE increases for COC
and CRL because of backscatter suppression. Larger PCEs are obtained for
ESTL and ESTC for which full backscatter is suppressed. For a polarization-
maintaining medium (rusted metal), one should note that imaging with
COC and CRL is not beneficial because a large part of the target signal is
filtered as well as the backscatter. However, target estimation leads to better
detection efficiency than for I, CRC, and COL. We note also that no sig-
nificant differences can be seen between imaging with circularly and linearly
polarized light. We further find that higher optical thickness can be achieved
in the case of semi-skimmed milk. This agrees with the fact that in Mie scat-
tering regime, the backscatter is smaller than in Rayleigh scattering regime,
and thus it requires more milk concentration to obtain a significant veil.
228 Ayman Alfalou and Christian Brosseau
Figure 91 (a) PCE as a function of target depth for painted metal in ocean water.
(b) Same as in (a) for lake water. Dubreuil et al. (2013).
Recent Advances in Optical Image Processing 229
lake water, respectively. The reference images are the painted metal in pure
water taken at different depths.
Likewise, our analysis suggests that adapted polarization imaging (CRL
and COC) improves target detection compared to intensity imaging. Target
estimation is the best way to achieve good detection when the turbidity of
the embedding medium is large. As the turbidity of water is fixed and given
the dimensions of our glass tank, smaller optical thicknesses can be achieved
in these experiments as compared to our previous experiments with diluted
milk. The turbidity of the lake water was found to be much smaller than that
of seawater.
Figure 92 Underwater imaging system used by Leonard, Alfalou, and Brosseau (2013).
230 Ayman Alfalou and Christian Brosseau
1m80
1m60
1m40
1m20
1m00
Figure 93 From the right, the first column indicates the object-to-camera distance; the
second column is the unpolarized image of the spherical object, the third column cor-
responds to the same object imaged with LPP, and the fourth column corresponds to
LCP. 1 L of skimmed milk was dispersed in the tank containing 6000 L of water.
Recent Advances in Optical Image Processing 231
of water), and varying the distance (number in the left column) of the camera
to the target (painted metal sphere) (Leonard, Alfalou, & Brosseau, 2013). In
the second column of Figure 93, we present the unpolarized image, whereas
the third and fourth columns of Figure 93 show, respectively, the
corresponding contrasts for the linearly parallel polarized (LPP) and linearly
crossed polarized (LCP) images. Polarization is shown to a useful tool for sig-
nificantly increasing the visibility of the object which is practically invisible in
the unpolarized image. There are many points to consider. First of all and as
shown in Figure 93, decreasing the object-to-camera distance increases the
recognition rate of the object because the visibility is increased. Second,
Figure 93 also indicates that the use of LCP allows better identification of per-
formances than LPP. It is also worth observing that Leonard, Alfalou, and
Brosseau (2013) presented analogous results dealing with color images for
an object-to-camera distance set to 1m80 (Figure 94) and milk addition.
As a final remark, we would like to point out that these experimental
results underlined the difficulties of dealing with real-time measurement
on moving targets or when the turbidity of the water is rapidly varying. Cur-
rent developments involving optical procedures using liquid crystals permit-
ting the obtention of fast polarization measurements and microlenses
permitting imaging the scene on different areas of the sensor for different
states of polarization are underway.
Figure 94 Color image intensity (left) and DOP (right) of the target in water with milk
addition. The object-to-camera distance is set to 1m80.
232 Ayman Alfalou and Christian Brosseau
optical setup shown in Figure 51. As discussed earlier, Huang et al. (1996)
described and implemented a method for removing a hole using an optical
setup based on the use of the single fundamental operation. They first
showed that the noise in a gray image can be suppressed (or attenuated)
by applying opening and/or closing operations to the image based on the
single fundamental operation O{.} defined by Equation (36). The reader
is referred to Huang et al. (1996) for the full details of the image processing.
The following example is obtained by making use of the optical setup
displayed in Figure 51 to suppress (or at least reduce) optically noise in a gray
image. The aim of Huang et al. (1996) was to remove the hole appearing in
the gray image shown in Figure 95a. For that specific purpose, these authors
used a choice of structure element with 2 2 pixel displayed in Figure 95b.
The result of the noise removing from Figure 95a after closing is displayed in
Figure 95c. The experimental results show that the single operation mor-
phological processor has good performances. In addition, it can be optically
implemented.
3.5.2 Fusion
Fu and Zhang (2012) described and implemented numerically an efficient
method of fringe pattern denoising using averaging based on nonlocal
self-similarity. Related work includes the multifocus image fusion and den-
oising scheme based on homogeneity similarity reported by Li, Chai, Yin,
and Liu (2012), and a fusion method based on the shearlet transform
described in Miao, Shi, Xu, Yang, and Shi (2011).
Image
Compression Encryption Increase size
Image
Encryption Compression Lose of information
Good compromise
Image
Compression & encryption simultaneousely Compression/encryption
perceptible loss of quality is considered a valid trade-off for the reduced data.
More generally, some forms of lossy compression can be thought of as an
application of transform codingin the case of multimedia data, perceptual
coding: it transforms the raw data to a domain that more accurately reflects
the information content. Three classes of lossy compression have been
developed: those by prediction, e.g., ADPCM,19 those by transform, e.g.,
JPEG,20 JPEG 2000,21 MPEG,22 and those based on fractal properties.
Decompression is the inverse operation of compression.
Encryption
In cryptography, encryption is the process of encoding messages (or infor-
mation) in such a way that only authorized parties can read it. Encryption
does not prevent hacking but it reduces the likelihood that the hacker will
be able to read the data that is encrypted. In an encryption scheme, the mes-
sage or information (referred to as plaintext) is encrypted using an encryp-
tion algorithm, turning it into an unreadable ciphertext. This is usually done
with the use of an encryption key, which specifies how the message is to be
encoded. Any adversary that can see the ciphertext should not be able to
determine anything about the original message. An authorized party, how-
ever, is able to decode the ciphertext using a decryption algorithm (that usu-
ally requires a secret decryption key) which adversaries do not have access to.
For technical reasons, an encryption scheme usually needs a key-generation
algorithm to randomly produce keys. As mentioned briefly earlier, in sym-
metric key schemes, the encryption and decryption keys are the same. Thus,
communicating parties must agree on a secret key before they wish to com-
municate. In public-key encryption schemes, the encryption key is publi-
shed for anyone to use and encrypt messages.
However, only the receiving party has access to the decryption key and is
capable of reading the encrypted messages. Public-key encryption is a rela-
tively recent invention: historically, all encryption schemes have been sym-
metric key (also called private key) schemes. Encryption, by itself, can
19
ADPCM (Adaptive Differential Pulse-Code Modulation) is a variant of differential pulse-code mod-
ulation that varies the size of the quantization step, to allow further reduction of the required band-
width for a given SNR.
20
JPEG (Joint Photographic Experts Group) is a group of experts that develops and maintains standards
for a suite of compression algorithms for computer image files.
21
JPEG 2000 is a standard which was meant as an update of the widespread JPEG image standard. The
lossy image compression mode of JPEG 2000 is based on the so-called discrete wavelet transformation
which is supposed to be more efficient than the JPEG algorithm based on the cosine transformation.
22
MPEG (Moving Picture Experts Group) is a working group with the mission to develop standards for
coded representation of digital audio and video and related data.
238 Ayman Alfalou and Christian Brosseau
protect the confidentiality of messages, but other techniques are still needed
to protect the integrity and authenticity of a message; for example, verifica-
tion of a message authentication code or a digital signature. In cryptography,
Kerckhoffs principle was stated by Auguste Kerckhoffs in the nineteenth
century: a cryptosystem should be secure even if everything about the sys-
tem, except the key, is public knowledge. Kerckhoffs principle23 was
reformulated (or perhaps independently formulated) by Claude Shannon as
the enemy knows the system. In contrast to security through obscurity,
it is widely embraced by cryptographers. Stated simply, the security of a
cryptosystem should depend solely on the secrecy of the key and the private
randomizer. Another way of putting it is that a method of secretly coding
and transmitting information should be secure even if everyone knows
how it works. Of course, despite the attackers familiarity with the system
in question, the attacker lacks knowledge as to which of all possible instances
is being presently observed.
The DES24 (Shamir, 1985) is a previously predominant symmetric-key
algorithm for the encryption of electronic data. DES is now considered to
be insecure for many applications. This is chiefly due to the 56-bit key size
being too small. The algorithm is believed to be practically secure in the form
of triple DES, although there are theoretical attacks. In recent years, the cipher
has been superseded by the advanced encryption standard. RSA25 (Biham &
Shamir 1999; Daemen & Rijmen, 2002) is a cryptosystem, which is known
as one of the first practicable public-key cryptosystems and is widely used for
secure data transmission. In such a cryptosystem, the encryption key is public
and differs from the decryption key which is kept secret.
In RSA, this asymmetry is based on the practical difficulty of factoring
the product of two large prime numbers, the factoring problem. However,
in recent years, other weak encryption methods have been used, e.g.,
ROT13 (rotation of 13 characters, without key), Caesars code (or Caesar
shift) which is a type of substitution cipher in which each letter in the
23
The system must be practically, if not mathematically, indecipherable. It must not be required to be
secret, and it must be able to fall into the hands of the enemy without inconvenience. Its key must be
communicable and retainable without the help of written notes, and changeable or modifiable at the
will of the correspondents. It must be applicable to telegraphic correspondence. It must be portable,
and its usage and function must not require the concourse of several people. Finally, it is necessary,
given the circumstances that command its application, that the system be easy to use, requiring neither
mental strain nor the knowledge of a long series of rules to observe.
24
DES, data encryption standard.
25
RSA stands for Ron Rivest, Adi Shamir, and Leonard Adleman, who first publicly described the algo-
rithm in 1977. Clifford Cocks, an English mathematician, had developed an equivalent system in
1973, but it was not declassified until 1997.
Recent Advances in Optical Image Processing 239
Brosseau, 2009) and image coding with an optical FT (Refregier & Javidi,
1995) or a fractional FT (Sahin, Ozaktas, & Mendlovic, 1995). There is a
wealth of research on encryption methods based on FT (Alfalou &
Brosseau, 2009). However, alternative approaches using the properties of
the DCT spectral fusion method are known to be very useful for encryption.
For example, Liu et al. (2011) proposed a color image encryption algorithm
by making use of the Arnold transform and DCT. The RGB components of
the color image are scrambled by the Arnold transform at the level of pixel
sequence. The scrambled RGB components are exchanged and mixed ran-
domly under the control of a matrix defined by a random angle. DCT is
employed for changing the pixel values of color image. In this encryption
scheme, the operations mentioned above are performed twice continuously.
The parameters of the Arnold transform and the random angle serve as the
key of the color image encryption method. Some numerical simulations are
made to test the validity and capability of the color encryption algorithm.
Numerical simulations have been reported to verify the color coding ability
of this algorithm (Liu et al., 2011).
The principle of other methods is briefly described.
Color image compression algorithm based on the DCT transform com-
bined to an adaptive block scanning
Remarkable success has been achieved in the low-loss compression of color
images (Douak, Benzid, & Benoudjit, 2011). After a preprocessing step
(mean removing and transforming the RGB components in a specific decor-
relating transform), the DCT transform is applied and followed by an iter-
ative phase (using the bisection method) including the thresholding,
quantization, dequantization, inverse DCT, inverse transform to get the
RGB components, and mean recovering. The efficiency of this method
was demonstrated with a series of experiments.
Key management of the double random-phase-encoding method using
public-key encryption
Saini and Sinha (2010) suggested that public-key encryption can be been
used to encode the key of the encryption process. In their technique, an
input image is encrypted by using the DRP encoding method using
extended fractional FT. The key of the encryption process is encoded by
using the RSA public-key encryption algorithm. The encoded is then trans-
mitted to the receiver side along with the encrypted image. In the decryp-
tion process, first the encoded key is decrypted using the secret key and then
the encrypted image is decrypted by using the retrieved key parameters. The
proposed technique has an advantage over DRP encoding method because
Recent Advances in Optical Image Processing 243
the problem associated with the transmission of the key has been eliminated
by using public-key encryption. The robust capability of this technique has
been illustrated by many computer simulations.
Image encryption based on the multiple-parameter discrete fractional FT
and chaos function
Lang, Tao, and Wang (2010) reported on a fractional Fourier method
for image encryption based on the multiple-parameter discrete fractional
Fourier transform and chaotic logistic maps in order to meet the require-
ments of the secure image transmission. In the proposed image encryption
scheme, the image is encrypted by juxtaposition of sections of the image in
the multiple-parameter discrete fractional Fourier domains and the align-
ment of sections is determined by chaotic logistic maps. This method does
not require the use of phase keys. The new method has been compared with
several existing methods and shows comparable or superior robustness to
blind decryption.
Multiple-image parallel optical encryption
Yong-Liang, Xin, Sheng, and Yao-Yao (2010) presented a method that
enables multiple-image optical encryption, in which a set of parallel plain-
texts can be extracted from the same designed ciphertext. This approach uses
a 6f optical setup. The principle of random phase encoding is utilized, and
the phase keys corresponding to different plaintexts are achieved indepen-
dently from the same designed ciphertext by cascade phase-retrieval algo-
rithm. Based on this general idea and on fusion, this method leads to an
independent decryption. The advantages of the approach could be con-
cluded as implementing decryption without cross talk, infinite encrypted
capacity and simple architecture. The plaintexts extracted mode can be also
extended from peer-to-peer to peer-to-multipeer.
One-step multiplexing optical encryption
Barrera and Torroba (2010) presented a combination of a free propagation
scheme and a classical 4f encrypting architecture to design a technique to
simultaneously encrypt multiple input objects. Unlike existing multiplexing
methods, encryption can be performed without setup modifications. An
interesting point of the approach presented in this paper is that the fusion
and encryption operations can be achieved in a single step. Moreover, it
requires only a single light source. The first encrypting masks are not placed
at the object plane, thus serving as an extra safety parameter. Additionally, the
multiplexing encryption scheme does not require setup modifications.
Authorized users need a master mask and the individual mask for the allowed
object, as well as the involved Fresnel distances. Due to the above described
244 Ayman Alfalou and Christian Brosseau
Figure 98 Illustration of the Liu et al. (2010) algorithm: (a) original image; (b) its Fourier
transform; and (c) image after decryption. Liu et al. (2010).
Recent Advances in Optical Image Processing 245
Alfalou, Brosseau, et al., 2011). These techniques open the way to a fuller
control of the reconstruction of target images. In the following, we will
focus our OIP analysis on the 4f system. The justification for such an
approach follows from the fact that image spectra can be manipulated
between the reading and formation steps of the process. Using such an
approach, the correlation between images has most recently been studied
in numerical simulations and experimental observations for face recognition
applications (Alfalou, Brosseau, & Alam, 2013; Katz et al., 2012). Very
recently, we pointed out that the 4f system provides a consistent way to
compress and encrypt an image with a specific filter (Alfalou & Brosseau,
2009). On the one hand, redundant information can be suppressed. On
the other hand, changing the distribution of the frequencies in images
can dramatically change the representation of the data and allows us to ren-
der them useless to a hacker (Alfalou & Brosseau, 2009).
Here, we restrict to a brief presentation of the approach reported by
Aldossari, Alfalou, and Brosseau (2013) to encrypt and compress simulta-
neously multiple images from a video sequence. The idea here is to use
an adaptation of a spectral fusion technique developed in Alfalou and
Brosseau (2010a, 2010b). On the one hand, the compression is based on
the RMS duration criterion of the spectrums image. That is, we determine
the size of the useful spectra for each target image by exploiting the RMS
criterion. This parameter is used to determine the allowed area of each target
image within the compressed spectrum. Moreover, this parameter is adapted
in order to minimize overlapping between the different spectra. Then, the
spectra are merged together by making use of a segmentation criterion. The
latter compares the local energy relative to each pixel for each spectrum.
Furthermore, it optimizes assignment of the considered pixel by taking into
account the adjacent areas to the considered pixel. This permits avoiding the
presence of isolated areas and small-sized areas (less than 10 pixels). On the
other hand, the encryption part is optimized by using and merging several
encryption keys (according to the number of multiple images considered).
The validation of this approach was realized by using a VLC architecture,
phase-only correlation filter, and peak-to-correlation energy as decision cri-
terion. The reader is referred to Aldossari et al. (2013) for greater detail about
the application of this technique for analysis of video sequences.
place in the area of OIP. With apologies, we have omitted several aspects of
this subject, and for the topics discussed here, we primarily used examples
from our research group. We focused our attention on optically
implementable methods which can be efficiently and rapidly processed.
From the experimental side, we showed that these methods are adapted
to denoise image, enhance edges, and realize filtering. The study of spatial
spectrum brings information about spatial distribution of intensity in the
image and their distinctive directions; i.e., high spatial frequencies are
responsible for fast and strong changes in image intensity (noise, edges),
while low spatial frequencies are responsible for smooth and slow changes
in image intensity. These techniques are not only helpful in 2D but also
allow us to manipulate 3D holograms. Furthermore, we showed several
examples which demonstrate that these techniques are of major importance
to the current concerns on the breast cancer problems. Taken together, the
works described in this review have provided a pleasingly complete analysis
of underwater polarization-filtered images. A recent application is the detec-
tion and classification of microcalcifications in a mammogram for early diag-
nosis of breast cancer. The achievement of these experiments in image
compression and encryption is also outstanding.
Overall, OIP is useful in applications in which a high parallelism and real-
time processing can be effectively realized. OIP is still in its early days, and
there are a number of directions into which the field is likely to move in the
coming years. PO associated with correlation techniques and Fourier optics
has particular sensitivity to detect and recognize objects in scattering or
poor-visibility media, addressing the venerable questions that persist in
the field of OIP, i.e., can we provide optimized optical techniques for den-
oising (deblurring, etc.) images in order to extract more and more features?
while introducing new questions, e.g., how optical neural networks, e.g.,
optical implementation of Hopfield model, can be done? What can we
do with all of this high speed processing? From an experimental perspective,
it would be equally interesting to use recent progress made in snapshot
polarization imaging to obtain measurements instantaneously (Kohlgraf-
Owens & Dogariu, 2009; Tyo et al., 2006). The diversity of this review sug-
gests that a rich spectrum of additional optical techniques of OIP is awaiting
discovery.
ACKNOWLEDGMENTS
Having a group working directly together or on related problems has been a source of
strength for us and the congenital atmosphere over the years has made our efforts so
Recent Advances in Optical Image Processing 247
enjoyable. We have been lucky to work with N. Abdallah, A. Arnold-Bos, B.-E. Benkelfat,
F. Bouzidi, P. Delrot, M. Dubreuil, M. Elbouz, M. Jridi, P. Katz, I. Leonard, A. Mansour,
Y. Ouerhani, Neji Nihel, and S. Qasmi. It was also great pleasure to work with A. Dogariu,
M. S. Alam, and J. Zallat. We also appreciate the support of Lab-STICC (Lab-STICC is
UMR CNRS 6285).
REFERENCES
Aggoun, A. (2006). Pre-processing of integral images for 3-D displays. Journal of Display Tech-
nology, 2, 393400. http://www.opticsinfobase.org/jdt/abstract.cfm?URIjdt-2-4-393.
Aggoun, A. (2011). Compression of 3D integral images using 3D wavelet transform. Journal of
Display Technology, 7, 586. http://www.opticsinfobase.org/jdt/abstract.cfm?URIjdt-
7-11-586.
Ahmed, N., Natarajan, T., & Rio, K. R. (1974). Discrete cosine transform. IEEE Transactions
on Computers, C-23, 90.
Alam, M. S. (2013). Optical pattern recognition. In M. Dekker (Ed.), Encyclopedia of optical
engineering (pp. 17781797). New York: Marcel Dekker.
Aldossari, M., Alfalou, A., & Brosseau, C. (2013). Optimized fusion method based on
adaptation of the RMS time-frequency criterion for simultaneous compression and
encryption of multiple images. In Proceedings of SPIE 8748, optical pattern recognition
XXIV, 87480B. http://dx.doi.org/10.1117/12.2014465.
Alfalou, A. (1999). Implementation of optical multichannel correlators: Application to pattern recog-
nition. Ph.D. thesis, Rennes-France: Universite de Rennes 1ENST Bretagne.
Alfalou, A., & Brosseau, C. (2009). Optical image compression and encryption methods.
Advances in Optics and Photonics, 1, 589. http://dx.doi.org/10.1364/AOP.1.000589.
Alfalou, A., & Brosseau, C. (2010a). Exploiting root-mean-square time-frequency structure
for multiple-image optical compression and encryption. Optics Letters, 35, 19141916.
http://dx.doi.org/10.1364/OL.35.001914.
Alfalou, A., & Brosseau, C. (2010b). Understanding correlation techniques for face recog-
nition: From basics to applications. In M. Oravec (Ed.), Face recognition book (p. 354):
India: INTECH. http://sciyo.com/articles/show/title/understanding-correlation-
techniques-for-face-recognition-from-basics-to-applications.
Alfalou, A., & Brosseau, C. (2013). Implementing compression and encryption of phase-
shifting digital holograms for three-dimensional object reconstruction. Optics Communi-
cation, 307, 6772.
Alfalou, A., Brosseau, C., Abdallah, N., & Jridi, M. (2011). Simultaneous fusion, compres-
sion, and encryption of multiple images. Optics Express, 19, 24023. http://dx.doi.org/10.
1364/OE.19.024023.
Alfalou, A., Brosseau, C., Abdallah, N., & Jridi, M. (2013). Assessing the performance of a
method of simultaneous compression and encryption of multiple images and its resistance
against various attacks. Optics Express, 21, 8025. http://dx.doi.org/10.1364/OE.21.
008025.
Alfalou, A., Brosseau, C., & Alam, M. S. (2013). Smart pattern recognition. In Proceedings
of SPIE 8748, optical pattern recognition XXIV, 874809. http://dx.doi.org/
10.1117/12.2018249.
Alfalou, A., Brosseau, C., Benkelfat, B.-E., Qasmi, S., & Leonard, I. (2012). Towards all-
numerical implementation of correlation. In Proceedings of SPIE 8398, optical pattern rec-
ognition XXIII, 839809. http://dx.doi.org/10.1117/12.919378.
Alfalou, A., Brosseau, C., Katz, P., & Alam, M. S. (2012). Decision optimization for face
recognition based on an alternate correlation plane quantification metric. Optics Letters,
37, 1562. http://dx.doi.org/10.1364/OL.37.001562.
248 Ayman Alfalou and Christian Brosseau
Alfalou, A., Elbouz, M., Mansour, A., & Keryer, G. (2010). New spectral image compression
method based on an optimal phase coding and the RMS duration principle. Journal of
Optics, 12, 115403. http://dx.doi.org/10.1088/2040-8978/12/11/115403.
Alfalou, A., Keryer, G., & De Bougrenet De la Tocnaye, J. L. (1999). Optical implementa-
tion of segmented composite filtering. Applied Optics, 38, 61296135. http://www.
opticsinfobase.org/ao/abstract.cfm?URIao-38-29-6129.
Alfalou, A., Mansour, A., Elbouz, M., & Brosseau, C. (2011). Optical compression scheme to
multiplex & simultaneously encode images. In G. Cristobal, P. Schelkens, &
H. Thienpont (Eds.), Optical and digital image processing fundamentals and applications
(p. 463): Hoboken, NJ (USA): Wiley, ISBN: 978-3-527-40956-3.
Andonian, A. T. (1982). Correction for rigid body motion in moire topography. Applied
Optics, 21, 12531258. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-21-
7-1253.
Armitage, D., Thackara, J. l., & Eades, W. D. (1989). Photoaddressed liquid crystal spatial
light modulators. Applied Optics, 28, 4763. http://www.opticsinfobase.org/ao/abstract.
cfm?URIao-28-22-4763.
Atchia, Y., Levy, H., Dufour, S., & Levi, O. (2013). Rapid multiexposure in vivo brain
imaging system using vertical cavity surface emitting lasers as a light source. Applied
Optics, 52, C64C71. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-52-7-
C64.
Babkina, T. M., & Voloshinov, V. B. (2001). A new method of acousto-optic image
processing and edge enhancement. Journal of Optics A: Pure and Applied Optics, 3, S54.
http://dx.doi.org/10.1088/1464-4258/3/4/359.
Baker, A. (1981). Image detection and enhancement. Applied Optics, 20, 3917. http://www.
opticsinfobase.org/ao/abstract.cfm?URIao-20-22-3917.
Banerjee, P. K., & Datta, A. K. (2012). Techniques of frequency domain correlation for
face recognition and its photonic implementation. In Q. Adamo & C. M. Epifano
(Eds.), Face recognition: Methods, applications and technology (pp. 165186): New York
(USA): Nova Publisher. https://www.novapublishers.com/catalog/product_info.php?
products_id28370 (Chapter 9).
Bara, S., Jaroszewicz, Z., Kolodziejczyk, A., & Moreno, V. (1991). Determination of basic
grids for subtractive moire patterns. Applied Optics, 30, 12581262. http://www.
opticsinfobase.org/ao/abstract.cfm?URIao-30-10-1258.
Barat, C., & Phlypo, R. (2010). A fully automated method to detect and segment a man-
ufactured object in an underwater color image. EURASIP Journal on Advances in Signal
Processing, 2010, 110. http://dx.doi.org/10.1155/2010/568092.
Barlaud, M., & Labit, C. (2002). Compression et codage des images et des videos. Paris, France: Ed.
Hermes-Lavoisier.
Barrera, J. F., & Torroba, R. (2010). One step multiplexing optical encryption. Optics
Communication, 283, 12681272. http://dx.doi.org/10.1016/j.optcom.2009.11.083.
Barrett, P. (1987). Implementing the Rivest Shamir and Adleman public key encryption
algorithm on a standard digital signal processor. In Lecture notes in computer science: 263.
Proceedings of advances in cryptology-Crypto85 (p. 311): London (UK): Springer-Verlag.
Bartolini, L., De Dominicis, L., Ferri de Collibus, M., Fornetti, G., Francucci, M.,
Guarneri, M., et al. (2007). Polarimetry as tool to improve phase measurement in
amplitude modulated laser for submarine archaeological sites inspection. In O3A: Optics
for arts, architecture, and archaeology. Proceedings of SPIE 6618 (p.66180I). http://dx.doi.org/
10.1117/12.725320.
Bartolini, L., De Dominicis, L., Fornetti, G., Francucci, M., Guarneri, M., Poggi, C., et al.
(2007). Improvement in underwater phase measurement of an amplitude-modulated
laser beam by polarimetric techniques. Optics Letters, 32, 14021404. http://www.
opticsinfobase.org/ol/abstract.cfm?URIol-32-11-1402.
Recent Advances in Optical Image Processing 249
Benkelfat, B.-E., El Wardi, S., Zghal, M., & Alfalou, A. (2012). 2D real-time arithmetic
operations using optical coherence properties: Image processing applications. Proceedings
of SPIE, 8397, 83970D.
Berriel, R., Bescos, J., & Santisteban, A. (1983). Image restoration for a defocused optical
system. Applied Optics, 22, 2772. http://www.opticsinfobase.org/ao/abstract.cfm?
URIao-22-18-2772.
Bicout, D., & Brosseau, C. (1992). Multiply scattered waves through a spatially random
medium: Entropy production and depolarization. Journal of Physics I, 2, 20472063.
http://dx.doi.org/10.1051/jp1:1992266.
Bicout, D., Brosseau, C., Martinez, A. S., & Schmitt, J. M. (1994). Depolarization of mul-
tiply scattered waves by spherical diffusers: Influence of the size parameter. Physical
Review E, 49, 17671770. http://link.aps.org/doi/10.1103/PhysRevE.49.1767.
Biham, E., & Shamir, A. (1999). Power Analysis of the Key Scheduling of the AES Candi-
dates. In Proceedings of the second AES candidate conference, printed by The National Institute of
Standards and Technology, Gaithersburg, MD (p. 115).
Billingsley, F. C. (1970). Applications of digital image processing. Applied Optics, 9, 289299.
Birch, P., Tan, S., Young, R., Koukoulas, T., Claret-Tournier, F., Budgett, D., et al. (2001).
Experimental implementation of a Wiener filter in a hybrid digitaloptical correlator. Optics
Letters, 26, 494. http://www.opticsinfobase.org/ol/abstract.cfm?URIol-26-8-494.
Blinn, J. F. (1993). Whats the deal with the DCT? IEEE Computer Graphics and Applications,
13, 78.
Boffety, M., Galland, F., & Allais, A. G. (2012). Influence of polarization filtering on image
registration precision in underwater conditions. Optics Letters, 37, 32733275. http://
www.opticsinfobase.org/ol/abstract.cfm?URIol-37-15-3273.
Born, M., & Wolf, E. (1970). Principles of optics (4th ed.). Oxford: Pergamon Press.
Brosseau, C. (1995). Statistics of the normalized Stokes parameters for a Gaussian stochastic
plane wave field. Applied Optics, 34, 47884793. http://dx.doi.org/10.1364/AO.34.
004788.
Brosseau, C. (1998). Fundamentals of polarized light: A statistical optics approach. New York:
Wiley, ISBN: 0471143022.
Brosseau, C., & Barakat, R. (1991). Jones and Mueller polarization matrices for random media,
Optics Communication, 84, 127133. http://dx.doi.org/10.1016/0030-4018(91)90214-X.
Brosseau, C., & Bicout, D. (1994). Entropy production in multiple scattering of light by a
spatially random medium. Physical Review E, 50, 49975005. http://link.aps.org/doi/
10.1103/PhysRevE.50.4997.
Brosseau, C., Givens, C. R., & Kostinski, A. B. (1993). Generalized trace condition on the
Mueller-Jones polarization matrix. Journal of the Optical Society of America. A, 10,
22482251. http://dx.doi.org/10.1364/JOSAA.10.002248.
Buytaert, J. A. N., & Dirckx, J. J. J. (2007). Design considerations in projection phase-shift
moire topography based on theoretical analysis of fringe formation. Journal of the Optical
Society of America. A, 24, 20032013. http://www.opticsinfobase.org/josaa/abstract.cfm?
URIjosaa-24-7-2003.
Buytaert, J. A. N., & Dirckx, J. J. J. (2008). Moire profilometry using liquid crystals for pro-
jection and demodulation. Optics Express, 16, 179193. http://www.opticsinfobase.org/
oe/abstract.cfm?URIoe-16-1-179.
Cao, F., Guichard, F., & Hornung, H. (2010). Information capacity: A measure of potential
image quality of a digital camera. Proceedings of SPIE, 7537, 75370F.
Cariou, J., Le Jeune, B., Lotrian, J., & Guern, Y. (1990). Polarization effects of seawater and
underwater targets. Applied Optics, 29, 16891695. http://www.opticsinfobase.org/ao/
abstract.cfm?URIao-29-11-1689.
Casasent, D. (1979). Performance evaluation of spatial light modulators. Applied Optics, 18,
2445. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-18-14-2445.
250 Ayman Alfalou and Christian Brosseau
Casasent, D., & Xia, S. F. (1986). Phase correction of light modulators. Optics Letters, 11, 398.
http://www.opticsinfobase.org/ol/abstract.cfm?URIol-11-6-398.
Castillo, I. M. D., Sanchez-de-la-Llave, D. D., Garca, R. R., Olivos-Perez, L. I.,
Gonzalez, L. A., & Rodrguez-Ortiz, M. M. (2001). Real-time self-induced nonlinear
optical Zernike-type filter in a bacteriorhodopsin film. Optical Engineering, 40,
23672368. http://dx.doi.org/10.1117/1.1412425.
Catrysse, P. B., Imai, F. H., Linne von Berg, D. C., & Sheridan, J. T. (2013). Imaging systems
and applications. Applied Optics, 52, ISA1ISA3. http://www.opticsinfobase.org/ao/
abstract.cfm?URIao-52-7-ISA1.
Catrysse, P. B., & Skauli, T. (2013). Pixel scaling in infrared focal plane arrays. Applied Optics,
52, C72C77. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-52-7-C72.
Chambah, M., Semani, D., Renouf, A., Courtellemont, P., & Rizzi, A. (2003). Underwater
color constancy: Enhancement of automatic live fish recognition. In R. Eschbach, &
G. G. Marcu (Eds.), Proceedings of SPIE: Vol. 5293. Color imaging IX: Processing, hardcopy,
and applications (pp. 157168).
Chang, T. Y., Hong, J. H., & Yeh, P. (1990). Spatial amplification: An image-
processing technique using the selective amplification of spatial frequencies. Optics
Letters, 15, 743745. http://www.opticsinfobase.org/ol/abstract.cfm?URIol-15-13-743.
Chang, W. H., & Nguyen, T. Q. (2008). On the fixed-point accuracy analysis of FFT
algorithms. IEEE Transactions on Signal Processing, 56, 4673.
Chen, H. C. (2005). A memory-efficient realization of cyclic convolution and its application
to discrete cosine transform. IEEE Transactions on Circuits and Systems for Video Technology,
15, 445.
Chen, W., Smith, C., & Fralick, S. (1977). A fast computational algorithm for the discrete
cosine transform. IEEE Transactions on Communications, 25, 1004.
Clemenceau, P., Dogariu, A., & Stryewski, J. (2000). Polarization active imaging. In Laser
radar technology and applications V: Vol. 4035. Proceedings of SPIE (pp. 401409). USA.
Collins, S. A. (1970). Lens-system diffraction integral written in terms of matrix optics. Journal
of the Optical Society of America, 60, 11681177.
Cooley, J. W., & Tukey, J. W. (1965). An algorithm for the machine calculation of complex
Fourier series. Mathematics of Computation, 19, 297.
Cox, I., Miller, M., Bloom, J., Fridrich, J., & Kalker, T. (2007). Digital watermarking and steg-
anography (2nd ed.). San Francisco, CA: Elsevier.
Cox, I. J., & Sheppard, C. J. R. (1986). Information capacity and resolution in an optical
system. Journal of the Optical Society of America. A, 3, 11521158.
Daemen, J., & Rijmen, V. (2002). The design of Rijndael: AESThe advanced encryption
standard. London (UK): Springer.
Darakis, E., Naughton, T., & Soraghan, J. J. (2007). Compression defects in different recon-
struction from phase-shifting digital holographic data. Applied Optics, 46, 45794586.
http://www.opticsinfobase.org/ao/abstract.cfm?URIao-46-21-4579.
Darakis, E., & Soraghan, J. J. (2006). Compression of interference patterns with application
to phase-shifting digital holography. Applied Optics, 45, 24372443. http://www.
opticsinfobase.org/ao/abstract.cfm?URIao-45-11-2437.
Darakis, E., & Soraghan, J. J. (2007a). Reconstruction domain compression of phase-shifting
digital holograms. Applied Optics, 46, 351356. http://www.opticsinfobase.org/ao/
abstract.cfm?URIao-46-3-351.
DeCusatis, C., & Das, P. (1995). Signal processing using wavelet transforms. Optics &
Photonics News, 6, 39. http://www.opticsinfobase.org/opn/abstract.cfm?URIopn-6-
12-39.
Denz, C., Dellwig, T., Lembcke, J., & Tschudi, T. (1996). Parallel optical image addition and
subtraction in a dynamic photorefractive memory by phase-code multiplexing. Optics
Letters, 21, 278280.
Recent Advances in Optical Image Processing 251
Dickey, T. D., Kattawar, G. W., & Voss, K. J. (2011). Shedding new light on light in the
ocean. Physics Today, 64, 4449.
Ding, J. J. (2001). Research of fractional Fourier transform and linear canonical transform. Taiwan:
National Taiwan University, Ph.D. dissertation.
Ding, L., Yan, Y., Xue, Q., & Jin, G. (2003). Wavelet packet compression for volume
holographic image recognition. Optics Communication, 216, 105113.
Douak, F., Benzid, R., & Benoudjit, N. (2011). Color image compression algorithm based
on the DCT transform combined to an adaptive block scanning. AEU: International
Journal of Electronics and Communications, 65, 1626. http://dx.doi.org/10.1016/j.aeue.
2010.03.003.
Dubreuil, M., Alfalou, A., & Brosseau, C. (2012). Imaging polarimetry and correlation tech-
niques for target detection. In Proceedings of Frontiers in optics 2012/Laser Science XXVIII,
OSA Technical Digest (online) (Optical Society of America, 2012), paper FW3A.30. http://
www.opticsinfobase.org/abstract.cfm?URIFiO-2012-FW3A.30.
Dubreuil, M., Delrot, P., Leonard, I., Alfalou, A., Brosseau, C., & Dogariu, A. (2013).
Exploring underwater target detection by imaging polarimetry and correlation tech-
niques. Applied Optics, 52, 9971005. http://www.opticsinfobase.org/ao/abstract.cfm?
URIao-52-5-997.
Duncan, R. C. (1962). Americas use of sea mines. White Oak: United States Naval Ordinance
Laboratory.
Duntley, S. Q. (1963). Light in the sea. Journal of the Optical Society of America, 53, 214233.
http://dx.doi.org/10.1364/JOSA.53.000214.
Duraffourg, L., Merolla, J. M., Goedgebuer, J. P., Mazurenko, Y., & Rhodes, W. T. (2001).
Compact transmission system using single-sideband modulation of light for quantum
cryptography. Optics Letters, 26, 1427.
Duvaut, P. (1994). Traitement du signal: Concepts et applications. Paris: Herme`s.
Ebersole, J. F. (1975). Optical image subtraction. Optical Engineering, 14, 436447.
Ebersole, J. F., & Wyant, J. C. (1976). Real-time optical subtraction of photographic imagery
for difference detection. Applied Optics, 15, 871876.
Egami, C., Suzuki, Y., Uemori, T., Sugihara, O., & Okamoto, N. (1997). Self-adaptive
spatial filtering by use of azo chromophores doped in low glass-transition-temperature
polymers. Optics Letters, 22, 14241426. http://www.opticsinfobase.org/ol/abstract.
cfm?URIol-22-18-1424.
Egan, J. P. (1975). Signal detection theory and ROC analysis. New York: Academic Press Series
in Cognition and Perception.
Elbouz, M., Alfalou, A., & Brosseau, C. (2011). Fuzzy logic and optical correlation-based
face recognition method for patient monitoring application in home video surveillance.
Optical Engineering, 50, 067003. http://dx.doi.org/10.1117/1.3582861.
Elbouz, M., Bouzidi, F., Alfalou, A., Brosseau, C., Leonard, I., & Benkelfat, B.-E. (2013).
Adapted all-numerical correlator for face recognition applications. In Optical pattern rec-
ognition XXIV: Vol. 8748. Proceedings of SPIE (p. 874807). USA. http://dx.doi.org/10.
1117/12.2014383.
Espinosa, J., Mas, D., Perez, J., & Roig, A. B. (2013). Open-access operating algorithms for
commercial videokeratographer and improvement of corneal sampling. Applied Optics,
52, C24C29. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-52-7-C24.
Fales, C. L., Huck, F. O., & Samms, R. W. (1984). Imaging system design for improved
information capacity. Applied Optics, 23, 872888.
Fan, L., Song, F., & Jutamulia, S. (2007). Edge detection with large depth of focus using
differential HaarGaussian wavelet transform. Optics Communications, 270, 169175.
http://www.sciencedirect.com/science/article/pii/S0030401806009485.
Fellgett, P. B., & Linfoot, E. H. (1955). On the assessment of optical images. Philosophical
Transactions of the Royal Society A, 247, 369407.
252 Ayman Alfalou and Christian Brosseau
Ferrari, J. A., Garbusi, E., Frins, E. M., & Priz, G. (2005). Spatial self-filtering with polarizer
sheets. Applied Optics, 44, 4146. http://www.opticsinfobase.org/ao/abstract.cfm?
URIao-44-1-41.
Fienup, J. R. (1982). Phase retrieval algorithms: A comparison. Applied Optics, 21,
27582769. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-21-15-2758.
Fienup, J. R. (2013). Phase retrieval algorithms: A personal tour [Invited]. Applied Optics, 52,
4556. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-52-1-45.
Fleck, B., Jahn, D., & Wenke, L. (1998). A nonlinear optical element for incoherent image
subtraction based on photoanisotropy. Optical Communications, 154, 339344.
Flores, J. L., & Ferrari, J. A. (2010). Orientation-selective edge detection and enhancement
using the irradiance transport equation. Applied Optics, 49, 619624. http://www.
opticsinfobase.org/ao/abstract.cfm?URIao-49-4-619.
Flores, J. L., Ferrari, J. A., & Garcia-Torales, G. (2011). Partially coherent optical processor
for enhancement of partial second order derivatives of an image: Analysis and implemen-
tation. Optical Communications, 284, 56305635. http://www.sciencedirect.com/sci
ence/article/pii/S0030401811009485.
Fu, S., & Zhang, C. (2012). Fringe pattern denoising using averaging based on nonlocal self-
similarity. Optics Communications, 285, 25412544. http://dx.doi.org/10.1016/j.
optcom.2012.01.059.
Garfinkel, L., Boring, C. C., & Heath, C. W., Jr. (1994). Changing trends. An overview of
breast cancer incidence and mortality. Cancer, 74, 222227.
Ghael, S. P., Sayeed, A. M., & Baraniuk, R. G. (1997). Improved wavelet denoising via
empirical Wiener filtering. In Wavelet applications in signal and image processing V:
Vol. 3169. Proceedings of SPIE (p. 389). USA. http://dx.doi.org/10.1117/12.292799.
Ghosh, A., & Murata, K. (1988). Real-time addition and subtraction of two images by vector
wave interferometry. Optics and Laser Technology, 20, 211215.
Gilbert, G. D., & Pernicka, J. C. (1967). Improvement of underwater visibility by reduction
of backscatter with a circular polarization technique. Applied Optics, 6, 741746. http://
www.opticsinfobase.org/ao/abstract.cfm?URIao-6-4-741.
Glenn, D. B. (1968). A fast Fourier transform algorithm for teal-valued series. Communications
of the ACM, 11, 703.
Gl
uckstad, J. (1995). Adaptive array illumination and structured light generated by spatial
zero-order self-phase modulation in a Kerr medium. Optics Communication, 120,
194203.
uckstad, J., Lading, L., Toyoda, H., & Hara, T. (1997). Lossless light projection. Optics Letters,
Gl
22, 13731375. http://www.opticsinfobase.org/ol/abstract.cfm?URIol-22-18-1373.
Goodman, J. W. (1966). Introduction to Fourier optics (2nd ed.). New York: McGraw-Hill.
Goodman, J. (1976). Some fundamental properties of speckle. Journal of the Optical Society of
America, 66, 11451150.
Goole (2013). http://fr.wikipedia.org/wiki/Moir%C3%A9_%28effet_de_contraste%29.
Gorman, G. A. (2011). Field deployable dynamic lightning system for turbid water imaging. M.S.
Thesis, Cambridge, USA: Massachusetts Institute of Technology.
Gourier, N., Hall, D., & Crowley, J. L. (2004). Estimating face orientation from robust
detection of salient facial features. In ICPR, International workshop on visual observation
of deictic gestures. Proceedings of Pointing 2004.
Granier, B., Refregier, P., & Figue, J. (1996). Restoration of images degraded by atmospheric
turbulence by a least-squares method and a Markov process. Optics Letters, 21, 423425.
http://www.opticsinfobase.org/ol/abstract.cfm?URIol-21-6-423.
Guglielmi, V., Fournier-Prunaret, D., Taha, A. K., Pinel, P., Rouabhi, S., & Beneteau, L.
(2002). Encryption using chaotic dynamics of two noninvertible maps. In Proceedings of
workshop on nonlinear dynamics of electronic systems (NDES02), Izmir, Turquie.
Recent Advances in Optical Image Processing 253
Gurov, I. P., & Sheynihovich, D. V. (2000). Interferometric data analysis based on Markov
nonlinear filtering methodology. Journal of the Optical Society of America. A, 17, 2127.
http://www.opticsinfobase.org/josaa/abstract.cfm?URIjosaa-17-1-21.
Hardie, R. C., & Barnard, K. J. (2012). Fast super-resolution using an adaptive Wiener filter
with robustness to local motion. Optics Express, 20, 21053.
Harris, J. L., Sr. (1966). Image evaluation and restoration. Journal of the Optical Society of
America, 56, 569. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-30-10-1258.
Hartmann, G. K., & Truver, S. (1991). Weapons that wait: Mine warfare in the United States
Navy. Annapolis: Naval Institute Press.
Hecht, E. (2001). Optics (4th ed.). New York: Addison-Wesley.
Helan, S., Burie, J., Bouwmans, T., & Bazeille, S. (2006). Object detection in underwater
images. In Proceedings of CMM06, Caracterisation du milieu marin, 1619/10.
Hennelly, B. M., & Sheridan, J. T. (2005). Generalizing, optimizing, and inventing numer-
ical algorithms for the fractional Fourier, Fresnel, and linear canonical transforms. Journal
of the Optical Society of America. A, 22, 917927. http://dx.doi.org/10.1364/JOSAA.22.
000917.
Horner, J. L. (1992). Metrics for assessing pattern-recognition performance. Applied Optics,
31, 165166. http://dx.doi.org/10.1364/AO.31.000165.
Horner, J. L., & Gianino, P. D. (1984). Phase-only matched filtering. Applied Optics, 23,
812816.
Hou, H. (1987). A fast recursive algorithm for computing the discrete cosine transform. IEEE
Transactions on Acoustics, Speech, and Signal Processing, 35, 1455.
Hou, W., Gray, D., Weidemann, A., Fournier, G., & Forand, J. (2007). Automated under-
water image restoration and retrieval of related optical properties. In Geoscience and remote
sensing symposium. Proceedings of IEEE IGARSS (pp. 18891892). http://ieeexplore.ieee.
org/stamp/stamp.jsp?tp&arnumber4423193&isnumber4422708.
Hou, S., Zhouand, S., & Siddique, M. A. (2012). Query by example video based on fuzzy
c-means initialized by fixed clustering center. Optical Engineering, 51, 047405. http://dx.
doi.org/10.1117/1.OE.51.4.047405.
Huang, K. S., Jenkins, B. K., & Sawchuk, A. A. (1989). Image algebra representation of
parallel optical binary arithmetic. Applied Optics, 28, 12631278. http://www.
opticsinfobase.org/ao/abstract.cfm?URIao-28-6-1263.
Huang, G., Jin, G., Wu, M., & Yan, Y. (1996). Novel gray image processing in a single oper-
ation optical morphological processor. Optics Communication, 130, 143152. http://
www.sciencedirect.com/science/article/pii/0030401896002155.
Huang, K. S., Sawchuk, A. A., Jenkins, B. K., Chavel, P., Wang, J. M., Weber, A. G., et al.
(1993). Digital optical cellular image processor (DOCIP): Experimental implementation.
Applied Optics, 32, 166173. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-
32-2-166.
Huang, G., Wu, M., Jin, G., Yao, Z., & Yan, Y. (1994). The parallel processor for the char-
acter extraction of digitizing image. Chinese Journal of Lasers, 283288. http://www.
oriprobe.com/journals/caod_483/1994_3.html.
Ikeda, O., Sato, T., & Kojima, H. (1986). Construction of a Wiener filter using a phase-
conjugate filter. Journal of the Optical Society of America. A, 3, 645. http://www.
opticsinfobase.org/josaa/abstract.cfm?URIjosaa-3-5-645.
Inbar, H., & Marom, E. (1995). A priori and adaptive Wiener filtering with joint transform
correlators. Optics Letters, 20, 1050. http://www.opticsinfobase.org/ol/abstract.cfm?
URIol-20-9-1050.
Iqbal, K., Salam, R., Osman, A., & Talib, A. (2007). Underwater image enhancement using
an integrated colour model. IAENG International Journal of Computer Science, 1819-
656X.32, 529534. http://eprints.usm.my/id/eprint/9675.
254 Ayman Alfalou and Christian Brosseau
Jaffe, J. (1990). Computer modeling and the design of optimal underwater imaging systems.
IEEE Journal of Oceanic Engineering, 15, 101111. http://jaffeweb.ucsd.edu/files/
JaffeIEEEJOE.pdf.
Jiang, S. S., & Sawchuk, A. A. (1986). Noise updating repeated Wiener filter and other adap-
tive noise smoothing filters using local image statistics. Applied Optics, 25, 2326. http://
www.opticsinfobase.org/ao/abstract.cfm?URIao-25-14-2326.
Joseph, J., Aranda, F. J., Rao, D. V. G. L. N., Akkara, J. A., & Nakashima, M. (1996). Optical
Fourier processing using photoinduced dichroism in a bacteriorhodopsin film. Optics
Letters, 21, 14991501. http://www.opticsinfobase.org/ol/abstract.cfm?URIol-21-
18-1499.
Kato, H., & Goodman, J. W. (1975). Nonlinear filtering in coherent optical systems through
halftone screen processes. Applied Optics, 14, 18131824. http://www.opticsinfobase.
org/ao/abstract.cfm?URIao-14-8-1813.
Kato, J. I., Yamaguchi, I., Nakamura, T., & Kuwashima, S. (1997). Video-rate fringe analyzer
based on phase-shifting electronic moire patterns. Applied Optics, 36, 84038412, http://
www.opticsinfobase.org/ao/abstract.cfm?URIao-36-32-8403.
Kato, U. I., Yamaguchi, I., & Tanaka, H. (1996). Nonlinear spatial filtering with a dye-doped
liquid-crystal cell. Optics Letters, 21, 767769. http://www.opticsinfobase.org/ol/
abstract.cfm?URIol-21-11-767.
Kattawar, G. W., & Rakovic, M. J. (1999). Virtues of Mueller matrix imaging for underwater
target detection. Applied Optics, 38, 64316438. http://www.opticsinfobase.org/ao/
abstract.cfm?URIao-38-30-6431.
Katz, P., Alfalou, A., Brosseau, C., & Alam, M. S. (2012). Correlation and independent
component analysis based approaches for biometric recognition. In A. Quaglia, &
C. M. Epifano (Eds.), Face recognition: Methods, applications and technology
(pp. 201229): New York (USA): NOVA Publishers. https://www.novapublishers.
com/catalog/product_info.php?products_id28370.
Kerlikowske, K., & Barclay, J. (1997). Outcomes of modern screening mammography.
Journal of the National Cancer Institute. Monographs, 22, 105111.
Khalil, A. M., & Fazio, P. (1973). Lens distortion equation using moire. Applied Optics, 12,
972975. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-12-5-972.
Khoury, J., Gianino, P. D., & Woods, C. L. (2000). Wiener-like correlation filters.
Applied Optics, 39, 231. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-39-
2-231.
Kim, A. D., & Moscoso, M. (2002). Backscattering of circularly polarized pulses. Optics
Letters, 27, 15891591. http://www.opticsinfobase.org/ol/abstract.cfm?URIol-27-
18-1589.
Kim, Y., Park, G., Jung, J. H., Kim, J., & Lee, B. (2009). Color moire pattern simulation and
analysis in three-dimensional integral imaging for finding the moire-reduced tilted angle
of a lens array. Applied Optics, 48, 21782187. http://www.opticsinfobase.org/ao/
abstract.cfm?URIao-48-11-2178.
Kohlgraf-Owens, T., & Dogariu, A. (2009). Spatially resolved scattering polarimeter. Optics
Letters, 34, 13211323. http://www.opticsinfobase.org/ol/abstract.cfm?URIol-34-9-
1321.
Kolba, D., & Parks, T. (1977). A prime factor FFT algorithm using high-speed convolution.
IEEE Transactions on Acoustics, Speech, and Signal Processing, 25, 281.
Komorowska, K., Miniewicz, A., Parka, J., & Kajzar, F. (2002). Self-induced nonlinear
Zernike filter realized with optically addressed liquid crystal spatial light modulator.
Journal of Applied Physics, 92, 56355642. http://dx.doi.org/10.1063/1.1515949.
Kondo, K., Ichioka, Y., & Suzuki, T. (1977). Image restoration by Wiener filtering in
the presence of signal-dependent noise. Applied Optics, 16, 2554. http://www.
opticsinfobase.org/ao/abstract.cfm?URIao-16-9-2554.
Recent Advances in Optical Image Processing 255
Lim, J. S., Kim, J., & Chung, M. S. (1989). Additive type moire with computer image
processing. Applied Optics, 28, 26772680. http://www.opticsinfobase.org/ao/
abstract.cfm?URIao-28-13-2677.
Liu, Z., Centurion, M., Panotopoulos, G., Hong, J., & Psaltis, D. (2002). Holographic
recording of fast events on a CCD camera. Optics Letters, 27, 22. http://www.
opticsinfobase.org/ol/abstract.cfm?URIol-27-1-22.
Liu, H. K., & Chao, T. H. (1986). Optical image subtraction techniques, 1975-1985.
Proceedings of SPIE, 638, 5565.
Liu, L., Cui, B., & Zhao, L. (1994). Optical cellular fuzzy logic gray-scale image proces-
sor. Optics Communications, 107, 453. http://dx.doi.org/10.1016/0030-4018(94)
90363-8.
Liu, Z., Xu, L., Liu, T., Chen, H., Li, P., Lin, C., et al. (2011). Color image encryption by
using Arnold transform and color-blend operation in discrete cosine transform domains.
Optics Communications, 284, 123.
Liu, Z., Zhang, Y., Li, S., Liu, W., Liu, W., Wang, Y., et al. (2013). Double image
encryption scheme by using random phase encoding and pixel exchanging in the gyrator
transform domains. Optics & Laser Technology, 47, 152.
Liu, L., Zhang, Z., & Zhang, X. (1994). One-operation image algebra and optoelectronic
cellular two-layer logic array. Journal of the Optical Society of America. A, 11,
17891797. http://www.opticsinfobase.org/josaa/abstract.cfm?URIjosaa-11-6-1789.
Liu, Z., Zhang, Y., Zhao, H., Ashfaq Ahmad, M., & Liu, S. (2010). Optical multi-image
encryption based on frequency shift. Optik, 122, 10101013. http://dx.doi.org/
10.1016/j.ijleo.2010.06.039.
Marom, E., & Inbar, H. (1996). New interpretations of Wiener filters for image recognition.
Journal of the Optical Society of America. A, 13, 1325. http://www.opticsinfobase.org/josaa/
abstract.cfm?URIjosaa-13-7-1325.
Mazzaferri, J., Campos, J., Escalera, J. C., Sheppard, C. J. R., & Ledesma, S. (2010). Edge
detector tolerant to object defocusing. Optics Communications, 283, 36393645. http://
www.sciencedirect.com/science/article/pii/S0030401810004839.
Mazzaferri, J., & Ledesma, S. (2007). Rotation invariant real-time optical edge detector.
Optics Communications, 272, 367376. http://www.sciencedirect.com/science/article/
pii/S0030401806012971.
McGlamery, B. L. (1975). Computer analysis and simulation of underwater camera system
performance. San Diego: University of California. Scripps Institution of Oceanography,
Visibility Laboratory, San Diego, California 92152, http://misclab.umeoce.maine.
edu/education/VisibilityLab/reports/SIO_75-2.pdf.
Meher, P. K. (2006). Systolic designs for DCT using a low-complexity concurrent con-
volutional formulation. IEEE Transactions on Circuits and Systems for Video Technology,
16, 1041.
Mendlovic, D. (2013). Toward a super imaging system. Applied Optics, 52, 561566. http://
www.opticsinfobase.org/ao/abstract.cfm?URIao-52-4-561.
Mendlovic, D., & Konforti, N. (1993). Optical realization of the wavelet transform for two-
dimensional objects. Applied Optics, 32, 6542. http://www.opticsinfobase.org/ao/
abstract.cfm?URIao-32-32-6542.
Menke, N., Yao, B., Wang, Y., Zheng, Y., Lei, M., Ren, L., et al. (2006). Optical image
processing using the photoinduced anisotropy of pyrrylfulgide. Journal of the Optical
Society of America. A, 23, 267271. http://www.opticsinfobase.org/josaa/abstract.cfm?
URIjosaa-23-2-267.
Miao, Q. G., Shi, C., Xu, P. F., Yang, M., & Shi, Y. b. (2011). A novel algorithm of image
fusion using shearlets. Optics Communications, 284, 15401547. http://dx.doi.org/
10.1016/j.optcom.2010.11.048.
Mie, G. (1908). Beitrage zur optik tr osungen. Annalen
uber medien, speziell kolloidaler metall
der Physik, 330, 377445. http://dx.doi.org/10.1002/andp.19083300302.
Recent Advances in Optical Image Processing 257
Papoulis, A. (1962a). The Fourier integral and its applications. New York: McGraw-Hill.
Pati, G. S., Unnikrishnan, G., & Singh, K. (1998). Multichannel image addition and subtrac-
tion using joint-transform correlator architecture. Optics Communications, 150, 3337.
Paturzo, M., Memmolo, P., Miccio, L., Finizio, A., Ferraro, P., Tulino, A., et al. (2008a).
Numerical multiplexing and demultiplexing of digital holographic information for
remote reconstruction in amplitude and phase. Optics Letters, 33, 26292631. http://
www.opticsinfobase.org/ol/abstract.cfm?URIol-33-22-2629.
Paz, S., Zlotnik, A., & Zalevsky, Z. (2013). Extending the depth of focus in tomography
systems for glass lattice three dimensional mapping. Applied Optics, 52, C50C57.
http://www.opticsinfobase.org/ao/abstract.cfm?URIao-52-7-C50.
Pennebaker, W. B., & Mitchell, J. L. (1993). JPEG: Still image data compression standard.
New York: International Thompson Publishing.
Phong, B. (1975). Illumination for computer generated pictures. Communications of the ACM,
18, 311317. http://dl.acm.org/citation.cfm?id360839.
Piederrie`re, Y., Boulvert, F., Cariou, J., Le Jeune, B., Guern, Y., & Le Brun, G. (2005).
Backscattered speckle size as a function of polarization: Influence of particle-size and
-concentration. Optics Express, 13, 50305039. http://www.opticsinfobase.org/oe/
abstract.cfm?URIoe-13-13-5030.
Piestun, R., & Miller, D. A. B. (2000). Electromagnetic degrees of freedom of an optical
system. Journal of the Optical Society of America. A, 17, 892902.
Qian, Y., Hu, F., Cheng, X., & Jin, W. (2011). Real-time image deblurring by optoelec-
tronic hybrid processing. Applied Optics, 50, 61846188.
Quatieri, T., Jr., & Oppenheim, A. V. (1981). Iterative techniques for minimum phase signal
reconstruction from phase or magnitude. IEEE Transactions on Acoustics, Speech, and Signal
Processing, 29, 11871193.
Rajput, S. K., & Nishchal, N. K. (2013). Known-plaintext attack-based optical cryptosystem
using phase-truncated Fresnel transform. Applied Optics, 52, 871.
Rasouli, S., Ghasemi, H., Tavassoly, M. T., & Khalesifard, H. R. (2011). Application of
parallel moire deflectometry and the single beam Z-scan technique in the measure-
ment of the nonlinear refractive index. Applied Optics, 50, 23562360. http://www.
opticsinfobase.org/ao/abstract.cfm?URIao-50-16-2356.
Refregier, P., & Javidi, B. (1995). Optical image encryption based on input plane and Fourier
plane random encoding. Optics Letters, 20, 767.
Rehn, H., & Kowarschik, R. (1998). Real-time non-linear spatial filtering with a leaky
OASLM. Optics & Laser Technology, 30, 3947. http://www.sciencedirect.com/sci
ence/article/pii/S0030399298000115.
Richardson, W. H. (1972). Bayesian-based iterative method of image restoration. Journal of
the Optical Society of America, 62, 55.
Richardson, E. G. (2003). H.264 and MPEG-4 video compression: Video coding for next-
generation multimedia. Hoboken, NJ (USA): John Wiley & Sons Ltd.
Riris, H., Carlisle, C. B., Warren, R. E., & Cooper, D. E. (1994). Signal-to-noise ratio
enhancement in frequency-modulation spectrometers by digital signal processing. Optics
Letters, 19, 144. http://www.opticsinfobase.org/ol/abstract.cfm?URIol-19-2-144.
Rizzi, A., Gatta, C., & Marini, D. (2004). From retinex to automatic color equalization:
Issues in developing a new algorithm for unsupervised color equalization. Journal of Elec-
tronic Imaging, 13, 7584.
Rolando, S., Goiffon, V., Magnan, P., Corbie`re, E., Molina, R., Tulet, M., et al. (2013).
Smart CMOS image sensor for lightning detection and imaging. Applied Optics, 52,
C16C23. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-52-7-C16.
Rosen, J., Katz, B., & Brooker, G. (2009). Fresnel incoherent correlation hologramA
review. Chinese Optics Letters, 7, 11341141. http://www.opticsinfobase.org/col/
abstract.cfm?URIcol-7-12-1134.
Recent Advances in Optical Image Processing 259
Russ, J. C. (2002). The image processing hand book (4th ed.). USA: CRC Press.
Ryle, J. P., McDonnell, S., Glennon, B., & Sheridan, J. T. (2013). Calibration of a digital
in-line holographic microscopy system: Depth of focus and bioprocess analysis. Applied
Optics, 52, C78C87.
Sabbah, S., Lerner, A., Erlick, C., & Shashar, N. (2005). Underwater polarization visionA
physical examination. In S. G. Pandalai (Ed.), Recent research and developments in experimen-
tal & theoretical biology (pp. 123176). Kerala, India: Transworld Research Network.
Sahin, A., Ozaktas, H. M., & Mendlovic, D. (1995). Optics Communications, 120, 128134.
Saini, N., & Sinha, A. (2010). Key management of the double random-phase-encoding
method using public-key encryption. Optical Laser Engineering, 48, 329334.
Salomon, D. (2004). Data compression: The complete reference (3rd ed.). New York: Springer.
Sayood, S. (2012). Introduction to data compression. Morgan Kaufmann series in multimedia
information and systems. (4th ed.): Elsevier.
Scharf, L. L. (1991). Statistical signal processing: Detection, estimation, and time series analysis.
Prentice Hall.
Schechner, Y., & Karpel, N. (2005). Recovery of underwater visibility and structure by
polarization analysis. IEEE Journal of Oceanic Engineering, 30, 570587. http://dx.doi.
org/10.1109/JOE.2005.850871.
Sendhil, K., Vijayan, C., & Kothiyal, M. P. (2005). Spatial phase filtering with a porphyrin
derivative as phase filter in an optical image processor. Optics Communications, 251,
292298. http://dx.doi.org/10.1016/j.optcom.2005.03.014.
Shabtay, G., Mendlovic, D., & Zalevsky, Z. (1998). Optical implementation of the contin-
uous wavelet transform. Applied Optics, 37, 2964. http://www.opticsinfobase.org/ao/
abstract.cfm?URIao-37-14-2964.
Shamir, A. (1985). On the security of DES. In Lecture notes in computer science: Vol. 218.
Advances in cryptologyCrypto85 (p. 280). New York: Springer-Verlag.
Shih, F., & Mitchell, O. (1989). Threshold decomposition of gray-scale morphology
into binary morphology. IEEE Transactions in Pattern Analysis Machine Intelligence, 11,
3142.
Shih, M. Y., Shishido, A., Chen, P. H., Wood, M. V., & Khoo, I. C. (2000). All-optical
image processing with a supranonlinear dye-doped liquid-crystal film. Optics Letters,
25, 978980. http://www.opticsinfobase.org/ol/abstract.cfm?URIol-25-13-978.
Shih, M. Y., Shishido, A., & Khoo, I. C. (2001). All-optical image processing by means of a
photosensitive nonlinear liquid-crystal film: Edge enhancement and image addition-
subtraction. Optics Letters, 26, 11401142.
Shilemay, M., Rozban, D., Levanon, A., Yitzhaky, Y., Kopeika, N. S., Yadid-Pecht, O.,
et al. (2013). Performance quantification of a millimeter-wavelength imaging system
based on inexpensive glow-discharge-detector focal-plane array. Applied Optics, 52,
C43C49. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-52-7-C43.
Shoop, B. L., Pezeshki, B., Goodman, J. W., & Harris, J. S., Jr. (1992). Noninterferometric
optical subtraction using reflection-electroabsorption modulators. Optics Letters, 17,
5860.
Simova, E. S., & Stoev, K. N. (1992). Phase-stepping automatic fringe analysis in holographic
moire. Applied Optics, 31, 59655974. http://www.opticsinfobase.org/ao/abstract.cfm?
URIao-31-28-5965.
Sizikov, V. S., & E kzemplyarov, R. A. (2013). Operating sequence when noise is being
filtered on distorted images. Journal of Optical Technology, 80(1), 2834.
Skauli, T. (2013). Information capacity as a figure of merit for spectral imagers: The trade-off
between resolution and coregistration. Applied Optics, 52, C58C63. http://www.
opticsinfobase.org/ao/abstract.cfm?URIao-52-7-C58.
Slawecki, D., & Li, W. (1992). DCT/IDCT processor design for high data rate image coding.
IEEE Transactions on Circuits and Systems for Video Technology, 2, 135.
260 Ayman Alfalou and Christian Brosseau
Smith, R. A. (1993). In A. G. Haus, & M. J. Yaffe (Eds.), A categorical course in physics, technical
aspects of breast imaging. Oak Brook, IL: Radiological Society of North America.
Smith, W. E., & Barrett, H. H. (1983). Radon transform and bandwidth compression.
Optics Letters, 8, 395397. http://www.opticsinfobase.org/ol/abstract.cfm?URIol-8-
7-395.
Sogandares, F. M., & Fry, E. S. (1997). Absorption spectrum (340640 nm) of pure water.
I. Photothermal measurements. Applied Optics, 36, 86998709. http://www.
opticsinfobase.org/ao/abstract.cfm?URIao-36-33-8699.
Sorensen, H., Heideman, M., & Burrus, C. (1986). On computing the split-radix FFT. IEEE
Transactions on Acoustics, Speech, and Signal Processing, 34, 152.
Stern, A. (2006). Sampling of linear canonical transformed signals. Signal Processing, 86,
14211425.
Strang, G. (1999). The discrete cosine transform. SIAM Review, 41, 135.
Sungwook, Y., & Swartziander, E. (2001). DCT implementation with distributed arith-
metic. IEEE Transactions on Computers, 50, 985.
Tahara, T., Ito, K., Kakue, T., Fujii, M., Shimozato, Y., Awatsuji, Y., et al. (2010). Parallel
phase-shifting digital holographic microscopy. Biomedical Optics Express, 1, 610.
Testorf, M., & Lohmann, A. W. (2008). Holography in phase space. Applied Optics, 47,
A70A77. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-47-4-A70.
Thoma, R., Oesterhelt, D., Hampp, N., & Brauchle, C. (1991). Bacteriorhodopsin films as
spatial light modulators for nonlinear-optical filtering. Optics Letters, 16, 651653. http://
www.opticsinfobase.org/ol/abstract.cfm?URIol-16-9-651.
Tisse, C.-L., Guichard, F., & Cao, F. (2008). Does resolution really increase image quality?
Proceedings of SPIE, 6817, 68170Q.
Torres-Mendez, L., & Dudek, G. (2005). Color correction of underwater images for aquatic
robot inspection. In A. Rangarajan, B. Vemuri, & A. L. Yuille (Eds.), Energy minimization
methods in computer vision and pattern recognition (pp. 6073): Berlin/Heidelberg: Springer,
ISBN: 978-3-540-30287-2. http://link.springer.com/chapter/10.1007%2F11585978_5.
Treibitz, T., & Schechner, Y. Y. (2009). Active polarization descattering. IEEE Transactions
on Pattern Analysis and Machine Intelligence, 31, 385399.
Tu, B., Liu, L., Liu, Y., Jin, Y., & Tang, J. (2013). High precision two-step calibration
method for the fish-eye camera. Applied Optics, 52, C37C42.
Tyo, J. S., Goldstein, D. L., Chenault, D. B., & Shaw, J. A. (2006). Review of passive
imaging polarimetry for remote sensing applications. Applied Optics, 45, 54535469.
http://www.opticsinfobase.org/ao/abstract.cfm?URIao-45-22-5453, https://www.
novapublishers.com/catalog/product_info.php?products_id28370.
Vander Lugt, A. (1964). Signal detection by complex spatial filtering. IEEE Transactions on
Information Theory, IT-10, 139.
Vest, C. M. (1974). Formation of images from projections: Radon and Abel transforms. Jour-
nal of the Optical Society of America, 64, 12151218.
Voelz, D. G. (2011). Computational Fourier optics: A MATLAB tutorial, SPIE tutorial texts
TT89. Bellingham, WA: SPIE Press.
Voss, K. J., & Fry, E. S. (1984). Measurement of the Mueller matrix for ocean water. Applied
Optics, 23, 44274439. http://www.opticsinfobase.org/ao/abstract.cfm?URIao-23-
23-4427.
Walker, J. G., Chang, P. C. Y., & Hopcraft, K. I. (2000). Visibility depth improvement in
active polarization imaging in scattering media. Applied Optics, 39, 49334941. http://
www.opticsinfobase.org/ao/abstract.cfm?URIao-39-27-4933.
Wang, Q. (2012). Optical image encryption with silhouette removal based on interference
and phase blend processing. Optics Communications, 285, 4294.
Wang, Q., Cheng, D., Wang, Y., Hua, H., & Jin, G. (2013). Design, tolerance, and fabri-
cation of an optical see-through head-mounted display with free-form surface elements.
Applied Optics, 52, C88C99.
Recent Advances in Optical Image Processing 261
Wang, Q., Guo, Q., & Lei, L. (2013). Asymmetric multiple-image hiding using phase
retrieval technique based on amplitude- and phase-truncation in fractional Fourier
domain. Optik, 124, 38983902. http://dx.doi.org/10.1016/j.ijleo.2012.12.018.
Wang, Q., Liu, L., Liang, F., & Wang, Z. (1995). Optical cellular logic two-level networks
for morphological image processing. Optics Communications, 115, 251255. http://dx.
doi.org/10.1016/0030-4018(94)00700-5.
Weaver, C. S., & Goodman, J. W. (1966). A technique for optically convolving two func-
tions. Applied Optics, 5, 1248.
Wells, W. H. (1969). Loss of resolution in water as a result of multiple small-angle scattering.
Journal of the Optical Society of America, 59, 686691. http://www.opticsinfobase.org/josa/
abstract.cfm?URIjosa-59-6-686.
Westfeld, A., & Pfitzmann, A. (2000). Attacks on steganographic systems. In A. Pfitzmann
(Ed.), Information hiding Third international workshop, LNCS 1768 (p. 61). Berlin,
Heidelberg: Springer-Verlag.
Westfeld, A., & Wolf, G. (1998). Steganography in a video conferencing system.
In D. Aucsmith (Ed.), Information hiding LNCS 1525. Berlin, Heidelberg: Springer-
Verlag.
White, S. (1989). Applications of distributed arithmetic to digital signal processing: A tutorial
review. IEEE ASSP Magazine, 6, 4.
Wiener, N. (1949). Extrapolation, interpolation, and smoothing of stationary time series.
New York: Wiley, ISBN: 0-262-73005-7,
Wolf, E. (2007). Introduction to the theory of coherence and polarization of light. Cambridge:
Cambridge University Press.
Xia, P., Shimozato, Y., Tahara, T., Kakue, T., Awatsuji, Y., Nishio, K., et al. (2013a).
Image reconstruction algorithm for recovering high-frequency information in parallel
phase-shifting digital holography [Invited]. Applied Optics, 52, A210A215. http://
www.opticsinfobase.org/ao/abstract.cfm?URIao-52-1-A210.
Xuan, N. P., Ferrier, J. L., Gazengel, J., Rivoire, G., Breki-Iovskhikh, G. L.,
Kudriavtseva, A. D., et al. (1988). Changes in the space structures of light beams induced
by nonlinear optical phenomena: Application to phase contrast and image processing.
Optics Communications, 68, 244250. http://www.sciencedirect.com/science/article/
pii/0030401888903938.
Yamaguchi, I., & Zhang, T. (1997). Phase-shifting digital holography. Optics Letters, 22,
12681270. http://www.opticsinfobase.org/ol/abstract.cfm?URIol-22-16-1268.
Yaroslavsky, L. P. (2012). Theoretical foundations of digital imaging using MATLAB. Boca
Raton, FL: CRC Press Taylor & Francis Group.
Yelleswarapu, C. S., Kothapalli, S.-R., & Rao, D. V. G. L. N. (2008). Optical Fourier tech-
niques for medical image processing and phase contrast imaging. Optics Communications,
281, 18761888.
Yelleswarapu, C. S., Kothapalli, S. R., Vaillancourt, Y., Aranda, F. J., Kimball, B., &
Rao, D. V. G. L. N. (2006). Phase contrast imaging using photo-thermally induced phase
transitions in liquid crystals. Applied Physics Letters, 89, 2111161. http://dx.doi.org/
10.1063/1.4737942.
Yelleswarapu, C., Wu, P., Kothapalli, S. R., Rao, D. V. G. L. N., Kimball, B.,
Sai, S. S., et al. (2006). All-optical spatial filtering with power limiting materials.
Optics Express, 14, 14511457. http://www.opticsinfobase.org/oe/abstract.cfm?
URIoe-14-4-1451.
Yokozeki, S., & Ohnishi, K. (1975). Spherical aberration measurement with shearing inter-
ferometer using Fourier imaging and moire method. Applied Optics, 14, 623627. http://
www.opticsinfobase.org/ao/abstract.cfm?URIao-14-3-623.
Yong-Liang, X., Xin, Z., Sheng, Y., & Yao-Yao, C. (2010). Multiple-image parallel optical
encryption. Optics Communication, 283, 27892793. http://dx.doi.org/10.1016/
j.optcom.2010.03.003.
262 Ayman Alfalou and Christian Brosseau
You, Y., Tonizzo, A., Gilerson, A. A., Cummings, M. E., Brady, P., Sullivan, J. M., et al.
(2011). Measurements and simulations of polarization states of underwater light in clear
oceanic waters. Applied Optics, 50, 48734893. http://www.opticsinfobase.org/ao/
abstract.cfm?URIao-50-24-4873.
Yu, F. T. S., & Jutamulia, S. (1998). Optical pattern recognition. Cambridge (UK): Cambridge
University Press.
Yu, F. T. S., & Tai, A. (1979). Incoherent image addition and subtraction: A technique.
Applied Optics, 18, 27052707.
Yuan, S., Jin, G., Wu, M., & Yan, Y. (1996). Neighborhood operation binary image algebra
for optical morphological image processing. Optics Communication, 123, 705715. http://
dx.doi.org/10.1016/0030-4018(95)00317-7.
Zalevsky, Z. (2010). Extended depth of focus imaging: A review. SPIE Reviews, 1, 018001.
Zalevsky, Z., Gaffling, S., Hutter, J., Chen, L., Iff, W., Tobisch, A., et al. (2013). Passive
time-multiplexing super-resolved technique for axially moving targets. Applied Optics,
52, C11C15.
Zalevsky, Z., & Mendlovic, D. (1996). Fractional Wiener filter. Applied Optics, 35, 3930.
http://www.opticsinfobase.org/ao/abstract.cfm?URIao-35-20-3930.
Zalevsky, Z., Shemer, A., Zlotnik, A., Ben Eliezer, E., & Marom, E. (2006). All-optical axial
super resolving imaging using a low-frequency binary-phase mask. Optics Express, 14,
26312643.
Zhang, Y., Pedrini, G., Osten, W., & Tiziani, H. J. (2004). Reconstruction of in-line digital
holograms from two intensity measurements. Optics Letters, 29, 17871789.
Zhao, L., Healy, J. J., & Sheridan, J. T. (2013). Unitary discrete linear canonical transform:
Analysis and application. Applied Optics, 52, C30C36. http://www.opticsinfobase.org/
ao/abstract.cfm?URIao-52-7-C30.
CHAPTER THREE
Contents
1. Loudon's Energy Velocity 264
2. Analysis of the Phase Function and Its Saddle Points 267
2.1 Numerical Investigation 268
2.2 Location of the Saddle Points and Approximation of the Phase Function 274
3. Uniform Asymptotic Expansions 288
3.1 Procedure for the Asymptotic Analysis of the Propagated Signal 289
3.2 The Sommerfeld Precursor Field 295
3.3 The Brillouin Precursor 300
3.4 The Signal Contribution 308
3.5 The Total Field 312
3.6 The Signal Arrival and the Signal Velocity 312
4. Extensions of the SommerfeldBrillouin Theory to the Singular and Weak
Dispersive Limits 319
4.1 Dispersive Signal Propagation in the Singular Dispersion Limit 320
4.2 Dispersive Signal Propagation in the Weak Dispersion Limit 323
5. Pulsed Plane-Wave Transmission Through an Interface 327
5.1 Formulation 329
5.2 Asymptotic Description 332
5.3 Transmitted Signal Evolution 338
6. Concluding Remarks 341
Acknowledgments 343
References 343
1
Rodney Loudon received his PhD in physics from the University of Oxford in 1958, a Postdoctoral
Fellow at the University of California at Berkeley from 1959 to 1960, and finally professor at the
University of Essex. He is a Fellow of the Royal Society and was awarded the Max Born Award
in 1992.
Part II. The Modern Asymptotic Theory 265
_ 2 U_ tot :
r S + 20 mNjrj (8)
Integration of this expression over a region R of space bounded by the sur-
face S with outward normal n^ followed by application of the divergence
theorem then yields
I Z Z
2 3 d
S nd
^ r + 20 mN jrj
2
_ d r U tot d3 r, (9)
S R dt R
b2 2 20
n2r n2i 1 , (12)
2 20 2 + 420 2
b2 0
nr ni (13)
2 20 2 + 420 2
relating the real and imaginary parts of the complex index of refraction
n() nr() + ini() for a Lorentz model dielectric have been used in deriv-
ing these expressions. The time-average energy velocity in a single-resonance
Lorentz model dielectric is then given by the ratio of these two time-averaged
quantities as
S c
vE ,
U tot ni (14)
nr +
0
2.5
Relative wave velocity, v/c
1.5
vp /c
1 vg /c
0.5
vE /c
0
0 0.2 w0 0.6 0.8 1 1.2 1.4 1.6 1.8 2
w r/s 1017
Figure 1 Comparison of the frequency dependence of the relative phase vp()/c
(dashed curve), relative group vg()/c (dotted curve), and relative energy vE()/c (solid
curve) velocities in a single-resonance Lorentz model dielectric with medium parame-
p
ters 0 4 1016 r/s, b 20 1016 r=s, and 0 0.28 1016 r/s.
Although the energy velocity has been derived only for the special case of
a time-harmonic wave field, it does find direct application to the more gen-
eral case of an electromagnetic pulse. In particular, Oughstun and Sherman
(1980) and Sherman and Oughstun (1981, 1995) have derived a physical
description of the precursor fields in a Lorentz model dielectric that is based
solely upon Loudons energy velocity and the attenuation of the dispersive
medium. This physical description then represents a generalization of the
group velocity description to dispersive attenuative media.
provide uniform asymptotic expansions for the two forerunners and the sig-
nal that remain valid for all 1. Evaluation of the integral representation of
the propagated pulse [Eq. (1.58) in Part I] remained at this point until early
1974. At that time, a complete reinvestigation of the SommerfeldBrillouin
theory was initiated by Oughstun2 and Sherman3 . Together, they found
more accurate expressions for the saddle point locations and applied newly
developed uniform asymptotic methods that enabled them to provide uni-
form asymptotic expansions for each of the three components of the prop-
agated field. This research relied upon the then recently published theorem
due to Olver (1970) which proved that the asymptotic expansion provided
by the method of steepest descents was independent of the steepest descent
path provided (among other things) that the path through the relevant saddle
point was inside the region with greater exponential decay than that at that
saddle point. This relaxation from the earlier steepest descent requirement,
referred to here as the saddle point method, had far-reaching implications
with regard to the signal velocity. In this section, we present the analysis
of the phase function and its saddle points as determined by Oughstun
and Sherman. Application of these results to well-defined uniform asymp-
totic methods is presented in the following section.
5
SPn+
(1.2 1016)
w0 5
0
w+ w + 10 15 w
(0) (1016 r/s)
SPn
5 (+1.2 1016)
Figure 2 Isotimic contours of the real part (, ) Re{(, )} of the complex phase
function in the right-half of the complex -plane at the fixed spacetime point 1.
Notice that the distant saddle points SP
d are symmetrically located at 1 20 i at this
luminal spacetime point. Reprinted with kind permission of Springer Science + Business
Media from Oughstun (2009), Fig. 12.4, p. 275.
q = 1.25
w
(1016 r/s)
(+1.0 1016)
5
SPn+
(1.0 1016)
w0 5 (0)
0
w+
10 15 (1016 r/s)
w+ SPd+
(+1.0 1016)
SPn
5
(1.0 1016)
Figure 3 Isotimic contours of the real part (, ) Re{(, )} of the complex phase
function in the right-half of the complex -plane at the fixed spacetime point 1.25.
Reprinted with kind permission of Springer Science + Business Media from Oughstun (2009),
Fig. 12.5, p. 275.
q = q SB 1.33425
w
(1016 r/s)
5
(+1.0 1016)
SPn (+1.0 1016)
(1.0 1016)
5
Figure 4 Isotimic contours of the real part (, ) Re{(, )} of the complex phase
function in the right-half of the complex -plane at the fixed spacetime point
SB 1.33425 when the upper near and distant saddle points are of equal
dominance, i.e., (+SPn, SB) (
SPd, SB). Reprinted with kind permission of Springer
Science + Business Media from Oughstun (2009), Fig. 12.6, p. 276.
w q q1 1.501
(1016 r/s)
5
(+1.0 1016)
(1.0 1016)
w0 5 (0)
0 SPn+ w
SPn
w+ 10 15 (1016 r/s)
(0) w+ SPd+
(+1.0 1016)
(1.0 1016)
Figure 5 Isotimic contours of the real part (, ) Re{(, )} of the complex phase
function in the right-half of the complex -plane at the fixed spacetime point 1.501
just prior to the coalescence of the two near saddle points SP n into a single second-
order saddle point when 1. Reprinted with kind permission of Springer Science
+ Business Media from Oughstun (2009), Fig. 12.7, p. 276.
q = 1.65
w
(1016 r/s)
5
(+1.0 1016)
(1.0 1016)
w0 5 (0)
w
0
SPn+ w+ 10 15 (1016 r/s)
w+ SPd+
(+1.0 1016)
(1.0 1016)
Figure 6 Isotimic contours of the real part (, ) Re{(, )} of the complex phase
function in the right-half of the complex -plane at the fixed spacetime point 1.65.
Reprinted with kind permission of Springer Science+Business Media from Oughstun (2009),
Fig. 12.8, p. 277.
q = 5.0
w
(1016 r/s)
(2.0 1016)
(3.4 1016) (0 )
0 w
w0 5 (1016 r/s)
w+
SPn+ w+ SPd+
(2.0 1016)
1 (3.4 1016)
Figure 7 Isotimic contours of the real part (, ) Re{(, )} of the complex phase
function in the right-half of the complex -plane at the fixed spacetime point 5.0.
Reprinted with kind permission of Springer Science+Business Media from Oughstun (2009),
Fig. 12.9, p. 278.
272 Natalie A. Cartwright and Kurt E. Oughstun
point with negative (, ) with the smallest absolute value] are indicated by
solid curves, whereas those contours which are less than the value at the
dominant saddle point are indicated by dashed curves. In addition, the
region of the complex -plane where (, ) is less than that at the dominant
saddle point has been shaded in light gray and the region where (, ) is less
than that at the nondominant saddle point, which is contained inside the
light gray-shaded region, is shaded in dark gray. Finally, the (, ) values
of the maximum and minimum isotimic contours, as well as the zero contour
(, ) 0, are indicated in each figure by the number contained in paren-
thesis that is either adjacent to or on top of the appropriate contour.
As predicted by Brillouin, for all values of the spacetime parameter 1
except one, there are four first-order saddle points of the complex phase
function (, ) for a single-resonance Lorentz model dielectric, symmetri-
cally located about the imaginary axis. Two of these saddle points, called near
saddle points, reside in the region jj 0 about the origin of the complex
-plane and the other two, called distant saddle points, reside in the region
jj 1 of the complex -plane that is removed from the origin.
The two near first-order saddle points SP n are seen in Figures 25 to lie
along the imaginary axis over the initial spacetime domain 2 [1, 1],
approaching each other as increases to 1, the upper near saddle point
p
SP+n crossing the origin at 0, where 0 1 + b2 =20 1:5 for
Brillouins choice of the medium parameters, and then coalescing into a sin-
gle second-order saddle point at 1, where the numerical value of this
critical spacetime point for Brillouins choice of the medium parameters is
just slightly larger than the spacetime value 1.501 illustrated in Figure 5.
As increases above 1, the two near first-order saddle points separate such
that they are symmetrically situated about the imaginary axis, approaching
the inner branch points , respectively, as ! 1.
The two distant saddle points SP d , on the other hand, are located in the
lower-half of the complex -plane for all 1 and are located at 1 i0
at the luminal spacetime point 1. As increases from unity, these two
saddle points symmetrically move in from infinity and approach the respec-
tive outer branch points 0 as increases to infinity, as evident in
Figures 27.
Initially, the distant saddle points SP d have less exponential decay asso-
ciated with them than does the upper near saddle point SP+n , as seen in
Figures 2 and 3, so that there exists some spacetime point SB such that
SPd , > SPn , when 1 < SB :
+
(15)
Part II. The Modern Asymptotic Theory 273
SPd , SB SPn , SB when SB :
+
(16)
Consequently, at SB, those three saddle points (SP+d , SP +
d , and SPn ) are
of equal importance (or dominance) in the asymptotic description of the
propagated wave field.
The remaining Figures 57 show that for values of 2 (SB, 1), the
upper near saddle point SP+n is dominant over the two distant saddle points
SPd , so that
SP
+
n
, >
SPd , when SB < < 1 : (17)
At 1, the two near first-order saddle points SP
n have coalesced into a
single second-order saddle point SPn which is dominant over the distant
saddle point pair, so that
SPn , 1 >
SPd , 1 when 1 : (18)
Finally, for all > 1, the near saddle points SP
n are dominant over the
distant saddle points SPd , so that
SPn , > SPd , when > 1 , (19)
as evident in Figures 67. Notice the change in scale of the real and imag-
inary coordinate axes in Figure 7, demonstrating how the topography of
0
(, ) becomes increasingly concentrated about the branch cuts
0
and ++ as increases above the critical spacetime value 1 with the near
saddle points SPn approaching the lower branch points , respectively, and
the distant saddle points SPd approaching the upper branch points .
0
q
! 20 20 i0 , as ! 1, (24)
Thus, the saddle points move into the branch points + and +0 in the right-
0
half plane and and in the left-half plane in the limit as ! 1.
An exact polynomial equation describing the location of the saddle
points can be obtained from Equation (23) as
2
1 8 + 8i0 2 1 7 4 20 + 620 2 1 + b2 2 6
2i0 1220 + 3b2 + 1620 2 1 5
+ 640 + 4820 + 2b2 20 + 12b2 20 + 1640 2 1
+ b2 20 2 1220
4
+ 4i0 640 + 4b2 20 + 820 20 + 420 b2 2 1
(26)
+ 20 + 220 220 2 b2
3
20 440 + 3b2 20 + 2420 20 + 1220 b2 2 1
b2 40 + 2020 20 + 920 b2
2
2i0 20 420 + 3b2 20 2 1 b2
+ 40 20 + b2 20 2 1 b2 0:
Because this eighth-order polynomial is extremely formidable, approximate
solutions of the saddle point Equation (23) were then developed for both the
distant and near saddle points that are a characteristic of a single-resonance
Lorentz model dielectric.
This particular form of the saddle point equation explicitly displays the
desired limiting behavior as ! 1 that is given in Equation (25). With this
limiting behavior in mind, the rational function appearing in the squared
term of Equation (27) may be approximated as
b2 + i0 0 b2
b 2
i
2 20 + 2i0
2 2 0 b2 2 2 b4
20 + 2i0 2i 20 + 2i0 0 2
2 2
0 + 2i0 1 + 2i0 :
2 2 2
SPd i0 1 + (29)
On the other hand, for sufficiently large values of , Equation (30a) and
q
(30b) have the limiting behaviors ! 21 20 and ! 0, so that
in the limit as approaches infinity
q
lim SPd 21 20 i0 0 , (32)
!1
and the distant saddle points SP d , respectively, approach the outer branch
0
points . This approximation to the distant saddle point locations (due
to Oughstun, 1978) then captures the exact limiting behavior in the two
opposite extremes at either 1 or 1. A sketch of the respective paths
followed by these two distant saddle points in the complex -plane is
presented in Figure 8.
The approximate local behavior of the phase function (, ) about
the distant saddle point SP+d in the right-half of the complex -plane, where
SP+d () + rei, is given by Oughstun and Sherman (1988) as
w branch w d w+ branch w+
cut cut
+
SPd SPd
2d
Figure 8 Behavior of the distant saddle points SP d in the complex -plane for a single-
resonance Lorentz model dielectric. The dotted curves indicate the respective directed
paths that these saddle points follow as increases to infinity. The dashed lines through
each saddle point indicate the local behavior of the isotimic contour (, ) (SPd , )
through that saddle point, the shaded region indicates the local region about each sad-
dle point where the inequality (, ) < (SPd , ) is satisfied, and the arrows indicate
local direction of ascent along the lines of steepest descent and ascent through each
saddle point. Reprinted with kind permission of Springer Science+Business Media from
Oughstun (2009), Fig. 12.39, p. 325.
278 Natalie A. Cartwright and Kurt E. Oughstun
r,, 1 i + 0 1 + + irei
b2 i0 1 + rei
i ,
2 2 + 20 1 + 2r cos + 20 1 r sin + r 2
(33)
with real part
r,, 1 0 1 + r sin
b2 0 1 + r sin
:
2 2 + 20 1 + 2r cos + 20 1 r sin + r 2
(34)
It is then seen that (r, , ) attains its maximum variation with r about the
distant saddle point SP+d when /4, 3/4, 5/4, 7/4. Consequently,
in the right-half plane, the lines of steepest descent through the distant sad-
dle point SP+d are at the angles 3/4 and 7/4, and the lines of
steepest ascent are at /4 and 5/4. Because of the even symmetry
of (, ) about the imaginary axis, the reverse holds true for the distant
saddle point SP d in the left-half plane. This local behavior about the distant
saddle points SP d is depicted in Figure 8, where the vectors indicate the
direction of ascent along the lines of steepest descent and ascent through
each respective distant saddle point, which are at angles of 45 to the coor-
dinate axes, and the shaded areas indicate the local regions about each saddle
point wherein the inequality (, ) < (SPd , ) is satisfied and in which
the path of steepest descent from the respective saddle point lies. These
results are in complete agreement with the numerical results presented in
Figures 27.
2 + i0 2 2
2
2 21 + 2i0 2 20 + 2i0 21 40
23
20 2i0 + i 2 0 2 221 + 20
1 0
are used in Equation (35), which then assumes the approximate polynomial
form
2 0 b2 b2 20 20 b2 2 3
0 + 2i0 2i 4 3 + 2 4 2 4 2 21 + 0
2 2 2
0 1 0 1 0
b2
2
2 b2
+ 1 2 2 4 02 20 2 2
0 0 1 0
b2
+ 2i0 1 2 20 b2 :
0
Because the coefficient of the cubic term in is small in comparison to the
other terms appearing in this expression, it may be neglected. The approx-
imate saddle point equation for the near saddle point locations then becomes
(Oughstun, 1978)
b2
2 20 + 2
20 20 2 20
+ 2i0
2
0, (36)
b2 b2
0 + 3 2
2 2
0 + 3 2
2 2
0 0
where
s
b2
0 n0 1 + 2 , (37)
0
and
20 2 2
1 41 + b 1: (38)
320 21
280 Natalie A. Cartwright and Kurt E. Oughstun
0 0
2
b
2 20 + 2
3 20
: (40b)
2 b2
0 + 3 2
2 2
0
In the limit as approaches infinity
q
lim SPn 20 20 i0 (41)
!1
and the near saddle points SP n , respectively, approach the inner branch
points , in agreement with the numerical results presented in
Figures 27. The critical spacetime point 1 at which the two near
first-order saddle points coalesce into a single second-order saddle point is
found from this expression to be given by
220 b2
1 0 + , (42)
20 320 420
which reduces to the approximate expression given by Brillouins analysis
through neglect of the term 420 in comparison to 320 in the denominator.
With Brillouins choice of the material parameters, this critical value is
numerically found (Oughstun, 1978) to lie in the range 1.50275 < 1 <
1.50300 while Equation (42) gives the value 1 1.50414.
The dynamical behavior of the near saddle points and the local complex
phase behavior about them, as described by Oughstuns approximation (39)
with
Part II. The Modern Asymptotic Theory 281
b2
, i 0 + 2 i 20 (43)
20 40
is now considered for the three separate cases 1 < 1, 1, and > 1.
Case 1 (1 < 1): Over this initial spacetime domain, the near saddle
point locations are given by
2
SPn i o 0 (44)
3
with the approximate expressions
2 0 2 12 31=2
2 20 + 2 b 2
2 2
2
o 420 @ 2 0 A
20 2 0 5
, (45)
20 + 3 b 2 0 + 3 b 2
2 2
0 0
2
b
2 20 + 2
3 20
, (46)
2 2 b2
0 + 3 2
2
0
that are appropriate over this domain. As depicted in Figure 9, the near sad-
dle points SP n are located along the imaginary axis, symmetrically situated
about the point 00 23 0 , where () varies slowly over this space
time domain.
The approximate local behavior of (, ) Re{(, )} about each
near saddle point SP 00
n , where i () + re
i
with 00
23 0 , is given by (Oughstun & Sherman, 1988)
r,, 00 + r sin 0
b2
00 3 00 2 00 2 00
+ + 20 + 3 + 40 r sin (47)
20 40
300 + 20 r 2 cos 2 r 3 sin3
,
from which it is seen that (r, , ) attains its maximum variation about each
near saddle point when 0, /2, , 3/2. The lines of steepest descent and
ascent through the near saddle points are then parallel to the coordinate axes,
as depicted in Figure 9, where the vectors indicate the direction of ascent
along these lines. The dashed lines through each saddle point indicate the
local behavior of the isotimic contour (, ) (SPn , ) through that
282 Natalie A. Cartwright and Kurt E. Oughstun
+
SPn
w
d
SPn
Figure 9 Phase behavior about the near saddle points SP n situated along the imaginary
axis of the complex -plane for a single-resonance Lorentz model dielectric over the
initial spacetime domain 1 < 1. The dashed lines through each saddle point indi-
cate the local behavior of the isotimic contour (, ) (SPn, ) through that saddle
point, the shaded region indicates the local region about each saddle point where the
inequality (, ) < (SPn , ) is satisfied, and the arrows indicate local direction of
ascent along the lines of steepest descent and ascent through each saddle point.
Reprinted with kind permission of Springer Science+Business Media from Oughstun
(2009), Fig. 12.40, p. 332.
saddle point, and the shaded region indicates the local region about each
saddle point where the inequality (, ) < (SPn , ) is satisfied. Notice
that the paths of steepest descent through the upper near saddle point
SP+n are at the angles 0, , whereas the paths of steepest descent through
the lower near saddle point SP n are at the angles /2, 2/2, as indicated
in the figure.
Notice that at the critical spacetime point 0 (where 1 < 0 < 1),
the near saddle points are located at SP+n (0) 0 and SPn 0 i 4 3 ,
0
where the solution for the upper near saddle point at this value is exact.
Furthermore
exactly. It is this latter property that makes this saddle point so important in
the subsequent asymptotic field behavior.
Part II. The Modern Asymptotic Theory 283
and the two first-order near saddle points have coalesced into a single
second-order saddle point.
From Equation (4.34) with 00 20/(3), the local behavior of
(, 1) about this second-order near saddle point is found to be given by
3
b2 40 3
r, ,1 + r sin 3 : (51)
0 40 272 2
Figure 10 Phase behavior about the near saddle point SPn in the complex -plane for a
single-resonance Lorentz model dielectric at the critical spacetime point 1 when
the two first-order saddle points SP
n have coalesced into a single second-order saddle
point. The dashed lines indicate the local behavior of the isotimic contour (, )
(SPn, ) through this second-order saddle point, the shaded area indicates the local
region about the saddle point where the inequality (, ) < (SPn, ) is satisfied,
and the arrows indicate local direction of ascent along the lines of steepest descent
and ascent through the saddle point. Reprinted with kind permission of Springer
Science + Business Media from Oughstun (2009), Fig. 12.41, p. 334.
284 Natalie A. Cartwright and Kurt E. Oughstun
Case 3 ( > 1): Over this final spacetime domain, the near saddle point
locations are given by
2
SPn i0 , (52)
3
where () and () are both real-valued and are given by Equation (40a).
As increases away from 1, the second-order saddle point SPn separates into
two near first-order saddle points SPn that move off of the imaginary axis
into the lower-half of the complex -plane, symmetrically situated about
the imaginary axis, as depicted in Figure 11. In the limit as ! 1, these
two near saddle points SP n approach the inner branch points ,
respectively.
_
2d
3a
SPn SPn+
w
branch
cut w
d w+
branch
cut w+
Figure 11 Behavior of the near saddle points SP n in the complex -plane for a single-
resonance Lorentz model dielectric over the final spacetime domain > 1. The dotted
curves indicate the respective directed paths that these first-order saddle points follow
as increases to infinity. The dashed lines through each saddle point indicate the local
behavior of the isotimic contour (, ) (SPn , ) through that saddle point, the
shaded region indicates the local region about each saddle point where the inequality
(, ) < (SPn, ) is satisfied, and the arrows indicate local direction of ascent along
the lines of steepest descent and ascent through each saddle point. Reprinted with kind
permission of Springer Science+Business Media from Oughstun (2009), Fig. 12.42, p. 335.
Part II. The Modern Asymptotic Theory 285
It is then seen that (r, , ) attains its maximum variation about the near
saddle point SP+n when /4,3 /4, 5/4, 7/4. The lines of steepest
descent through this saddle point are at /4, 7/4 and the lines of
steepest ascent are at 3/4, 7/4. Because of the even symmetry of
(,) about the imaginary axis, the lines of steepest ascent and descent
through the near saddle point SP n are reversed, as illustrated in Figure 11.
b2 0 1
SPd , 0 1 + 1 , (54)
2 + 20 1 2
2
and
286 Natalie A. Cartwright and Kurt E. Oughstun
" #
b2 =2
SPd ,
1 + 2 , (55)
+ 20 1 2
respectively, for all 1. It is then seen that at the luminal spacetime point
1, (SPd , 1) 0 (exactly), and that as increases away from unity,
(SPd , ) is negative with monotonically increasing magnitude, where
lim !1 SPd , 1 as the distant saddle points SP d approach the
0
outer branch points , respectively.
The real part of the complex phase behavior at the near saddle points is
described by the set of approximate expressions
2
2 b2 2
SPn , o 0 0 + o 0
3 20 40 3
2
o 0 + 20 ; 1 1 ,
3
(56)
3 2
40 b
SPn ,1 ; 1 , (57)
272 0 40
2 b2
SPn , 0 0 +
3 20 40
8 3 2 h i
1 + 2 1
; 1 :
2
9 0 3
(58)
The imaginary part of the complex phase behavior at the near saddle
points SPn identically vanishes over the spacetime domain 1 1,
and for 1 is given by
b2 4 2
SPn ,
0 2 + :
2
20 40 3
(59)
These expressions then show that at the upper near saddle point SP+n ,
(SP+n , ) is initially negative over the spacetime domain 2 [1, 0),
increasing to zero as increases to 0, identically vanishes at 0, and then
grows negative monotonically over the short spacetime interval 2
(0, 1]. At the lower near saddle point SP n , (SPn, ) is initially positive
and monotonically decreases to the approximate value 43 b2 =272 0 40
Part II. The Modern Asymptotic Theory 287
1 q SB q0
0 q
-
q1 (wSP+- ,q)
n
Figure 12 Sketch of the behavior of the real part (, ) of the complex phase function
(, ) at the relevant saddle points for a single-resonance Lorentz model dielectric.
Reprinted with kind permission of Springer Science+Business Media from Oughstun
(2009), Fig. 12.43, p. 339.
288 Natalie A. Cartwright and Kurt E. Oughstun
2
provided that 20 b2 =40 1. With Brillouins choice of the material
parameters, this critical value is numerically found (Oughstun, 1978) to
be SB 1.334 while Equation (61) estimates the value as SB 1.255.
A related quantity of interest is the value of the real angular frequency
q
SB 21 20 that is defined by the relation
SPd+ , SB SB , (62)
where (0 ) 0 ni(0 ) along the real axis. According to this definition,
the angular frequency SB is the real coordinate value at which the isotimic
contour (, ) (SP+d , ) through the distant saddle point SP+d crosses the
positive 0 -axis when SB. Unfortunately, finding the solution of
Equation (62) for SB is an extremely formidable, if not impossible, task.
However, because the isotimic contour through SP+d at the angle /2 to
the 0 -axis remains at essentially this angle when it intersects the 0 -axis,
as seen in Figure 4, an approximate expression for SB is given by the real
coordinate value of the distant saddle point SP+d at SB, so that
(Oughstun & Sherman, 1988)
s
b2 52
SB SB 0 2 + 2 + 02 : (63)
0 30
p
Notice that SB 20 + b2 =40 when 0 > b 0, and that
SB b + 20 =b when b > 0 0. With Brillouins choice of the material
parameters, this critical frequency value is numerically found (Oughstun,
1978) to be SB 8.70 1016 r/s while the final approximate expression
above gives SB 7.22 1016 r/s whereas SB (SB) 8.41 1016 r/s.
Because the distant saddle points SP d are dominant for some values of
and either the upper near saddle point SP+n or both of the near saddle points
SPn are dominant for other values of , there is no single path P() that is an
Olver-type path (see Oughstun, 1978, 2009) with respect to a single saddle
point and that evolves continuously with for all 1. In order to obtain an
asymptotic representation that remains uniformly valid for all 1, the con-
tour P() must evolve continuously for all 1 and, in the vicinity of the
spacetime point SB when the saddle point dominance changes, the
path must pass through both the dominant and nondominant saddle points
involved in the dominance change. Moreover, the path P() must be divis-
ible into a sum of subpaths Pj(), each of which is an Olver-type path with
respect to one of the saddle points.
For spacetime values in the range 2 [1, 1) during which the two near
saddle points SP 00
n are situated along the imaginary -axis, the lower near
saddle point SPn is dominant over all of the other saddle points (see
Figures 24). That saddle point is not useful, however, because the
Olver-type paths with respect to it are not deformable to the original con-
tour C (and vice versa) due to the presence of the branch cuts and is hence-
forth ignored.
There are many paths having the required properties that pass through both
the upper near saddle point SP+n and the distant saddle points SP d for space
time points in the domain 2 [1,1). There are also many paths having the
required properties that pass through both of the near saddle points SP n and
the distant saddle points SPd for > 1. Finally, there are many paths having
the required properties that pass through the single second-order near saddle
point SPn and both of the distant saddle points SP d when 1. As a conse-
quence, the contour P() can always be chosen so that it passes through the
upper near saddle point SP+n and the distant saddle points SP d for 2 [1, 1]
and through all four saddle points for > 1 so that it evolves in a continuous
manner as varies over the entire subluminal spacetime domain 1 and can
be divided into the desired subpaths with respect to each relevant saddle point.
An example of such a path P() and its component subpaths Pj() is illus-
trated in Figure 13. For values of in the range 2 [1,1], the component
subpaths (from left to right in the figure) are P + +
d (), Pn (), and Pd (), and for
> 1 the component subpaths are Pd (), Pn (), Pn (), and Pd (). The sub-
+ +
paths P
d () and Pn () are Olver-type paths with respect to the saddle points
SPd and SPn , respectively.
Provided that the path P() and its component subpaths Pj() satisfy the
above constraints, it is unimportant which particular paths are used; by
Part II. The Modern Asymptotic Theory 291
w''
C
+
SPn
Pd Pn+
w
w w w+ w+
SPn
SP SPd+
d Pd+
q < qSB
1<
w
Original contour of integration C
SPn+
w
Pd Pn+
w w w+ w+
SPn
SP d
SPd+
Pd+
w
Pd SPn SPn+ Pn+
w w w+ w+
Pn
SPd SPd+
Pd+
q1 < q
where I
d (z, ) and In (z, ) denote the contour integrals taken over the
Olver-type paths P
d and Pn , respectively. In order to obtain an asymptotic
approximation of the propagated signal AH(z,t) in a Lorentz model dielectric,
it now only remains to obtain asymptotic approximations of the various con-
tour integrals appearing on the right-hand sides of Equation (67a) and (67b).
If the distant saddle points SP
d do not pass too near to the pole at c,
then an asymptotic approximation of the quantity I d (z, ) + Id (z, ) in the
+
is obtained, where As(z, t) is the contribution due to the two distant saddle
points alone, referred to as the first, or Sommerfeld, precursor, and Rd is the
remainder.
If the near saddle points SP+n for 1 < 1 and SP n for > 1 do not
pass too close to the pole singularity at c > 0, then asymptotic approxima-
tions of I+n (z, ) for 1 < 1 and In (z,) + In (z,) for > 1 in the form
+
are obtained, where Ab(z,t) is the contribution due to the first upper near
saddle point and then both near saddle points, referred to as the second,
or Brillouin, precursor, and Rn is the remainder.
Part II. The Modern Asymptotic Theory 293
Consider now the above absorption band situation when the distant sad-
q
dle point SP+d approaches (as varies) the pole at c > 21 20 that is
bounded away from + 1. An asymptotic approximation of the quantity
Id(z,) + Id+(z,) in the form
Id z, + Id+ z, As z, t + Cd+ z,t + Rd z, (70)
is then obtained, where As(z, t) is given by the same expression as is As(z, t) in
Equation (68) for all 1. The quantity C+d (z, t) appearing in Equation (70)
is the contribution due to the proximity of the saddle point to the pole,
which is asymptotically negligible if the distant saddle points SP+d do not
approach the pole at c so that Equation (70) reduces to Equation (68).
Similarly, in the below absorption band situation when near saddle
point SP+n approaches (as varies) the simple pole at c with
q
0 < c < 20 20 , asymptotic approximations of the form I+n (z, ) for
1 < < 1 and I
n (z, ) + In (z, ) for > 1 in the form
+
Lorentz model dielectric. This approximation is uniformly valid for all sub-
luminal spacetime points 1, provided that the simple pole singularity
at c is bounded away from the limiting locations taken by the distant
saddle points SP d as ! 1 and by SPn as ! 1 from below. The con-
+ +
for 1 1, and
103
8
2
As (z,t)
8
1 1.05 1.1 1.15 1.2 1.25 1.3
q = ct/z
Figure 14 Dynamic behavior of the Sommerfeld precursor As(z, t) for the step- function-
modulated sine wave with below-resonance-applied signal frequency c 0.50 rad/s
at a relative observation distance z/zd 3.8 into the single-resonance Lorentz dielectric
with Brillouin's choice of material parameters as described by the uniform asymptotic
theory with numerically determined saddle point locations.
298 Natalie A. Cartwright and Kurt E. Oughstun
and
b2
00 SPd , i : (80)
+ i0 1
3
Substitution of Equations (79) and (80) into (78a)(78d) gives the approx-
imate expression for the Sommerfeld precursor
" #1=2
b2 =2
As z,t 1+ 2
2b + 20 1 2
( " #)
z b2 =21
exp 0 1 + 1 + 2
c + 20 1 2
( " #
0 5 + 3c 1 5 3c 1
2 + c 2 + 20 1 + 2 c 2 + 20 1 + 2
" !#
z b2 =2
J0 1 + 2
c + 20 1 2
2 3
3 3
c 20 1 2 + c 20 1 2
6 2 2 7
+4 5
c + 0 1 +
2 2 2
+ c + 0 1 +
2 2 2
" !#)
z b2 =2
J1 1 + 2
c + 20 1 2
(81)
as z ! 1 for all 1.
Part II. The Modern Asymptotic Theory 299
For values of > 1 bounded away from unity, the two Bessel functions
appearing in (81) may be replaced by their large argument asymptotic
approximation
r
2
J cos (82)
2 4
as ! 1 with jarg()j < . Substitution of (82) into (81) results in the non-
uniform asymptotic approximation of the Sommerfeld precursor (Oughstun &
Sherman, 1988, 1994)
r ( " #)
1 c z b2 =21
As z, t exp 0 1 + 1 + 2
b 2z c + 20 1 2
82 3
> 3 3
< c 20 1 2 + c 20 1 2
6 2 2 7
4 2 5
>
: c 2
+ 2
0 1 + 2
+ c 2
+ 2
0 1 +
" ! #
z b2 =2
cos 1 + 2 +
c + 20 1 2 4
" #
0 5 + 3c 1 5 3c 1
+
2 + c 2 + 20 1 + 2 c 2 + 20 1 + 2
" ! #)
z b2 =2
sin 1 + 2 + ,
c + 20 1 2 4
(83)
r
bc
2 13=4 e20 c 1
z
As z,t
(" 2z
b2 11=2 c
b2 11=2 c 2 + 420
#
b2 11=2 + c z p
cos b 2 1 + (84)
b2 11=2 + c 2 + 420 c 4
"
1
+ 20 1=2
b2 1 c 2 + 420
# )
1 z p
sin b 2 1 + ,
b2 11=2 + c 2 + 420 c 4
the near saddle points were provided by Oughstun and Sherman, as shown in
Section 2. Here, the uniform asymptotic method of Chester et al. (1957) is
applied to obtain the asymptotic expansion of the integral representation of
the propagated field [(1.58) in Part I] about the near saddle points SPn ().
The asymptotic expansion of AH(z, t) as z ! 1 about the upper first-
order near saddle point SP+N() for 1 < < 1, about the second-order near
saddle point SPN for 1, and about both first-order near saddle points
SPN() for > 1 (see Figures 27) yields the dynamical evolution of the
second forerunner, or Brillouin precursor, Ab(z, t). A direct application of
the method of steepest descent does not result in a uniform approximation
to the propagated field because the order of the saddle points changes dis-
continuously at 1. A uniform expansion, valid for all 1, is obtained
through use of the method due to Chester et al. (1957). Here, the transfor-
mation that retains the behavior of the saddle points about 1 is
1 3
1 0 + , 0, (86)
3
where 0, 1 are functions to be determined, and the branch of the transfor-
mation (86) to be used is the one that defines a conformal mapping of some
disc that contains both saddle points SPn () (see Bleistein & Handelsman,
1975). The amplitude function is expanded as a finite number of terms plus a
remainder that is regular and equal to zero at the transformed saddle points
1 d
G0 , h1 + h2 + 2 1 H0 , , (87)
c d
where h1,h2, and H0 are to be determined. Substitution of (86) and (87) into
(1.58) in Part I results in the uniform asymptotic expansion of the Brillouin
precursor Ab(z,t) for the step-function-modulated sine wave (Oughstun &
Sherman, 1988, 1994)
h i
c 1=3 2=3
z i2=3 i2=3 c
Ab z, t Re exp 0 e Ai 1 e
( " c z #) z
1 i i
h + + h
2 SPN+ c SPN c
c 2=3 2=3
i2=3 1 i2=3 c
+ e Ai 1 e
(z " z #)))
1 i i
h + h ,
2 SPN+ c SPN c
1=2
1
(88)
302 Natalie A. Cartwright and Kurt E. Oughstun
as z ! 1 for all 1, to leading order. Here, Ai() and Ai(1)() denote the
Airy function and its first derivative, respectively, and the coefficients appe-
aring in Equation (88) are given by
1h i
0 SPN+ , + SPN , , (89a)
2
h i1=3
1=2 3
1 SPN+ , SPN , , (89b)
4
!1=2
1=2
21
h
2 , (89c)
SPN+ ,
for 6 1. At the critical value 1 when the two near saddle points coalesce
into a single second-order saddle point, these coefficients take on the lim-
iting values
!1=3
2
lim h 3 hs 1 , (90a)
!1 SPN , 1
" #
1 i i i
lim h + + h hs 1 , (90b)
!1 2 SP + c SPN c SPN c
N
" #
1 i i d i
lim 1=2 h + h h2s 1 ,
!1 2
1
SPN+ c SPN c d SPN c
(90c)
0.2
0.15
0.1
Ab(z,t)
0.05
0.05
0.1
1 1.5 2 2.5
q = ct/z
Figure 15 Dynamic behavior of the Brillouin precursor field Ab(z, t) for the step-function-
modulated sine wave with below-resonance-applied signal frequency c 0.5 0 at a
relative observation distance z 3.8 zd into the single-resonance Lorentz dielectric with
Brillouin's choice of material parameters as described by the uniform asymptotic theory
with numerically determined near saddle point locations.
is zero only at the single spacetime point ct/z 0; this does not mean
that the field is static about this point (L. Felsen & Marcuvitz, 1973).
A physically meaningful effective oscillation frequency has been defined
by Oughstun (2009) over the initial spacetime domain 2 (1, 1] by the
temporal width of the build-up to the peak amplitude point that occurs
between the spacetime points SB and 0. The difference between
these two spacetime points is given by 0 0 SB 420 b2 =30 40 .
This difference corresponds to the effective half-period Teff 2/eff of the
field over this spacetime domain through the relation 0 (c/z)(Teff/2),
so that
30 40 c
eff 0 : (91)
420 b2 z
Notice that this effective angular oscillation frequency of the Brillouin pre-
cursor at 0 asymptotically approaches zero as z ! 1, in agreement
with the limiting behavior of Re{SP+n ()} as ! 1+ .
Although numerically determined distant saddle point positions (as a
function of ) may always be used in the exact expressions for the complex
phase function and its second derivative appearing in Equation (88) in order
to obtain the precise asymptotic behavior of the Brillouin precursor for a
304 Natalie A. Cartwright and Kurt E. Oughstun
1
SPn+ , 20 3 0 0
3
b2
+ 20 3 0 2 20 3 + 3 0
,
540 40
(92)
2
b
00 SPn+ , 20 1 + 3 0
: (93)
0 40
For 1 < 1, the spectrum of the envelope function at the near saddle
points is given by
1
i 3 0
20 ic
u~H SPn c 3
SPn c 1
2c + 3 0
20 2
9
(94)
1=2
20 c 0 c=z
Ab z,t
1 40 1 + 6 0
b 2c + 3 0 20 2
9
nz h
exp 20 3 0 0
3c
b 2 h io
+ 20 3 0 20 3 3 0 ,
180 40
(95)
as z ! 1. For 0, this nonuniform asymptotic approximation simplifies
somewhat to
1=2
20 c 0 c
Ab z, t " !2 #
20 6z
b 2c +
3
z 20 b2 20 4
exp 0 + + 0 ,
c 3 20 40 3 3
(96)
as z ! 1, where
1=2
420 20 40
+ 0 :
92 3b2
at which point the Brillouin precursor varies with the propagation distance
z > 0 only as z1/2 as z ! 1, making this spacetime point in the field
evolution entirely unique.
306 Natalie A. Cartwright and Kurt E. Oughstun
3.3.2 Case 2: 5 1
At 1, the two near saddle points have coalesced into a single
second-order saddle point SPn located at SPn 1 i 2 3 with
0
0 00
1 0 41 + b =30 1 1, where (SPn, 1) (SPn, 1)
2 2 2 2 2
420 b2 220 b2
0 1 + ,
90 40 90 40
and the second precursor field attenuates exponentially with increasing
propagation distance z > 0 at the fixed spacetime point 1.
In the special (limiting) case when 0 0, it follows that 1 0 exactly
and the two first-order near saddle points coalesce into a single second-order
Part II. The Modern Asymptotic Theory 307
saddle point at the origin. The peak amplitude in the Brillouin precursor
then decays with the propagation distance z > 0 only as z1/3 [see (3.26)
in Part I] instead of the usual z1/2 behavior described by Equation (97)
when 0 > 0 (Oughstun, Cartwright, Gauthier, & Jeong, 2010).
where the final approximation is valid for all > 1 such that 3()
20(() 1). The spectrum of the envelope function at the near saddle
points is given by
2
i 0 + i c
u~H SPn c 3
SPn c 4
c 2 + 20 2
9
(104)
for 1. Thus, for values of bounded away from 1 such that
(z/c)2/3j1()j 1, substitution of Equations (101)(104), together with
308 Natalie A. Cartwright and Kurt E. Oughstun
the large argument asymptotic expansion of the Airy function and its deriv-
ative, into the uniform expansion (88) reduces to the nonuniform expansion
given by Oughstun and Sherman (1988, 1994) valid for spacetime values
> 1
r
20 c
0 c
2
3z
Ab z, t
4 2 2 4 2 2
b c + 0 + c + 0
2 2
9 9
z 2 b2
exp 0 0 + 4 1
2
c 3
0
4 0
4
+ 20 2 1 2c + 20 2 2
9 3 9
2 2 #
z 0 b 42 2
cos 0 + + +
c 20 04
3 02
4
2 2
4 z b 42
+ 0 sin 0 + 0 4
3 #) c 20 0 3
2
+ +
20 4
(105)
as z ! 1. This nonuniform asymptotic approximation reduces to the clas-
sical result [see Equation (2.36c) in Part I] given by Brillouin (1914, 1960) if
() and () are further approximated by expressions that are valid only for
spacetime points not too distant from 0. Hence, Brillouins classical
expression for the second precursor field over the spacetime domain
> 1 is an approximation, valid for near 1, of an expression that becomes
invalid as approaches 1 from above. As a result, Brillouins expression
for the asymptotic behavior of the second precursor field (or second forerun-
ner) for the unit step-function-modulated signal over the spacetime
domain > 1 is not applicable.
thus giving rise to the signal contribution. With this assumption, Brillouin
reasoned that the signal contribution describes a smooth transition from the
second forerunner to the residue contribution [(2.37) in Part I]. Brillouins
analysis of the signal contribution is limited not only by its restrictions on the
carrier frequency c, but also because a uniform asymptotic method is
required to account for a saddle point in the vicinity of a singularity of
the amplitude function. Such a uniform method provides the smooth tran-
sition of the field into the residue contribution that occurs at the spacetime
point s when the path of integration crosses the pole.
Let s > 1 denote the spacetime point at which the path of integration
crosses the simple pole located at c, determined by the equation
c , s SP , s , (106)
where Im{} and SP denotes either +D() or +N(). The uniform
expansion method developed by L. B. Felsen and Marcuvitz (1959) and
L. B. Felsen (1963) for a saddle point in the vicinity of a simple pole, and
later generalized by Bleistein (1966) for a saddle point in the vicinity of a
singularity, is obtained using the same transformation as for the method
of steepest descent
, SP , s2 , (107)
but with an expansion of the amplitude function as
1 d
+ Ts: (108)
c ds s b
Here, s is real-valued along the path of steepest descent, is the residue of the
amplitude function at c, b is the location of the transformed pole, and T(s) is
regular at both s 0 (the transformed saddle point) and at s b.
Substitution of Equations (107) and (108) into the integral representation
of the propagated field given in Equation (1.58) in Part I yields
( )
1 hz iZ 1 z=cs2
AH z,t Re exp SP , e ds
2 c P s b
( ) (109)
1 hz iZ
T sez=cs ds ,
2
Re exp SP ,
2 c P
where P() is the path of integration that has been deformed through the
valleys of the accessible saddle points. The second term in Equation (109)
is the method of steepest descent applied to a first-order saddle point located
310 Natalie A. Cartwright and Kurt E. Oughstun
4
Natalie Cartwright received her PhD in mathematics in 2004 from the University of Vermont under
the supervision of Kurt Oughstun.
Part II. The Modern Asymptotic Theory 311
Z 1
2
ex dx,
2
erfc p (111)
0.03
0.02
0.01
Ac(z,t)
0.01
0.02
0.03
1 1.5 2 2.5
q = ct/z
Figure 16 The uniform asymptotic approximation of the signal contribution for a step-
function-modulated sine wave with below absorption band frequency c 0.5 0 at an
observation distance of z 1.2 106 m 3.8 zd into the single-resonance Lorentz
dielectric with Brillouin's choice of material parameters. The asterisk denotes the
spacetime value s at which the path P() crosses c.
312 Natalie A. Cartwright and Kurt E. Oughstun
0.2
0.15
0.1
AH (z,t)
0.05
0.05
0.1
1 1.5 2 2.5
q = ct/z
Figure 17 The uniform asymptotic approximation of the dynamical evolution of the
propagated field contribution for a step-function-modulated sine wave with below
absorption band frequency c 0.5 0 at an observation distance of z 1.2 106
m 3.8 zd into the single-resonance Lorentz dielectric with Brillouin's choice of material
parameters.
wave field. The value of s depends upon which Olver-type path is chosen
for P(). If that path is taken to lie along the path of steepest descent through
the saddle point nearest the pole, then, because (, ) Im{(, )} is
constant along the path of steepest descent, it follows that the value of s
is defined by the expression (Brillouin, 1914, 1960)
sp ,s c , s , (114)
where sp sp() denotes the saddle point that interacts with the pole sin-
gularity. At s, however, the pole contribution is asymptotically negli-
gible in comparison to the saddle point contribution because P() is an
Olver-type path with respect to that saddle point. Consequently, the
particular value of s at which the pole contribution occurs is of little or no
significance to the asymptotic behavior of the propagated wave field
AH(z, t). An example of such an Olver-type path at a fixed spacetime point
> 1 when the two near saddle points SP n are dominant over the two distant
saddle points SPd in a single-resonance Lorentz model dielectric is depicted in
Figure 18. The path P() through the pair of near saddle points SP n and SPn
+
can lie anywhere within the shaded region of the figure. With the path P()
shown in this figure, < s if the pole p lies in the angular frequency interval
c2 < p < c4, > s if p lies within either of the angular frequency inter-
vals 0 p < c2 or p > c4, and s if either p c2 or p c4. If the
314 Natalie A. Cartwright and Kurt E. Oughstun
w w w+ w+
SPd SPd+
P(q)
Figure 18 A deformed contour of integration P() passing through both the near and
distant saddle points for a fixed spacetime value > 0. This contour is an Olver-type
path with respect to the near saddle point SP+n in the right-half of the complex -plane,
and is an Olver-type path with respect to the near saddle point SP n in the left-half of the
complex -plane. The lighter shaded area indicates the region of the complex -plane
wherein the inequality (, ) < (SPn , ) is satisfied and the darker shaded area indi-
cates the region of the complex -plane wherein the inequality (, ) < (SPd, ) is
satisfied. Reprinted with kind permission of Springer Science+Business Media from
Oughstun (2009), Fig. 15.1, p. 510.
sp , c c , (115)
behavior of the propagated wave field and the pole contribution is asymp-
totically negligible by comparison. For spacetime values > c such that
the inequality (sp, ) < (c) is satisfied, however, the pole contribution
is the dominant contribution to the asymptotic behavior of the propagated
wave field and the saddle point contribution is asymptotically negligible by
comparison. For example, for the spacetime point depicted in Figure 18,
Equation (115) is satisfied if either c c1 or c c6 and the value of is
then c for either of these two pole locations. Furthermore, for the space
time value depicted in Figure 18, < c and the inequality (SP+n , ) >
(p) is satisfied if c1 < c < c6, and > c and the inequality (SP+n , )
< (c) is satisfied if either 0 c < c1 or c > c6. Consequently, the
pole contribution is asymptotically negligible in comparison to the saddle
point contribution to the propagated wave field at the spacetime value
depicted in Figure 18 if c1 < c < c6, whereas the pole contribution
is the dominant contribution to the asymptotic behavior of the propagated
wave field and the saddle point contribution is asymptotically negligible in
comparison to it if either 0 c < c1 or c > c6.
Based upon these results, it is seen that the signal arrival for a fixed value of
c occurs at the spacetime point c satisfying Equation (115). Notice
that this definition of the signal arrival yields a signal velocity (Oughstun &
Sherman, 1975, 1980, 1988)
c
vc c (116)
c c
that is independent of the initial pulse envelope function and the propaga-
tion distance z, depending only upon the dispersive medium properties and
the value of c. Notice that this pulse velocity measure always satisfies rel-
ativistic causality; that is, vc(c) c for all c. In addition, notice that
Brillouins definition of the signal velocity in Equation (2.38) in Part I is
in terms of s and so is referred to here as vs in order to distinguish it from
the correct definition given in Equation (116).
In summary, the signal arrival in a single-resonance Lorentz model
dielectric separates naturally into two distinct cases dependent upon the
value of the real angular frequency c of the pole in comparison to the crit-
ical angular frequency value SB defined in Equation (63). For values of c
in the angular frequency interval 0 c SB, the signal arrival is due to the
crossing of the isotimic contour () (sp) with the simple pole singu-
larity at c, where sp denotes the location of the upper near saddle
point SP+n for 1 < < 1, the second-order near saddle point SPn at
316 Natalie A. Cartwright and Kurt E. Oughstun
1, and the near saddle point SP+n for all > 1. For such values of c, the
signal due to the pole contribution at c is preceded by the first and
second precursor fields, and the signal evolves essentially undisturbed as
increases above c. For pole values c > SB, however, the signal arrival first
occurs due to the crossing of the isotimic contour () (SP+d , ) through
the distant saddle point SP+d with the simple pole singularity. This first arrival
occurs at some spacetime point 2 (1, SB) for finite c. At some later
spacetime point 2 (SB, 0), this pole is again crossed, but in the opposite
direction, by the isotimic contour () (SP+n , ) through the upper near
saddle point SP+n , rendering the pole contribution asymptotically less dom-
inant than the second precursor field. Finally, for some still later spacetime
point > 0, the pole is again crossed in the original direction by the isotimic
contour () (SP+n , ) through the near saddle point SP+n so that it
finally becomes asymptotically dominant over all other contributions to
the propagated wave field AH(z,t) for all remaining spacetime values. Con-
sequently, for pulsed sources with c > SB there is the existence of a
so-called prepulse (Oughstun & Sherman, 1988, 1994) due to the interrup-
tion of the signal evolution by the second precursor field which becomes
dominant over the pole contribution for some short spacetime interval.
This prepulse formation is seen to be an integral part of the dynamic evo-
lution of the second precursor field superimposed upon the evolution of
the signal contribution. The spacetime evolution of the signal contribution
when c > SB may then be considered to be separated into three parts: the
so-called prepulse which is preceded by the first precursor and then followed
by the second precursor field superimposed upon the signal contribution,
which is then finally followed by the signal which remains dominant for
all later spacetime points. It is important to keep in mind that the prepulse
is not independent of the signal evolution. Indeed, the prepulse formation is
simply a consequence of the superposition of the signal (or pole) contribu-
tion with the second precursor field which becomes dominant over the sig-
nal for a finite spacetime interval.
The spacetime point c at which the pole contribution is of equal
dominance with the second precursor field is defined by the relation
SPn+ ; c c ; c 0 , (117)
where SP+n () denotes the location of the upper near saddle point SP+n for
c 2 [0, 1), the second-order near saddle point SPn at c 1, and the near
saddle point SP+n for all c > 1. For all spacetime points > c, the pole
Part II. The Modern Asymptotic Theory 317
vc1/c
0.8
vc2/c
0.6
vE/c
v/c
vc/c
0.4
0.2
0
0 w0 5 w1 wSB 10 15
wc (r/s) 1016
Figure 19 Angular frequency dependence of the relative main signal velocity vc(c)/c
1/c(c), relative anterior presignal velocity vc1(c)/c 1/c1(c), and relative poste-
rior presignal velocity vc2(c)/c 1/c2(c) in a single-resonance Lorentz model dielec-
tric characterized by Brillouin's choice of the medium parameters. The dashed curve
describes the frequency dependence of the relative energy transport velocity
vE(c)/c in the dispersive medium. Reprinted with kind permission of Springer Science
+Business Media from Oughstun (2009), Fig. 15.3, p. 518.
22 2
when 1 0 + 300 p4 . At the spacetime point 0, the dominant
0
near saddle point SP+n crosses the origin [SP+n (0) 0] so that its contribu-
tion to the asymptotic behavior of the propagated wave field experiences
zero exponential attenuation [ SPn+ 0 , 0 0], the peak amplitude
point in the wave field decaying only as z1/2 with z ! 1. At the subse-
quent spacetime point 1, this contribution to the asymptotic wave
experiences a small (but nonzero) amount of exponential attenuation with
propagation distance as well as a z1/3 algebraic decay as z ! 1, provided
that 0 > 0. In the singular dispersion limit as 0 ! 0, however, the two near
saddle points SPn coalesce into a single second-order saddle point at the ori-
gin when 1 ! 0, resulting in a peak amplitude that experiences zero
exponential attenuation, the amplitude now decaying only as z1/3 as
z ! 1, as described by Brillouin (1960) (see Section 3.3 of Part I).
The numerically determined peak amplitude decay with relative prop-
agation distance z/zd is presented in Figure 20 for an input Heaviside unit
10 0
Peak amplitude of the Brillouin precursor
101
de = 3.02 1013 r/s
d 0 = 0.1 de
d 0 = 0.01 de
102 d 0 = 0.001 d e
103
0 20 40 60 80 100
z/zd
zd 1 c (125)
0.02
z/zd = 10
Brillouin
0.01 precursor
A(z,t)
0
Sommerfeld
precursor
0.01
0.02
2 3 4 5 6
t (1010 s)
Figure 21 Propagated signal with below-resonance angular carrier frequency c
0.769 0 in a single-resonance Lorentz model dielectric in the singular dispersion limit.
The -symbol in the plot marks the time t0 0z/c following the onset of the Brillouin
precursor. Reprinted with kind permission of Springer Science+Business Media from
Oughstun (2009), Fig. 15.09, p. 655.
Part II. The Modern Asymptotic Theory 323
1 z1 t1 2 z2 t2 , (126)
1 z1 2 z2 , (127)
both of which must be satisfied for all . For the absorptive part, the equiv-
alence relation
1
z2 z1 , 8 (128)
2
results. If the two media differ only through their densities, then
p 1
n 1 + Ng ! 1 + Ng as N ! 0,
2
because () (/c)ni() for real , where g 4q2e =me =
2 20 + 2i0 . In this case, ni 12 Ngi and the above equivalence
relation becomes
N1
z2 z1 , (129)
N2
the accuracy of this result increasing as N1, N2 ! 0 with N2 > N1. The
corresponding equivalence relation for the phase part then results in
N1 z1
t2 t1 + 1 , (130)
N2 c
p
which is simply a uniform displacement in time. Because p 4Nq2e =me ,
the pair of equivalence relations given in Equations (129) and (130) may be
expressed as
Part II. The Modern Asymptotic Theory 325
2p1
z2 z1 , (131)
2p2
!
2p1 z1
t2 t1 + 1 : (132)
2p2 c
The first part of this equivalence relation is the essential part. For example, if
N1/N2 1 102, then z2 z1 102 and t2 t1 (0.33 108 s/m)z1.
In that case, the propagated wave field structure illustrated in Figure 21 also
applies to the case when the medium plasma frequency p is reduced by
the factor 10 and the propagation distance z is increased by the factor 100
provided that the time origin is adjusted according to Equation (130).
The accuracy of this equivalence relation is illustrated in the sequence of
propagated signal plots presented in Figures 2224. In each case, the initial
signal has below-resonance angular carrier frequency c 3.0 1014 r/s
and is propagating in a single-resonance Lorentz model dielectric with res-
onance frequency 0 3.90 1014 r/s and phenomenological damping
constant 0 3.02 1012 r/s. The plasma frequency in case 1, illustrated
in Figure 22, is p1 3.05 1014 r/s, resulting in an absorption depth
14
0.1 wp1 = 3.05 10 r/s
4
zd = 7.238 10 m
0.05
A(z,t)
0.05
0.1
0 1 2 3 4
12
t z/c ( 10 s)
Figure 22 Propagated wave field at five absorption depths in a single-resonance
Lorentz medium with plasma frequency p1 3.05 1014 r/s. Notice that A(z, t) 0
for all t < z/c.
326 Natalie A. Cartwright and Kurt E. Oughstun
12
0.1 wp2 = 3.05 10 r/s
zd = 0.458 m
0.05
A(z,t)
0.05
0.1
0 1 2 3 4
12
t z/c ( 10 s)
Figure 23 Propagated wave field at five absorption depths in a single-resonance
Lorentz medium with plasma frequency p2 3.05 1012 r/s. Notice that A(z, t) 0
for all t < z/c.
Part II. The Modern Asymptotic Theory 327
10
wp3 = 3.05 10 r/s
0.1 3
zd = 4.582 10 m
0.05
A(z,t)
0.05
0.1
0 1 2 3 4
12
t z/c ( 10 s)
Figure 24 Propagated wave field at five absorption depths in a single-resonance Lorentz
medium with plasma frequency p3 3.05 1010 r/s. Notice that A(z, t) 0 for all t < z/c.
AS(z,t)
0
AB(z,t)
0
Ac(z,t)
0
A(z,t)
Much later, Gitterman and Gitterman (1976) again considered the trans-
mission of a step-modulated signal into a Lorentz half-space. They used both
low- and high-frequency expansions of the dielectric permittivity to obtain
closed-form expressions for the locations of the saddle points of the phase
function, applying nonuniform asymptotic methods to produce a set of
explicit time-domain descriptions of the transmitted signal, each with its
own region of validity. The combination of their approximations for the
saddle point locations with nonuniform asymptotic theory severely limits
the accuracy of these descriptions, even within each of their regions of valid-
ity (see Cartwright, 2011). The same problem, but limited to the case of nor-
mal incidence, was considered by Mokole and Samaddar (1999). Following
the work of Oughstun and Sherman (1994) and based upon accurate
approximations for the saddle point locations and employing uniform
asymptotic methods, they provided a time-domain asymptotic description
of the early-time response of the transmitted field. In addition, they recast
the integral representation of the reflected field into a form that is
readily amenable to numerical calculation.
A complete uniform asymptotic description of a step-modulated sinusoid
obliquely incident on a Lorentz half-space, valid as the propagation (or pen-
etration) distance into the Lorentz medium increases, was finally given by
Cartwright (2011). The asymptotic approximations are relatively straight-
forward extensions of the asymptotic approximations derived in the study
of linearly polarized electromagnetic plane-wave pulse propagation through
unbounded media. The expressions are completely uniform in time and thus
improve and extend the early-time results of Colby (1915), improve and
unify the limited approximations provided by Gitterman and Gitterman
(1976), and extend the work of Mokole and Samaddar (1999) to all times
and arbitrary angle of incidence. These closed-form solutions provide both
the arrival time and angle of refraction of prominent physical features of the
transmitted pulse.
5.1 Formulation
Consider an incident, linearly polarized electromagnetic plane-wave pulse
propagating in Region 1 (z < 0) that impinges upon the planar boundary
at z 0 with Region 2 (z > 0), where Region 1 is vacuum (E1 E0)
and Region 2 is a single-resonance Lorentz medium, as depicted in
Figure 26. The relative dielectric permittivity of Region 2 is described by
the frequency-dependent expression
330 Natalie A. Cartwright and Kurt E. Oughstun
x
Hr
Er kR Et kT
Region 1
e1 = e0 , m1 = m0 aR Ht
aT
n z
a
Region 2
Ei kI e2 = e2(w) , m2 = m0
Hi
Figure 26 Geometric illustration of a pulsed TE-polarized plane wave obliquely incident
upon a planar interface at z 0.
b2
E2 1 , (133)
2 20 + 2i0
[see Equation (1.30) in Part I]. In both regions, the magnetic permeability is
that of free space ( 0).
If the incident signal Ei(r, t) is a transverse electric (TE)-polarized plane
wave with angle of incidence , where 0 < /2, then the transmitted
electric and magnetic field components are given exactly by Ex Ez
Hy 0 and the integral representations (see Marozas, 1998 for a thorough
derivation)
0 1 0 1
Ey Z ia + 1 T? ;
B C 1 B C
@ Hx A E i @ T? ;cos N2 ;=0 A
2 ia1 (134)
Hz T? ;sin =0
expikT ; r it
d:
Here, a is greater than the abscissa of absolute convergence of the incident
signal, E~ is the temporal Fourier transform of the electric field of the
i
incident signal, r is the position vector in Region 2, and T?() is the trans-
mission coefficient for the TE mode, given by
2
T? ; : (135)
1 + N2 ;
The wavevector in Region 2 is given by
D E
kT ; sin,0, N2 ;cos , (136)
c c
where the function N2(; ) is defined as
Part II. The Modern Asymptotic Theory 331
s
b2 =cos
N2 ; 1 2 (137)
20 + 2i0
p
with 0 0 =E0 denoting the impedance of free space.
If the incident signal Ei(r, t) is a transverse magnetic (TM)-polarized
plane wave with angle of incidence , where 0 < /2, then the
transmitted electric and magnetic field components are given exactly by
Ey Hx Hz 0 and the integral representations
0 1 0 1
Ex Z ia + 1 Tjj ;cos N2 ;=n2
B C 1 B C
E~ B C
i
@ Ez A @ Tjj ;sin=n2 A
2 ia1 (138)
Hy Tjj ;n2 =0
expikT ; r it
d:
2n2
Tjj ; (139)
n22 + N2 ;
p
where n2 E2 =E0 is the complex index of refraction in Region 2.
As for the SommerfeldBrillouin problem, the incident electric field at
z 0 is given by the canonical step-modulated sinusoid
E i t utsinc t, (140)
where u(t) denotes the Heaviside step function and c 2 + is the fixed car-
rier frequency of the signal. Substitution of the temporal Fourier transform
of the incident electric field E~ into either Equation (134) or (138), and
i
of Equation (141). Here, the scaled variables are indicated by a caret (e.g.,
^ =0 ), the complex phase function ,
^ is defined as
hz x i
^ ; i
, ^ sin ,
^ cos N2 ; (142)
d d
with
d x sin + zcos (143)
the penetration depth into Region 2 along the refracted ray, and ct/d is a
spacetime variable defined for all d > 0. These exact integral representa-
tions of the transmitted components are now in the form of those considered
in the study of a step-modulated sine wave propagating through an
unbounded single-resonance Lorentz material (see Section 3), the only dif-
ferences being in the additional amplitude functions and the functional form
of the phase function. Thus, the same uniform asymptotic methods that are
used in the study of a step-modulated sine wave propagating through an
unbounded, single-resonance Lorentz material may be used in the study
of a step-modulated sine wave impinging on a single-resonance Lorentz
half-space. As described by Cartwright (2011), the similarities and differ-
ences between the two problems are now given along with the final
asymptotic expressions for the transmitted pulse.
Rather than perform integration around these branch cuts, saddle point
methods are the preferred method of analysis when > 1.
Part II. The Modern Asymptotic Theory 333
H S + B + c : (146)
The first term in each asymptotic expansion, uniformly valid for all 1, is
given in the following (Cartwright, 2011).
334 Natalie A. Cartwright and Kurt E. Oughstun
as d ! 1, for all 1, where Jn() denotes the Bessel function of the first
kind of integer order n. The coefficients appearing in (147) are given by
n o n o
A; Im ^ SPD+ , B; iRe ^ SPD+ , (148a)
2 !1=2
1 4 ig ^ SPD+ 1 4A3
0 ;
2 ^ SPD+ ^ c 2A i2
^ SP +
D
!1=2 3
(148b)
ig^ SP 1 4A 3
+ D
2 5,
^ SPD
^c 2A i ^ SPD
2 !1=2
1 4 ig ^ SPD+ 1 4A3
1 ;
4A ^ SPD+ ^ c 2A i2
^ SP + D
!1=2 3
(148c)
ig^ SPD 1 4A3
2 5,
^ SPD
^c 2A i ^ SPD
( (
d0 1 c 1=3 i2=3
B r,ct=d Re exp 0 ; e Ai;
c 2 d0
i
^ SPN+ ;h + ;
g
^ SPN+
^c
!
i
+ ^ SPN ;h ;
g
^ SPN
^c
1 c 2=3 i4=3 1
+ 1=2 e Ai ;
21 ; d0
i
g^ SPN+ ;h + ;
^ SPN+
^c
!))
i
g^ SPN ;h ; ,
^ SPN
^c
(149)
where Ai() denotes the Airy function and the coefficients appearing in
Equation (149) are given by
h i1=3
1h i
1=2 3
0 ; ^ SPN+ +
^ SPN , 1 ; ^ SPN+
^ SPN ,
2 4
(150a)
" 2=3 # !1=2
1=2
c 21
; 1 ei2=3 , h ;
2
,
d0 ^ SPN+
(150b)
for 61. At the critical value 1, when the two near saddle points coalesce
into a single second-order saddle point N, these coefficients take on the
limiting values
!1=3
2
lim h ; 3 h1 1 ;, (151a)
!1 N
" #
1 i i i
lim ^ SPN+ h + +
g ^ SPN h
g gN h1 ,
!1 2 ^ SPN+ ^c ^ SPN
^c N ^c
(151b)
336 Natalie A. Cartwright and Kurt E. Oughstun
" #
1 i i
lim ^ SPN+ h +
g ^ SPN h
g
!1 21=2 ^ SPN+ ^c ^ SPN
^c
1
(151c)
2 d ^
ig
h1 :
d ^ ^ c
^ N
B
Ey (V/m)
c
0.2
0.1
0
0.1
0.2
1 2 3 4 5 6
q
Figure 27 The three contributions S, B, c and the total transmitted field Ey S +
B + c of the electric field component of the TE-polarized plane-wave pulse with
10 at the observation point (x, z) (0.1 m, 0.2 m).
338 Natalie A. Cartwright and Kurt E. Oughstun
A 104
6
8
1 3 4 5
q
B 105
8
6
Hz, TE case (A/m)
6
1 2 3 4 5
q
Figure 28 The magnetic field components (a) Hx and (b) Hz for the TE-polarized plane-
wave pulse. The same material and pulse parameters and observation point used for
Figure 27 are used here.
^ 3 1:
lim Re (155)
!=2
half-space, the near saddle point SP+N also crosses the origin, thereby pre-
serving this special feature of the Brillouin precursor, but now the peak
amplitude point occurs at the spacetime point
z cos x sin
0 N2 0; + (158)
d d
and is refracted at an angle of
!
tan sin
0T arctan arctan p (159)
RefN2 0;g cos 2 + b^2
into Region 2, in agreement with the numerical computations performed by
Blaschak and Franzen (1995). These authors studied precursor propagation
assuming a trapezoidal modulated sinusoid at oblique incidence on a Lorentz
p
half-space with material parameters b 24 1016 s1 , 0 4 1016 s1,
and 0 0.28 1016 s1. With an angle of incidence 80, the authors
found, using purely numerical estimates, the leading edge of the Sommerfeld
precursor to be at an angle of ST 80 and the peak of the Brillouin pre-
cursor to be 0T 40, in agreement with that given by both Equations (157)
and (159). The angle of transmission of the entire Brillouin precursor BT is
found by substitution of SP+N(; ) into Equation (156). As ! 1,
SPN+ ! ^ 1 and BT ! 0 so that the Brillouin precursor is continuously
bent toward the normal with increasing , in agreement with that given
in Gitterman and Gitterman (1976).
The signal contribution, characterized by the smooth turn on of oscilla-
tions at the carrier frequency of the incident pulse, is refracted at the constant
angle
tan
T arctan
c
(160)
RefN2 ^ c ;g
for all . The velocity of the main signal is defined as vc c/c, where c is
the spacetime point at which the signal contribution is of equal dominance
with the Brillouin precursor (Oughstun & Sherman, 1980), given implicitly
by the Equation
^ c ;:
Re SPN+ c , c ; Re (161)
An example of the dependence of the refractive angles ST, BT, cT on is
shown in Figure 29 for the same material and signal parameters, angle of
incidence 10, and observation point (x, z) (0.1 m, 0.2 m) used
Part II. The Modern Asymptotic Theory 341
30
Transmission angle 25
S
20 a T
B
a T
15 c
a T
10
in Figure 27. The open circle represents the angle of refraction 0T of the
peak amplitude point 0 at which B experiences algebraic attenuation.
The asterisk represents the spacetime point c at which the main signal
velocity is defined.
The relative velocity of the main signal vS/c 1/c for various angles of
incidence is shown in Figure 30 using the same material and signal param-
eters and observation point (x, z) (0.1 m, 0.2 m) used in Figure 27.
Here, the angles of incidence are 0, 10, 30, 50, 70. As with the
unbounded case, the signal velocity obtains its minimum approximately
^ 0 and is subluminal for all .
at ^
6. CONCLUDING REMARKS
The fundamental analysis that was begun by Sommerfeld and
Brillouin in 1914 concerning the propagation velocity of a pulse in a Lorentz
model dielectric has now seen its conclusion in 2007 with the publication of
342 Natalie A. Cartwright and Kurt E. Oughstun
0.7
0.6
0.5
0.4
vs/c
0.3
0.2
0.1
0
0 1 2 3 4 5
w, s-1
Figure 30 The relative velocity of the main signal vS/c 1/c for angles of incidence
0, 10, 30, 50, 70depicted by the solid, dotted, dashed, and dashed-dotted cur-
ves, respectively. The same material and pulse parameters and observation point used
for Figure 27 are used here.
ACKNOWLEDGMENTS
The research review presented in this chapter by the authors was supported, in part, by the
United States Air Force Office of Scientific Research (AFOSR).
REFERENCES
Baerwald, H. (1930). U ber die fortpflanzung von signalen in disperdierenden medien. Annals
of Physics, 7, 731760.
Blaschak, J. G., & Franzen, J. (1995). Precursor propagation in dispersive media from short-
rise-time pulses at oblique incidence. Journal of the Optical Society of America B: Optical
Physics, 12(7), 15011512.
Bleistein, N. (1966). Uniform asymptotic expansions of integrals with stationary point near
algebraic singularity. Communications on Pure and Applied Mathematics, XIX(4), 353370.
Bleistein, N., & Handelsman, R. (1975). Asymptotic expansions of integrals. New York: Dover.
Brillouin, L. (1914). U ber die fortpflanzung des licht in disperdierenden medien. Annals of
Physics, 44, 204240.
Brillouin, L. (1960). Wave propagation and group velocity. New York: Academic.
Cartwright, N. A. (2004). Uniform asymptotic description of the unit step function modulated sinu-
soidal signal. PhD dissertation: University of Vermont.
Cartwright, N. A. (2011). Electromagnetic plane-wave pulse transmission into a Lorentz
half-space. Journal of the Optical Society of America. A, 28(12), 26472654.
Cartwright, N. A., & Oughstun, K. E. (2006). Uniform asymptotics applied to ultrawideband
pulse propagation. SIAM Review, 49(4), 628648.
Cartwright, N. A., & Oughstun, K. E. (2009). Ultrawideband pulse propagation through a
lossy plasma. Radio Science, 44. http://dx.doi.org/10.1029/2009RS004201.
Chester, C., Friedman, B., & Ursell, F. (1957). An extension of the method of steepest
descents. Proceedings of the Cambridge Philosophical Society, 53, 599611.
Colby, W. (1915). Signal propagation in dispersive media. Physics Review, 5, 253265.
Felsen, L., & Marcuvitz, N. (1973). Radiation and scattering of waves. Englewood Cliffs, NJ:
Prentice-Hall, Inc.
Felsen, L. B. (1963). Radiation from a uniaxially anisotropic plasma half-space. IEEE Trans-
actions on Antennas and Propagation, AP-11, 469484.
Felsen, L. B., & Marcuvitz, N. (1959). Modal analysis and synthesis of electromagnetic fields.
Polytechnic Institute of Brooklyn Microwave Research Institute Report.
Gitterman, E., & Gitterman, M. (1976). Transient processes for incidence of a light signal on a
vacuum-medium interface. Physical Review A, 13(2), 763776.
Handelsman, R. A., & Bleistein, N. (1969). Uniform asymptotic expansions of integrals that
arise in the analysis of precursors. Archive for Rational Mechanics and Analysis, 35, 267283.
344 Natalie A. Cartwright and Kurt E. Oughstun
Jeong, H., Dawes, M. C., & Gauthier, D. J. (2006). Direct observation of optical precursors
in a region of anomalous dispersion. Physical Review Letters, 96, 143901.
Loudon, R. (1970). The propagation of electromagnetic energy through an absorbing dielec-
tric. Physica A, 3, 233245.
Marozas, J. A. (1998). Angular spectrum representation of ultrawideband electromagnetic pulse
propagation in lossy, dispersive dielectric slab waveguides. PhD dissertation: University of
Vermont.
Mokole, E. L., & Samaddar, S. N. (1999). Transmission and reflection of normal incident,
pulse electromagnetic plane waves upon a Lorentz half-space. Journal of the Optical Society
of America B: Optical Physics, 16(5), 812831.
Neufeld, J. (1970a). Revised formulation of the macroscopic Maxwell theory. I. Fundamen-
tals of the proposed formulation. II Nuovo Cimento, LXV B(1), 3368.
Neufeld, J. (1970b). Revised formulation of the macroscopic Maxwell theory. II. Propaga-
tion of an electromagnetic disturbance in dispersive media. II Nuovo Cimento, LXVI B(1),
5176.
Olver, F. W. J. (1970). Why steepest descents? SIAM Review, 12(2), 228247.
Oughstun, K. E. (1978). Propagation of optical pulses in dispersive media. PhD dissertation: The
Institute of Optics, University of Rochester.
Oughstun, K. E. (2005). Dynamical evolution of the Brillouin precursor in Rocard-Powles-
Debye model dielectrics. IEEE Transactions on Antennas and Propagation, 53(5),
15821590.
Oughstun, K. E. (2006). Electromagnetic and optical pulse propagation 1. New York: Springer.
Oughstun, K. E. (2009). Electromagnetic and optical pulse propagation 2. New York: Springer.
Oughstun, K. E., Cartwright, N. A., Gauthier, D., & Jeong, H. (2010). Optical precursors in
the singular and weak dispersion limits. Journal of the Optical Society of America B: Optical
Physics, 27(8), 16641670.
Oughstun, K. E., & Sherman, G. C. (1975). Optical pulse propagation in temporally disper-
sive Lorentz media. Journal of the Optical Society of America, 65(10), 1224A.
Oughstun, K. E., & Sherman, G. C. (1980). Comparison of the signal velocity of a pulse with
the energy velocity of a time-harmonic field in Lorentz media. In Proceedings of the URSI
symposium on electromagnetic wave theory, Munich.
Oughstun, K. E., & Sherman, G. C. (1988). Propagation of electromagnetic pulses in a linear
dispersive medium with absorption (the Lorentz medium). Journal of the Optical Society of
America B: Optical Physics, 5(4), 817849.
Oughstun, K. E., & Sherman, G. C. (1989). Uniform asymptotic description of electromag-
netic pulse propagation in a linear dispersive medium with absorption (the Lorentz
medium). Journal of the Optical Society of America. A, 6(9), 13941420.
Oughstun, K. E., & Sherman, G. C. (1994). Pulse propagation in causal dielectrics. Berlin:
Springer-Verlag.
Palombini, C. L., & Oughstun, K. E. (2010). Optical precursor fields in nonlinear pulse
dynamics. Optics Express, 18(22), 2310423120.
Sherman, G. C., & Oughstun, K. E. (1981). Description of pulse dynamics in Lorentz media
in terms of the energy velocity and attenuation of time-harmonic waves. Physical Review
Letters, 47, 14511454.
Sherman, G. C., & Oughstun, K. E. (1995). Energy velocity description of pulse propagation
in absorbing, dispersive dielectrics. Journal of the Optical Society of America B: Optical
Physics, 12, 229247.
Sommerfeld, A. (1914). U ber die fortpflanzung des lichtes in disperdierenden medien. Annals
of Physics, 44, 177202.
Stratton, J. A. (1941). Electromagnetic theory. New York: McGraw-Hill.
Wyns, P., Foty, D. P., & Oughstun, K. E. (1988). Numerical analysis of the precursor fields in
linear dispersive pulse propagation. Journal of the Optical Society of America. A, 6(9),
14211429.
CHAPTER FOUR
Contents
1. Introduction 346
2. Quantum States of Light 351
2.1 Mode Description 351
2.2 Gaussian States 353
2.3 Definite Photon-Number States 358
2.4 Particle Description 358
2.5 Mode Versus Particle Entanglement 359
3. MZ Interferometry 362
3.1 Phase-Sensing Uncertainty 362
3.2 Coherent State Interferometry 364
3.3 Fock State Interferometry 365
3.4 Coherent + Squeezed-Vacuum Interferometry 366
3.5 Definite Photon-Number State Interferometry 367
3.6 Other Interferometers 369
4. Estimation Theory 373
4.1 Classical Parameter Estimation 374
4.2 Quantum Parameter Estimation 382
5. Quantum Limits in Decoherence-Free Interferometry 387
5.1 QFI Approach 388
5.2 Bayesian Approach 390
5.3 Indefinite Photon-Number States 393
5.4 Role of Entanglement 399
5.5 Multipass Protocols 400
6. Quantum Limits in Realistic Interferometry 401
6.1 Decoherence Models 403
6.2 Bounds in the QFI Approach 408
6.3 Bayesian Approach 419
6.4 Practical Schemes Saturating the Bounds 421
7. Conclusions 426
Acknowledgments 427
References 427
1. INTRODUCTION
Without much exaggeration one may say that optics is basically the sci-
ence of light interference. Light interference effects were behind the final
acceptance of classical wave optics and abandoning the Newtonian corpuscu-
lar theory of light. Conceptual insight into the process of interfering light
waves prompted development of a number of measurement techniques
involving well controlled interference effects and gave birth to the field of
optical interferometry (Hariharan, 2003). At the fundamental level, classical
interferometry is all about observing light intensity variations (intensity
fringes) resulting from a change in relative phases between two (or more)
overlapping light waves, e.g., when a single light beam is split into a number
of paths with tunable optical path-length differences and made to interfere on
the screen. The number of applications is stunning and ranges from basic
length measurements via spectroscopic interferometric techniques to the most
spectacular examples involving stellar interferometry and gravitational-wave
detectors (Pitkin, Reid, Rowan, & Hough, 2011).
Coherent properties of light as well as the degree of overlap between the
interfering beams determine the visibility of the observed intensity fringes
and are crucial for the quality of any interferometric measurement. Still,
when asking for fundamental limitations on precision of estimating, e.g.,
a phase difference between the arms of a MachZehnder (MZ) interferom-
eter, there is no particular answer within a purely classical theory, where
both the light itself as well as the detection process are treated classically.
In classical theory, intensity of light can in principle be measured with arbi-
trary precision and as such allows to detect in principle arbitrary small phase
shifts in an interferometric experiment.
This, however, is no longer true when semiclassical theory is considered
in which the light is still treated classically but the detection process is quan-
tized, so that instead of a continuous intensity parameter the number of
energy quanta (photons) absorbed is being measured. The absorption pro-
cess within the semiclassical theory has a stochastic character and the number
of photons detected obeys Poissonian or super-Poissonian statistics (Fox,
2006). If light intensity fluctuations can be neglected, the number of photons
detected, N, follows the Poissonian statistics with photon-number standard
p
deviation N hN i, where hN i denotes the mean number of photons
detected. This implies that the determination of the relative phase difference
Quantum Limits in Optical Interferometry 347
These early results provided a great physical insight into the possibilities
of quantum-enhanced interferometry and the class of states that might be of
practical interest for this purpose. The papers lacked generality, however, by
considering specific measurement estimation strategies, in which the error in
the estimated phase was related via a simple error propagation formula to the
variance of some experimentally accessible observable, e.g., photon-number
difference at the two output ports of the interferometer, or by studying the
width of the peaks in the shape of the phase-measurement probability
distribution.
Given a particular state of light fed into the interferometer, it is a priori
not clear what is the best measurement and estimation strategy yielding the
optimal estimation precision. Luckily, the tools designed to answer these
kinds of questions had already been present in the literature under the name
of quantum estimation theory (Helstrom, 1976; Holevo, 1982). The Quantum
Fisher Information (QFI) as well as the cost of Bayesian inference provide a
systematic way to quantify the ultimate limits on performance of phase-
estimation strategies for a given quantum state, which are already optimized
over all theoretically admissible quantum measurements and estimators. The
concept of the QFI and the Bayesian approach to quantum estimation are
reviewed in Section 4. As a side remark, we should note, that by treating
the phase as an evolution parameter to be estimated and separating explicitly
the measurement operators from the estimator function, quantum estima-
tion theory circumvents some of the mathematical difficulties that arise if
one insists on the standard approach to quantum measurements and attempts
to define the quantum phase operator representing the phase observable being
measured (Barnett & Pegg, 1992; Lynch, 1995; Noh, Fouge`res, & Mandel,
1992; Summy & Pegg, 1990).
The growth of popularity of the QFI in the field of quantum metrology
was triggered by the seminal paper of Braunstein and Caves (1994) advocat-
ing the use of QFI as a natural measure of distance in the space of quan-
tum states. The QFI allows to pin down the optimal probe states that are
the most sensitive to small variations of the estimated parameter by esta-
blishing the fundamental bound on the corresponding parameter sensitivity
valid for arbitrary measurements and estimators. Following these lines of
reasoning the 1/N bound, referred to as the Heisenberg limit, on the phase
estimation precision using N-photon states has been claimed fundamental
and the NOON states were formally proven to saturate it (Bollinger,
Itano, Wineland, & Heinzen, 1996). Due to close mathematical analo-
gies between optical and atomic interferometry (Bollinger et al., 1996;
Quantum Limits in Optical Interferometry 349
Lee, Kok, & Dowling, 2002) similar bounds hold for the problem of atomic
transition-frequency estimation and more generally for any arbitrary unitary
parameter estimation problem, i.e., the one in which an N-particle state
evolves under a unitary UN, U expiH^ , with H^ being a general
single-particle evolution generator and the parameter to be estimated
(Giovannetti, Lloyd, & Maccone, 2004, 2006).
A complementary framework allowing to determine the fundamental
bounds in interferometry is the Bayesian approach, in which one assumes
the estimated parameter to be a random variable itself and explicitly defines
its prior distribution to account for the initial knowledge about before per-
forming the estimation. In the case of interferometry, the typical choice is
the flat prior p() 1/2 which reflects the initial ignorance of the phase.
The search for the optimal estimation strategies within the Bayesian
approach is possible thanks to the general theorem on the optimality of
the covariant measurements in estimation problems satisfying certain group
symmetry (Holevo, 1982). In the case of interferometry, a flat prior guaran-
tees the phase-shift, U(1), symmetry and as a result the optimal measurement
operators can be given explicitly and they coincide with the eigenstates of
the Pegg-Barnett phase operator (Barnett & Pegg, 1992). This makes it
possible to optimize the strategy over the input states and for simple cost
functions allows to find the optimal probe states (Berry & Wiseman,
2000; Buzek, Derka, & Massar, 1999; Luis & Perina, 1996). In particular,
for the 4sin 2 =2 cost function which approximates the variance for small
phase deviations , the corresponding minimal estimation uncertainty has
been found to read /N for large N providing again a proof of the
possibility of achieving the Heisenberg scaling, yet with an additional coef-
ficient. It should be noted that the optimal states in the above approach that
have been found independently in Berry and Wiseman (2000); Luis and
Perina (1996); Summy and Pegg (1990) have completely different structure
to the NOON states which are optimal when QFI is considered as the figure
of merit. This is not that surprising taking into account that the NOON
states suffer from the 2/N ambiguity in retrieving the estimated phase,
and hence are designed only to work in the local estimation approach when
phase fluctuations can be considered small. Derivations of the Heisenberg
bounds for phase interferometry using both the QFI and Bayesian
approaches are reviewed in Section 5. We also discuss the problem of
deriving the bounds for states with indefinite photon number in which case
replacing N in the derived bounds with the mean number of photons hN i
is not always legitimate, so that in some cases the naive Heisenberg
350 Rafal Demkowicz-Dobrza
nski et al.
Koody nski, & Guta, 2012; Escher, de Matos Filho, & Davidovich, 2011;
Demkowicz-Dobrza nski & Maccone, 2014). These tools allow to investi-
gate optimality of estimation strategies for basically any decoherence model
and typically provide the maximum allowable constant factor improvements
p
forbidding better than 1= N asymptotic scaling of precision. Detailed
presentation of the above mentioned results is given in Section 6.
Other approaches to derivation of fundamental metrological bounds
have been advocated recently. Making use of the calculus of variations it
was shown in Knysh, Chen, and Durkin (2014) how to obtain exact formu-
las for the achievable asymptotic precision for some decoherence models,
while in Alipour (2014); Alipour, Mehboudi, and Rezakhani (2014) a var-
iants of QFI have been considered in order to obtain easier to calculate, yet
weaker, bounds on precision. While detailed discussion of these approaches
is beyond the scope of the present review, in Section 6.2.3 we make use of
the result from Knysh et al. (2014) to benchmark the precision bounds
derived in the case of phase diffusion noise model.
The paper concludes with Section 7 with a summary and an outlook on
challenging problems in the theory of quantum-enhanced metrology.
with n fn1 , . .. , nM g and jni jn1 i jnM i representing a Fock state
with exactly ni photons occupying the i-th mode. States jnii may be further
expressed in terms of the respective creation and annihilation operators a^{i , a^i
obeying ^a i , a^{j ij :
a^{n p p
jni i pi j0i, a^jni i ni jni 1i, a^{i jni i ni + 1jni + 1i, (2)
n!
where j0i is the vacuum state with no photons at all.
In the context of optical interferometry, modes are typically taken to be
distinguishable by their spatial separation, corresponding to different arms of
an interferometer, whereas the various optical devices such as mirrors,
beam-splitters or phase-delay elements transform the state on its way
through the interferometer. Eventually, photon numbers are detected in
the output modes allowing to infer the value of the relative phase difference
between the arms of the interferometer.
In many applications, the above standard state representation may not be
convenient and phase-space description is used insteadin particular, the
Wigner function representation (Schleich, 2001; Wigner, 1932). Adopting
the convention in which the quadrature operators read x^i a^{i + a^i and
a {i a^i , the Wigner function may be regarded as a quasi-probability
p^i i^
distribution on the quadrature phase space:
Z
1 M 0 M 0 ip0 xxx
^ 0
pp
^
W x,p d x d p Tr e , (3)
2 2 M
where x fx1 , . . ., xM g, x^ fx^1 , . .. , x^M g and similarly for p and p.
^ As a
consequence, the Wigner function is real, integrates to 1 over the whole
phase space and its marginals yield the correct probability densities of each
of the phase space variables. Yet, since it may take negative values it cannot
be regarded as a proper probability distribution. Most importantly, it may be
reconstructed from experimental data either by tomographic methods
(Schleich, 2001) or by direct probing of the phase space (Banaszek,
Quantum Limits in Optical Interferometry 353
where for a more compact notation we have introduced: the phase space var-
iable z fx1 ,p1 , ..., xM , pM g, the vector containing mean quadrature values
R
hz^i fhx^1 i, hp^1 i, ... , hx^M i, hp^M ig, hz^i i Trz^i d2M z W zzi , and
the 2M dimensional covariance matrix :
1
i, j hz^i z^j + z^j z^i i hz^i ihz^j i: (5)
2
Gaussian states remain Gaussian under arbitrary evolution involving
Hamiltonians at most quadratic in the quadrature operators, which includes
all passive devices such as beam splitters and phase shifters as well as single-
and multimode squeezing operations. Below we focus on a few classes of
Gaussian states highly relevant to interferometry.
These are the states produced by any phase-stabilized laser, what makes them
almost a fundamental tool in the theoretical description of many quantum
optical experiments. Moreover, coherent sates have properties that resemble
features of classical light, and thus enable to establish a bridge between the
quantum and classical descriptions of light.
In the standard representation, an M-mode coherent state
ji j1 i jM i is a tensor product of single-mode coherent states,
whereas a single-mode coherent state is an eigenstate of the respective anni-
hilation operator:
2
X1
n
ji ejj =2 p jni: (8)
n0 n!
From the formula above, it is clear that coherent states do not have a definite
photon number and if a photon number n is measured its distribution follows
2n
the Poissonian statistics Pn ejj jjn! with average hni jj2 and stan-
2
dard deviation n jj. Thus, the relative uncertainty n/hni in the mea-
p
sured photon-number scales like 1= hni and hence in the classical limit of
large hni the beam power may be determined up to arbitrary precision.
Moreover, the evolution of the coherent state amplitude is identical to
the evolution of a classical-wave amplitude. In particular, an optical phase
delay transforms the state ji into j eii. These facts justify a common
jargon of calling coherent states the classical states of light, even though
for relatively small amplitudes different coherent states may be hard to dis-
2
tinguish due to their non-orthogonality jhjij2 ejj . More generally,
we call cl a classical state of light if and only if it can be written as a mixture
of coherent states:
Z
cl d2M Pjihj (9)
Quantum Limits in Optical Interferometry 355
j, ri D ^
^ Srj0i, (10)
^ exp1 r a^2 1 r a^{2 is the squeezing operator, r jrjei is a com-
where Sr 2 2
plex number and jrj and are the squeezing factor and the squeezing angle,
respectively. In fact, any pure Gaussian one-mode state may be written in
the above form. For 0, uncertainties in the quadratures x and p read
x er and p erreduction of noise in one quadrature is accompanied
by an added noise in the other one. This may be conveniently visualized in
the phase-space picture by error disks representing uncertainty in quadra-
tures in different directions, see Figure 1. In such a representation squeezed
states correspond to ellipses while coherent states are represented by circles.
Importantly, the fact that the uncertainty of one of the quadratures can be
less relatively to the other makes it possible to design an interferometric
scheme where the measured photon-number fluctuations are below that
of a coherent state and allows for a sub-shot noise phase estimation precision,
X
1 n
1 Hn 0 tanh r 2 i n
jri p p e 2 jni, (12)
cosh r n0 n! 2
1 X 1
ji2 1n ei n tanh n jn, ni: (14)
cosh n0
A notable feature of the twin-beam state, which may be clearly seen from
Equation (14), is that it is not a product of squeezed states in modes a
and b, but rather it is correlated in between them being a superposition of
terms with the same number of photons in both modes. Its first moments
Quantum Limits in Optical Interferometry 357
Entangled states are defined as all states that are not separable. A crucial fea-
ture of entanglement is that it depends on the division into subsystems. For
example, consider three qubits A, B and C and their joint quantum state
P
ABC 1i, j0 12 jiih jjjiih jjj0ih0j. This state is separable with respect
to the ABjC cut but is entangled with respect to the AjBC cut.
As a first example, consider a two-mode Fock state j1iaj1ib which rep-
resents one photon in mode a and one photon in mode b. This state written
in the mode formalism of Equation (1) is clearly separable. On the other
hand, photons are indistinguishable bosons and if we would like to write
their state in the particle formalism of Equation (18) we have to symmetrize
over all possible permutations of particles, thus obtaining the state
1
j1ia j1ib p jai1 jbi2 + jbi1 jai2 : (22)
2
In this representation, the state is clearly entangled. We may thus say that the
state contains particle entanglement but not the mode entanglement.
If, however, we perform the HongOuMandel experiment and send
the j1iaj1ib state through a balanced beam-splitter which transforms mode
p p
annihilation operators as a^ ! ^ ^
a + b= 2, b^! ^ ^
a b= 2, the resulting
state reads:
Quantum Limits in Optical Interferometry 361
1 1
j1ia j1ib ! p j2ia j0ib j0ia j2ib p jai1 jai2 jbi1 jbi2 , (23)
2 2
which is both mode and particle entangled. Mode entanglement emerges
because the beam-splitter is a joint operation over two modes that intro-
duces correlations between them. On the other hand, it is a local operation
with respect to the particles, i.e., it can be written as U U in the particle
representation, and does not couple photons with each other. Thus, using a
beam-splitter one may change mode entanglement but not the content of
particle entanglement.
As a second example, consider two modes of light, a and b, each of
them in coherent state with the same amplitude , jiajib. This state
clearly has no mode entanglement. Since this state does not have a definite
photon number, in order to ask questions about the particle entanglement
we first need to consider its projection on one of the N-photon
subspacesone can think of a nondemolition total photon-number
measurement yielding result N. After normalizing the projected state we
obtain:
s
1 X
N
N
j N i jia jib N p jnia jN nib , (24)
N
2 n0 n
1 O N
j N i p jaii + jbii (25)
2N i1
and is clearly a separable state. The fact that products of coherent states con-
tain no particle entanglement is in agreement with our definition of classical
state given in Equation (9) being a mixture of products of coherent states.
A classical state according to this definition will contain neither mode nor
particle entanglement.
As a last example, consider the case of particular interest for quantum
interferometry, i.e., a coherent state of mode a and a squeezed vacuum state
of mode b: jiajrib. Again this state has no mode entanglement. On the other
hand it is particle entangled. To see this, consider, e.g., the two-photon
sector, which up to irrelevant normalization factor reads:
jijr iN 2 / 2 j2ia j0ib + tanh r j0ia j2ib 2 jai1 jai2 + tanh r jbi1 jbi2 (26)
362 Rafal Demkowicz-Dobrza
nski et al.
3. MZ INTERFEROMETRY
We begin the discussion of quantum-enhancement effects in
optical interferometry by discussing the paradigmatic model of the MZ
interferometer. We analyze the most popular interferometric schemes
involving the use of coherent and squeezed states of light accompanied
by a basic measurement-estimation procedure, in which the phase is esti-
mated based on the value of the photon-number difference between the
two output ports of the interferometer. Such a protocol provides us with
a benchmark that we may use in the following sections when discussing the
optimality of the interferometry schemes both with respect to the states of
light used as well as the measurements and the estimation procedures
employed.
na
a a
a
b b nb
b
Figure 2 The MachZehnder interferometer, with two input light modes a, b and two
output modes a0 , b0 . In a standard configuration, a coherent state of light ji is sent into
mode a. In order to obtain quantum enhacement, one needs to make use of the b input
port as well, sending, e.g., the squeezed vacuum state jri.
Quantum Limits in Optical Interferometry 363
1 { ^ ^{ i { ^ J^ 1 ^ {
a b + b a^, J^y b^ a^ a^{ b,
J^x ^ z a { a^ b^ b,
^ (28)
2 2 2
which fulfill the angular momentum commutation relations J^i , J^j iEijk J^k
while the corresponding square of the total angular momentum reads:
N^ N^ {
J^ + 1 , N^ a^{ a^ + b^ b,
^
2
(29)
2 2
The optimal operation points are again clearly /2, 3/2, since at them
the first term under the square root, which is nonnegative, vanishes. For a
fair comparison with other strategies, we should fix the total average number
of photons hN i, which is regarded as a resource, and optimize the split of
energy between the coherent and the squeezed vacuum beams in order
to minimize . This optimization can only be done numerically, but
the solution can be well approximated analytically in the regime of hN i
1.
p
In this regime, the squeezed vacuum should carry approximately hN i=2
p
of photons, so the squeezing factor obeys sinh 2 r 14 e2r hN i=2 while
the majority of photons belongs to the coherent beam. The resulting
precision reads
q
p p
hN i
1 hN i=2 hNi + hN i=2 hN i
1 1
jijri p (39)
hNi hNi=2 hN i3=4
and proves that indeed this strategy offers better than shot noise scaling of
precision.
While the above example shows that indeed quantum states of light
may lead to an improved sensitivity, the issue of optimality of the proposed
scheme has not been addressed. In fact, keeping the measurement-
estimation scheme unchanged, it is possible to further reduce the estima-
tion uncertainty by sending a more general two-mode Gaussian states of
light with squeezing present in both input ports and reach the / 1=hN i
scaling of precision (Olivares & Paris, 2007; Yurke et al., 1986). We skip
the details here, as the optimization of the general Gaussian two-mode
input state minimizing the estimation uncertainty of the scheme considered
is cumbersome. More importantly, using the tools of estimation theory
introduced in Section 4, we will later show in Section 5 that even with
jijri class of input states it is possible to reach the / 1=hN i scaling of
precision, but this requires a significantly modified measurement-
estimation scheme (Pezze & Smerzi, 2008) and different energy partition
between the two input modes.
the twin-Fock state where N/2 photons are simultaneously sent into each of
the input ports, is also capable of providing the Heisenberg scaling of pre-
cision (Holland & Burnett, 1993), but requires a different measurement-
estimation scheme which goes beyond the analysis of the average
photon-number difference at the output, see Section 5.3.3.
One should bear in mind that the expressions for attainable precision
presented in this section are based on a simple error propagation formula cal-
culated at a particular operating point. Therefore, in order to approach any
of the precisions claimed, one needs to first lock the interferometer to oper-
ate close to an optimal point, which also requires some of the resources to be
consumed. Rigorous quantification of the total resources needed to attain a
given estimation precision starting with a completely unknown phase may
be difficult in general. We return to this issue in Section 5, where we are able
to resolve this problem by approaching it with the language of Bayesian
inference.
from one mirror to the other inside the interferometer. The resulting
inputoutput relation reads:
0
a^ 1
^0
b 2 T i sin T cos
p
2i 1 T sin p
T a^
(45)
T 2i 1 T sin b^
and up to an irrelevant global phase may be rewritten as:
0
a^ cos =2 i sin =2 a^
,
b^0 i sin =2 cos =2 b^ ! (46)
T
2 arcsin p :
T 2 + 41 T sin 2
T 2 + 41 T sin 2
p :
@ 4T 1 T cos (47)
@
As a consequence, the above expression allows to translate all the results
derived for the MZ interferometer to the FabryPerot case.
Quantum Limits in Optical Interferometry 371
Figure 4 Formal equivalence of Ramsey and MZ interferometry. The two atomic levels
which are used in the Ramsey interferometry play analogous roles as the two arms of
the interferometer, while the /2 pulses equivalent from a mathematical point of view
to the action of the beam-splitters. Quantum-enhancement in Ramsey interferometry
may be obtained by preparing the atoms in a spin-squeezed input state reducing
the variances of the relevant total angular momentum operators, in a similar fashion
as using squeezed states of light leads to an improved sensitivity in the MZ
interferometry.
: (49)
t
Beating the projection noise requires the input state of the atoms to share
some particle entanglement. From an experimental point of view, the most
promising class of states are the so-called one-axis or two-axis spin-squeezed
states, which may be realized in BEC and atomic systems interacting with
light (Kitagawa & Ueda, 1993; Ma, Wang, Sun, & Nori, 2011). In fact, these
states may be regarded as a definite particle-number analogues of optical
squeezed states. In particular, starting with atoms in a ground state the
two-axis spin-squeezed states may be obtained via:
^2 ^2
j i e 2 J + J jgiN , (50)
where J^ J^x iJ^y . The above formula resembles the definition of an opti-
cal squeezed state given in Equation (10), where a^, a^{ operators are replaced
by J^ , J^+ . With an appropriate choice of squeezing strength as a function
Quantum Limits in Optical Interferometry 373
4. ESTIMATION THEORY
In this section, we review the basics of both classical and quantum
estimation theory. We present Fisher Information (FI) and Bayesian
374 Rafal Demkowicz-Dobrza
nski et al.
4.1.1 FI Approach
In this approach, p(X) is regarded as a family of PDFs parametrized by
the parameter to be estimated based on the registered data x. The per-
formance of a given estimator N x is quantified by the Mean Square Error
(MSE) deviation from the true value :
D Z
2 2 E 2
N N x dN x p x N x : (51)
Quantum Limits in Optical Interferometry 375
so that on average it yields the true parameter value. The optimal unbiased
estimator is the one that minimizes 2N for all . Looking for the optimal
estimator may be difficult and it may even be the case that there is no single
estimator that minimizes the MSE for all .
Still, one may always construct the so-called Cramer-Rao Bound (CRB)
that lower-bounds the MSE of any unbiased estimator N (see, e.g., Kay,
1993 for a review):
1
2 N , (53)
N F p
where F is the FI, and can be expressed using one of the formulas below:
Z * 2 + 2
1 @ p x 2 @ @
F p dx ln p ln p : (54)
p x @ @ @2
The basic intuition is that the bigger the FI is the higher estimation precision
may be expected. The FI is nonnegative and additive for uncorrelated events,
h i h i h i
so that F p1
, 2
F p 1
+ F p2
, for p
1, 2
x1 , x2 p1 2
x1 p x2
h i
and in particular: F pN N F p , which can be easily verified using
the last expression in definition (54). The FI is straightforward to calculate
and once an estimator is found that saturates the CRB it is guaranteed to
be optimal. In general, estimators saturating the CRB are called efficient.
The sufficient and necessary condition for efficiency is the following condition
on the PDF and the estimator (Kay, 1993):
@
ln p x N F p k x : (55)
@
An estimator satisfying the above equality exist only for a special class of
PDFs belonging to the so called exponential family of PDFs, for which:
a0
lnp x a + bx + cdx, , (56)
c 0
where a(), c() and b(x), d(x) are arbitrary functions and primes denote dif-
ferentiation over . In general, however, the saturability condition cannot
be met.
376 Rafal Demkowicz-Dobrza
nski et al.
which, in contrast to Equation (51), is also averaged over all the values of the
parameter with the Bayesian prior p(). Making use of the Bayes theorem
we can rewrite the above expression in the form
Z Z
2 2
N dx px d pjx N x : (61)
From the above formula, it is clear that the optimal estimator is the one that
minimizes terms in square bracket for each x. Hence, we can explicitly
derive the form of the Minimum Mean Squared Error (MMSE) estimator
Z
@ 2
d pjx N x 0
@ N Z
MMSE
) N x d pjx hipjx (62)
the knowledge we possess about the parameter after inferring the informa-
tion about it from the sampled data x. The corresponding MMSE reads:
Z Z 2
2
N dx px d pjx hipjx
Z (63)
2
dx px pjx ,
X
1
C, C cn cos n with : (65)
n0
Quantum Limits in Optical Interferometry 379
n1 n1
unfair coin, which heads and tails occurring with probabilities p and
q 1 p, respectively.
Probability that n out of N photons get transmitted is governed by the
binomial distribution
N
pp n
N
pn 1 pN n : (69)
n
p1 p
2 p N : (70)
N
Luckily the binomial probability distribution belongs to the exponential
family of PDFs specified in Equation (56), and by inspecting saturability con-
dition Equation (55) it may be easily checked, that the simple estimator
p N n n=N saturates the CRB. It is also worth mentioning that the optimal
estimator also coincides with the ML estimator, hence in this case the ML
estimator is optimal also for finite N and not only in the asymptotic regime.
In the context of optical interferometry, we will deal with an analogous
situation, where photons are sent into one input port of an interferometer
and p, q correspond to the probabilities of detecting a photon in one of the
two output ports. For a MZ interferometer, see Section 3, the probabilities
depend on the relative phase delay difference between the arms of an inter-
ferometer : p sin 2 =2, q cos 2 =2, which is the actual parameter
of interest. Probability distribution as a function of then reads:
N
p n
N
sin=22n cos=22N n (71)
n
and the corresponding FI and the CRB take the form
h i 1
2
F pN N , N : (72)
N
Interestingly, FI does not depend on the actual value , what suggests that
the achievable estimation precision may be independent of the actual param-
eter value. However, for such a parametrization, the CRB saturability
condition Equation (55) does not hold and there is no unbiased estimator
saturating the bound. Nevertheless, using Equation (58), we may still
write a locally unbiased estimator saturating CRB at 0: N n
0 tan 0 =2 + 2n=N sin 0 , which is possible provided sin0 6 0.
Quantum Limits in Optical Interferometry 381
Since the CRB given in Equation (72) can only be saturated locally it is
worth looking at the ML estimator which we know will perform optimally
in the asymptotic regime N ! 1. Solving Equation (59) we obtain
r
ML n
N n argmax lnp n 2 arctan
N
(73)
N n
and in general we notice that there are two equivalent maxima. This ambi-
guity is simply the result of invariance of the PDF p(n) with respect to the
change ! and might have been expected. Hence in practice we would
need some additional information, possibly coming from prior knowledge or
other observations, in order to distinguish between this two cases and be enti-
tled to claim that the ML estimator saturates the CRB asymptotically for all .
We now analyze the same estimation problem employing the Bayesian
approach. Let us first consider the case where p is the parameter to be esti-
mated and the relevant conditional probability pN njp pN p n is given by
Equation (69). Choosing a flat prior distribution p(p) 1 and the mean
square as a cost function, we find using Equations (62) and (63) that the
MMSE estimator and the corresponding minimal averaged MSE equal:
n + 1 D 2 E 1
p N n , pN , (74)
N +2 6N + 2
which may be compared with the previously discussed FI approach where
the optimal estimator was p n n=N and the resulting variance when
R1
averaged over all p would yield h2 p N i 0 dp pp1 N 1=6N . Hence,
in the limit N ! 1 these results converge to the ones obtained previously
in the FI approach. This is a typical situation that in the case of large amount
of data the two approaches yield equivalent results (van der Vaart, 2000).
We now switch to parametrization and consider pN nj pN n as in
Equation (71). Assuming flat prior distribution p() 1/(2) and the pre-
viously introduced natural cost function in the case of circular parameter
C , 4sin 2 =2, due to estimation ambiguity we realize
that the condition for the optimal estimator given in Equation (68) is satisfied
C
for a trivial estimator N n 0 which does not take into account the
measurement results at all. This can be understood once we realize that
the ambiguity in the sign of estimated phase and the possibility of esti-
mating the phase with the wrong sign is worse than not taking into account
the measured data at all. In order to obtain a more interesting result, we
need to consider a subset of possible values of over which the
382 Rafal Demkowicz-Dobrza
nski et al.
where FQ is the QFI while the Hermitian operator L[] is the so-called
Symmetric Logarithmic Derivative (SLD), which can be unambiguously defined
for any state via the relation _ 12 L + L . Crucially,
QFI is solely determined by the dependence of on the estimated param-
eter, and hence allows to analyze parameter sensitivity of given probe states
without considering any particular measurements norP estimators. Explicitly,
the SLD when written in the eigenbasis of i i jei i hei j
reads:
X 2hei j_ ej
L jei ej , (78)
i, j
i + j
where the sum is taken over the terms with nonvanishing denominator.
Analogously to the FI (54), the QFI is an additive quantity when calculated
h i
on product states and in particular FQ N N FQ . Thus the N
term in the denominator of Equation (77) may be equivalently interpreted
as the number of independent repetitions of an experiment with a state to
form the data sample x of size N, or a single shot experiment with a multi-
party state N
.
Crucially, as proven by Braunstein and Caves (1994) and Nagaoka
(2005), there always exist a measurement strategya projection measure-
ment in the eigenbasis of the SLDfor which the FI calculated for the
resulting probability distribution equals the QFI, and consequently the
1
For clarity of notation, in what follows we drop the tilde symbol when writing estimator variance.
384 Rafal Demkowicz-Dobrza
nski et al.
bounds (53) and (77) coincide. Hence the issue of saturability of the QCRB
amounts to the problem of saturability of the corresponding classical CRB.
As discussed in detail in Section 4.1.1, the bound is therefore globally satu-
rable for a special class of probability distributions belonging to the so called
exponential family, and if this is not the case the saturability is achievable
either in the asymptotic limit of many independent experiments N ! 1
or in the local approach when one estimates small fluctuation of the param-
eter in the vicinity of a known
value 0.
For pure states, , the QFI in Equation (77) simplifies to
2 dj i
FQ 4 _ j_ _ j , j_ i : (79)
d
Yet, for general mixed states, calculation of the QFI involves diagonalization
of the quantum state , in order to calculate the SLD, and becomes tedious
for probe states living in highly dimensional Hilbert spaces.
Interestingly, the QFI may be alternatively calculated by considering
purifications ji of a given family of mixed states on an extended Hilbert
space TrE , where by E we denote an ancillary space
needed for the purification. It has been proven by Escher et al. (2011) that
the QFI of any is equal to the smallest QFI of its purifications
n
2 o
FQ min FQ 4 min _ j_ _ j : (80)
Even though minimization over all purification may still be challenging, the
above formulation may easily be employed in deriving upper bounds on
QFI by considering some class of purifications. Since upper bounds on
QFI translate to lower bounds on estimation uncertainty this approach
proved useful in deriving bounds in quantum metrology in presence of
decoherence (Escher, Davidovich, Zagury, & de Matos Filho, 2012;
Escher et al., 2011).
Independently, in Fujiwara and Imai (2008) another purification-based
QFI definition has been constructed:
FQ 4 min _ j_ : (81)
Despite apparent difference, Equations (80) and (81) are equivalent and one
can prove that any purification minimizing one of them is likewise optimal
for the other causing the second term of Equation (80) to vanish. Although
for any suboptimal jiEquation (80) must provide a strictly tighter bound
Quantum Limits in Optical Interferometry 385
on QFI than Equation (81), the latter definition, owing to its elegant form
allows for a direct and efficient procedure for derivation of the precision
bounds in quantum metrology (Demkowicz-Dobrza nski et al., 2012;
Koody nski & Demkowicz-Dobrza nski, 2013). Derivations of the precision
bounds using the above two techniques in the context of optical interferom-
etry are discussed in Section 6.
For completeness, we list below some other important properties of
the QFI. QFI does not increase under a parameter-independent quantum
channel
FQ FQ , (82)
where is an arbitrary completely positive map. QFI appears in the lowest
order expansion of the measure of fidelity F between two quantum states
p2
p p
F 1 ,2 Tr 1 2 1 , (83)
1
F , + d 1 FQ d2 + Od4 : (84)
4
QFI is convex
!
X X X
FQ pi i
pi FQ i
, pi 1, pi 0, (85)
i i i
which reflects the fact that mixing quantum states can only reduce achiev-
able estimation sensitivity. In a commonly encountered case, specifically in
the context of optical interferometry, when the estimated parameter is
encoded on the state by a unitary
U U{ , U eiH^ , (86)
where jeii, i form the eigendecomposition of . Note that in this case the
QFI does not depend on the actual value of . For the pure state estimation
case, jihj, the QFI is proportional to the variance of H^ :
FQ j i 42 H 4hjH^ ji hjH^ ji
2
(88)
386 Rafal Demkowicz-Dobrza
nski et al.
and the QCRB (77) takes a particular appealing form resembling the form of
the energy-time uncertainty relation:
1
2 2 H : (89)
4N
We conclude the discussion of QFI properties by mentioning a very recent
and elegant result proving that for unitary parameter encodings, QFI is pro-
portional to the convex roof of the variance of H^ (Toth & Petz, 2013; Yu,
2013):
X
FQ 4 min pi 2 H i , (90)
fpi , j i ig
i
and leaves us with the problem of minimization hCi over a general POVM
R
M with standard constraints M 0, d
2 M 1. Note that for clarity we
M U U { , (92)
Let us first consider the situation in which jiin is a state resulting from
sending N photons on the balanced beam-splitter, as discussed in
Section 3.3. This time, however, we do not insist on a particular measure-
ment nor estimation scheme, but just want to calculate the corresponding
QCRB on the sensitivity. Written in the photon-number basis the state
takes the form:
s
XN
1 N
jiin jnijN ni, (97)
n0
2N n
for which FQ N and this results in the shot noise bound on precision:
1
jN ij0i p : (98)
N
Recall, that this bound is saturated with the simple MZ interferometric
scheme discussed in Equation (36), which is a proof that for the considered
probe state this measurement-estimation scheme is optimal. From a particle-
description point of view, see Section 2.4, the above considered state is a
pure product state with no entanglement between the photons. More gen-
erally, the shot noise limit, sometimes referred to as the standard quantum
limit, is valid for all N-photon separable states (Giovannetti, Lloyd, &
390 Rafal Demkowicz-Dobrza
nski et al.
Maccone, 2011; Pezze & Smerzi, 2009), and going beyond this limit
requires making use of inter-photon entanglement, see Section 5.4.
Let us now investigate general input states, which possibly may be
entangled. Consider the state of the form jNOONi p12 jN ij0i + j0ijN i
which is commonly referred to as the NOON state (Bollinger et al.,
1996; Lee et al., 2002). QFI for such state is given by FQ N2 and
consequently
1
, (99)
N
which is referred to as the Heisenberg limit. In fact, NOON state gives the best
possible precision as it has the biggest variance of J^z among the states with a
given photon number N (Bollinger et al., 1996; Giovannetti et al., 2006).
Still, the practical usefulness of the NOON states is doubtful. The difficulty
in their preparation increases dramatically with increasing N, and with pre-
sent technology the experiments are limited to relatively small N, e.g., N 4
(Nagata, Okamoto, OBrien, Sasaki, & Takeuchi, 2007) or N 5 (Afek,
Ambar, & Silberberg, 2010). Moreover, even if prepared, their extreme sus-
ceptibility to decoherence with increasing N, see Section 6, makes them
hardly useful in any realistic scenario unless N is restricted to small values.
Taking into account that experimentally accessible squeezed states of light
offer a comparable performance in the decoherence-free scenario, see
Section 5.3, and basically optimal asymptotic performance in the presence
of decoherence, see Section 6, there is not much in favor of the NOON
states apart from their conceptual appeal.
!
X
N
hC i 2 2Re cn cn1 n, n1 : (100)
n1
R
Because is Hermitian, the completeness condition d ^ ^{
2 U U 1
implies that nn 1, while due to the positive semidefiniteness condi-
p
tion 0, jnm j nn mm 1. Therefore, the real part in the subtracted
P
term in Equation (100) can at most be N n1 jcn jjcn1 j, which will be the
case for n,m ei(nm), where n arg cn . This is a legitimate positive
semidefinite operator, as it can be written as jeNiheNj, with
P
jeN i N n0 e jn,N ni (Chiribella et al., 2005; Holevo, 1982). Thus,
in
For the uncorrelated input state (97), the average cost reads:
s
!
X 1
1 N N N N !1 1
hC i 2 1 N : (102)
2 n0 n n+1 N
Since in the limit of small estimation uncertainty the considered cost func-
tion approaches the MSE, we may conclude that in the limit of large N:
N !1 1
p , (103)
N
which coincides with the standard shot-noise limit derived within the QFI
approach.
Note a subtle difference between the above solution and the solution of
the optimal Bayesian transmission coefficient estimation problem discussed
in Section 4.1.3 with the parametrization employed. The formulas for
probabilities in Section 4.1.3 can be regarded as arising from measuring each
of uncorrelated photons leaving the interferometer independently, while in
the present considerations we have allowed for arbitrary quantum measure-
ments, which are in general collective. Importantly, we account for the
adaptive protocols in which a measurement on a subsequent photon depends
on the results obtained previously (Koody nski & Demkowicz-Dobrza nski,
2010)practically these are usually additional controlled phase shifts
392 Rafal Demkowicz-Dobrza
nski et al.
allowing to keep the setup at the optimal operation point (Higgins, Berry,
Bartlett, Wiseman, & Pryde, 2007). This approach is therefore more general
and in particular, does not suffer from a ambiguity that forced us to
restrict the estimated region to [0, ) in order to obtain nontrivial results
given in Equation (75).
Let us now look for the optimal input states. From Equation (102) it is
clear that we may restrict ourselves to real cn. Denoting by c the vector con-
taining the state coefficients cn, we rewrite the formula for hC i in a more
appealing form
More formally, let jiar jiajir, be the original coherent state used for
sensing the phase, accompanied by a coherent reference beam ji. The
corresponding output state reads j iar jeiiajir. Now, the phase
plays the role of the relative phase shift between the two modes, with a clear
physical interpretation. The combined phase shift in the two modes, i.e.,
an operation Ua Ur ein^a + n^r , has no physical significance as it is not
detectable without an additional reference beam. Hence, before calculating
the QFI or any other quantity determining fundamental precision bounds,
one should first average the state jiar over the combined phase shift and
treat the resulting density matrix as the input probe state
Z 2 X 1
d a r
U U jiar hjjihjUa{ Ur{ pN N : (109)
0 2 N 0
This operation destroys all the coherences between sectors with different
total photon number, and the resulting state is a mixture of states N with
different total photon numbers N, appearing with probabilities pN.
In the absence of a reference beam, i.e., when 0, the above averaging
kills all the coherences between terms with different photon numbers in the
mode a:
Z X1
d i i jj2 jj2n
je ihe jj0ih0j e jnihnjj0ih0j (110)
2 n0
n!
This, however, implies that knowing the solution for the optimal definite
photon-number probe states, by adjusting the probabilities pN with which
different optimal N are being sent, we may determine the optimal strategies
with the average photon-number fixed.
396 Rafal Demkowicz-Dobrza
nski et al.
Taking the QFI approach for a moment, we recall that the optimal
N-photon state, the NOON state, yields FQ(N) N2. Let us consider a
strategy where a vacuum state and the NOON state are sent with probabil-
ities 1 p and p, respectively, with the constraint on the average photon
number pN hN i. The corresponding QFI reads FQ (1 p) 0 + pN2.
Substituting p hN i/N we get FQ NhN i. Therefore, while keeping hN i
fixed we may increase N arbitrarily and in principle reach FQ 1,
suggesting the possibility of arbitrary good sensing precisions (Rivas &
Luis, 2012; Zhang, Jin, Cao, Liu, & Fan, 2013). Note in particular, that
a naive generalization of the Heisenberg limit to hN1 i, does not hold,
and the strategies beating this bound are referred to as sub-Heisenberg
strategies (Anisimov et al., 2010). A universally valid bound may be
q
written as 1= hN^ i (Hofmann, 2009), but the question remains,
2
Z 2 !
2 d2 dW z
jh j + d ij 1 242 d 2M
z (114)
4 d
Z ! 1
dW z 2 2M 2
242 d z : (115)
d
In terms of the covariance matrix and the first moments hzi, the formula
takes an explicit form:
2 !! 12
dhziT 1 dhzi 1 d 1
+ Tr : (116)
d d 4 d
Formal optimization of the above equation was done by Pinel et al. (2012),
however, the result was a one-mode squeezed-vacuum state, which in order
to carry phase information needs to be assisted by a reference beam. Unfortu-
nately, performing a common phase-averaging procedure described in
Section 5.3.1 in order to calculate the precision in the absence of additional
phase reference destroys the Gaussian structure of the state and makes the opti-
mization intractable. Luckily, for path symmetric states, i.e., the states invariant
under the exchange of interferometer arms, the phase averaging procedure does
not affect the QFI ( Jarzyna & Demkowicz-Dobrza nski, 2012). Hence, assum-
ing the path-symmetry the optimal Gaussian state is given by jrijri with
sinh 2 r hNi=2two squeezed vacuums send into the input ports of the
interferometerand its corresponding QFI leads to the QCRB
1 1
p (117)
hN ihN i + 2 hN i
The state also achieves the Heisenberg limit for a large number of photons in
the setup but it does not require any external phase reference. It is also worth
noting that this state, while being mode-separable is particle entangled and is
feasible to prepare with current technology for moderate squeezing
strengths. However, the enhancement over optimal squeezedcoherent
strategy is rather small and vanish for large number of photons. Precisions
obtained in squeezedcoherent and squeezedsqueezed scenarios are
depicted in Figure 6.
One can also study Gaussian states within the Bayesian framework.
Optimal seed operator can be easily generalized from the definite
P
photon-number case to 1 N 0 jeN iheN j. Conceptually, the whole treat-
ment is the same as in the definite photon-number case. However, the
expressions and calculations are very involved and will not be
presented here.
Quantum Limits in Optical Interferometry 399
1
N
1
N
N
Figure 6 Limits on precision obtained within QFI approach when using two optimally
squeezed states in both modes jrijri (black, solid), coherent and squeezed-vacuum
states jijri (black, dashed). For comparison, precision achievable with simple coherent
and squeezed vacuum MZ interferometric scheme discussed in Section 3.4 is also
depicted (gray, dashed).
where max denotes state from the set fi gi1, ..., N for which QFI takes the
largest value. But for a one photon state, the maximum value of QFI is equal
to 1, so
1
FQ N, p : (119)
N
P i i
For general separable states i pi 1 N , it is sufficient to use
the convexity of QFI, together with Equation (119) to obtain the same con-
clusion. Above results imply that QFI, or precision, can be interpreted as a
particle-entanglement witness, i.e., all states that give precision scaling better
than the shot noise must be particle entangled (Hyllus et al., 2012; Toth, 2012).
The seemingly unentangled state jijri when projected on the definite
total photon-number sector, indeed contains particle entanglement as was
demonstrated in Section 2.5. This fact should be viewed as the fundamental
source of its ability for performing quantum-enhanced sensing. It is also
worth stressing, that unlike mode entanglement, particle entanglement is
invariant under passive optical transformation like beam splitters, delay lines
and mirrors, which makes it a sensible quantity to be treated as a resource for
quantum-enhanced interferometry.
evolves into j i p12 j01i + eiN j10i. The phase is acquired N times
faster compared with a single pass case, mimicking the behavior of a single
pass experiment with a NOON state. Hence, the precision may in prin-
ciple be improved by a factor of N. Treating the number of single photon
passes as a resource, it has been demonstrated experimentally (Higgins
et al., 2007) that in the absence of noise such a device can indeed achieve
the Heisenberg scaling without resorting to entanglement and efficiency of
various multipass protocols has been analyzed in detail in Berry et al.
(2009). This is not to say, that all quantum strategies are formally equivalent
to single-photon multipass strategies. As will be discussed in the next
section, the NOON states are extremely susceptible to decoherence, in
particular loss, and this property is shared by the multipass strategies. Other
quantum strategies prove more advantageous in this case, and they do not
have a simple multipass equivalent (Demkowicz-Dobrza nski, 2010;
Kaftal & Demkowicz-Dobrza nski, 2014).
Figure 8 Schematic description of the decoherence processes discussed that affect the
performance of an optical interferometer: (i) phase diffusion representing stochastic fluc-
tuations of the estimated phase delay, (ii) losses in the respective a/b arms represented
by fictitious beam splitters with 0 a/b 1 transmission coefficients, and (iii) imperfect
visibility indicated by a mode mismatch of the beams interfering at the output beam
splitter.
X
N
in jiin hj, jiin cn jnijN ni: (120)
n0
Figure 9 General metrological scheme in case of photons being treated as formally dis-
tinguishable particles. Each photon travels through a phase encoding transformation
U. Apart from that all photons are subject to either correlated (i) (phase diffusion),
or uncorrelated (ii), (iii) (loss, imperfect visibility) decoherence process.
N {
in , U U : (121)
where cn are parameters of the input state given as in (120). The above equa-
tion indicates that due to the phase diffusion the off-diagonal elements of
are exponentially suppressed at a rate increasing in the antidiagonal
directions.
Spagnolo, Toffoli, Sciarrino, & De Martini, 2010), and most notably, when
analyzing bounds on quantum enhancement in gravitational-wave detectors
(Demkowicz-Dobrza nski et al., 2013).
Loss decoherence map may be formally described using the following
set of Kraus operators (Dorner et al., 2009):
s s
1 a a^{ a^ la 1 b lb b^{ b^ ^lb
la
Kla , lb a a^ b b (126)
la ! lb !
where the values of index la/b corresponds to the number of photons lost in
mode a/b, respectively. For a general N-photon input state of the form
(120), the density matrix representing the output state of the lossy interfer-
ometer reads
M
N NX
N 0
U in U{ pla , lb la , lb la , lb ,
(127)
N 0 0 la 0
lb N N 0 la
P 2 la , lb
where pla , lb N n0 jcn j bn is the binomially distributed probability of
losing la and lb photons in arms a and b, respectively, with
la , lb n nla la N n N nlb
bn 1 a b 1 b lb , (128)
la a lb
while the corresponding conditional pure states read:
q
NXlb
cn ein
l , l p
l , l
bn a b jn la ,N n lb i: (129)
a b
nla
pla , lb
The direct sum in Equation (127) indicates that the output states of different
total number of surviving photons, N 0 , belong to orthogonal subspaces,
which in principle could be distinguished by a nondemolition, photon-
number counting measurement.
In the particle-approach when photons are considered as formally distin-
guishable particles, the loss process acts on each of the photons indepen-
dently, see Figure 9, so that the overall decoherence process has a tensor
product structure N, with being a single particle loss transformation.
At the input stage, each photon occupies a two-dimensional Hilbert space
spanned by vectors jai, jbi representing the photon traveling in the mode
a/b, respectively. In order to describe loss, however, and formally keep
406 Rafal Demkowicz-Dobrza
nski et al.
Note that similarly to the loss model we have modeled the noise with the
use of fictitious beam-splitters to visualize the effects of decoherence. Now,
as we know that a beam-splitter acts on the photons contained in its two-
mode input state in an uncorrelated manner, the effective map on the full
N-photon input state is N. In case of atomic systems, this would be a typ-
ical local dephasing model describing uncorrelated loss of coherence
between the two relevant atomic levels (Huelga et al., 1997). Still, there
is a substantial difference from the loss models as the dephased photons
are assumed to remain within the spatially confined beams of the
interferometer arms.
We can relate the two models by a simple observation, namely that if
the photons lost in the loss model with a b were incoherently
injected back into the arms of the interferometer, we would recover the
local dephasing model with the corresponding parameter . It should
therefore come as no surprise, when we derive bounds on precisions for
the two models in Sections 6.2.1 and 6.2.2, that for the same the
local dephasing (imperfect visibility) model implies more stringent bounds
on achievable precision than the loss model. Intuitively, it is better to get
rid of the photons that lost their coherence and do not carry information
about the phase, rather than to inject them back into the setup.
408 Rafal Demkowicz-Dobrza
nski et al.
The structure of the output state is more complex than in the case of
phase diffusion and loss models. This is because the local dephasing noise not
only transforms the input state into a mixed state, but due to tracing out
some degrees of freedom, the output state
X
2
N
in Ki1 KiN UN in UN { Ki{1 Ki{N (135)
i1 , ..., iN 1
s
1 2 1
p , (136)
2 N
where is the dephasing parameter, see Section 6.1.3. For the optimal
p p
uncorrelated input state, j in i jai + jbi= 2N we get 1= 2 N ,
and hence the quantum precision enhancement which is the ratio of the
bound on precision achievable for the optimal strategy and the precision
p
for the uncorrelated strategy is bounded by a constant factor of 1 2 .
CS Method
The derivation of the formula (136) presented below makes use of the CS
method (Demkowicz-Dobrza nski et al., 2012), which requires viewing the
quantum channel representing the action of the interferometer from a geo-
metrical perspective. The set of all physical quantum channels,
: LHin ! LHout , that map between density matrices defined on
the input/ouput Hilbert spaces (Hin=out ) constitutes a convex set
(Bengtsson & Zyczkowski, 2006). This is to say that given any two quantum
channels 1, 2, their convex combination p1 + (1 p)2, 0 p 1
is also a legitimate quantum channel. Physically corresponds then to a
quantum evolution that is equivalent to a random application of 1, 2
transformations with probabilities p, 1 p, respectively.
As derived in Section 6.1.3, within the imperfect visibility (local
dephasing) decoherence model: N(in), and hence the relevant
quantum channel, has a simple tensor structure. Consider a single-photon
channel , which -dependence we may depict as a trajectory within
the set of all single-photon quantum maps, see Figure 10. The question
of sensing the parameter can now be translated to the question of deter-
mining where on the trajectory a given quantum channel lies.
Consider a local CS of a quantum channel trajectory in the vicinity of
a given point 0, 0 + (Demkowicz-Dobrza nski et al., 2012;
Matsumoto, 2010),
X
p x x + O 2 , (137)
x
Figure 10 The space of all quantum channels, , which map between density matrices
specified on two given Hilbert spaces, : LHin ! LHout , represented as a con-
vexset. The estimated parameter specifies a trajectory, (black curve), in such
a space. From the point of view of the QFI, any two channel trajectories, e.g., and
(gray curve), are equivalent at a given 0 as long as they and their first derivatives
with respect to coincide there. Moreover, they can be optimally classically simulated at
0 by mixing two channels lying on the intersection of the tangent to the trajectory and
the boundary of the set: .
parameter directly from XN, i.e., ! X N ! , can perform only better
than the scheme where the information about is firstly encoded into the
quantum channel which acts on the probe state and afterwards decoded from
the measurement results performed on the output state. This way, we may
always construct a classically scaling lower bound on the precision, or equiv-
alently an upper bound on the QFI of (135):
h i
FQ Fcl pN N Fcl p , (139)
which is determined by the classical FI (54) evaluated for the probability dis-
tribution p(X). Importantly, Demkowicz-Dobrza nski et al. (2012) have
shown that in this particular case thanks to the parameter being unitary
encoded, it is always optimal to choose a CS depicted in Figure 10, which
employs for each only two channels lying at the points of intersection
of the tangent to the trajectory with the boundary of the quantum maps set.
Such an optimal CS leads to the tightest upper bound specified in
Equation (139): FQ N = + , where are the distances to
d
the boundary marked in Figure 10, 0 d j0 .
Looking for parameters amounts to a search of the distances one can
go along the tangent line to the trajectory of such that the corresponding
map is still a physical quantum channel, i.e., a completely positive trace
preserving map. This is easiest to do making use of the ChoiJamiolkowski
isomorphism (Choi, 1975; Jamiokowski, 1972) which states that with
each quantum channel, : LHin ! LHout , we can Nassociate a positive
operator 2 L(Hout Hin ), so that ( I )(jIihIj), where
PdimHin
jIi i1 jiijii is a maximally entangled state on Hin Hin , while I is
the identity map on L(Hin ). Since the complete positivity of is equivalent
d
to positivity of the operator, one needs to analyze 0 d j0 and
find maximum so that the above operator is still positive-semidefinite.
Taking the explicit form of the for the case of optical interferometry
with imperfect visibility, see Section 6.1.3, one can show that
p
1 2 = (Demkowicz-Dobrza nski et al., 2012), which yields the
ultimate quantum limit on precision given by Equation (136).
QS Method
Unfortunately, in the case of loss the simple CS method yields a trivial bound
0, since the tangent distances to the boundary of the set of quantum
channels are 0 in this case. It is possible, however, to derive a useful
bound via the QS method which is a natural generalization of the CS
method. The QS method has been described in detail and developed for
general metrological schemes with uncorrelated noise by Koody nski
and Demkowicz-Dobrza nski (2013) stemming from the works of
Demkowicz-Dobrza nski et al. (2012) and Matsumoto (2010).
As shown in Figure 11, local QS amounts to reexpressing the action of
for 0 + by a larger -independent map that also acts on the
auxiliary -dependent input , up to the first order in :
TrE + O2 : (142)
P
Note that for x p xjxi hxj, and jxihxj x, QS becomes
equivalent to the CS of Equation (137), so that CS is indeed a specific
instance of the more general QS. An analogous reasoning as in the case
of CS leads to the conclusion that we may upper-bound the QFI of the over-
all output state, here (127) for the case of losses, as
414 Rafal Demkowicz-Dobrza
nski et al.
Figure 11 The quantum simulation (QS) of a channel. The action of the channel is
simulated up to the first order in the vicinity of a given point 0 using a -independent
channel and an auxiliary state that contains the full information about the esti-
mated parameter .
h i h n h ioi h i
FQ N
in F Q Tr E N
N
in F Q N
N FQ ,
since TrE N[] is just a parameter independent map, under which the over-
all QFI may only decreasesee Equation (82). Last equality follows from the
additivity property of the QFI, which, similarly to Equation (139), constrains
h i
FQ N to scale at most linearly for large N. Similarly to the case of CS, in
order to get the tightest bound one should find QS that yields the smallest
FQ[ ], which in principle is a nontrivial task.
Fortunately, Koody nski and Demkowicz-Dobrza nski (2013) have
demonstrated that the search for the optimal channel QS corresponds to
the optimization over the Kraus representation of a given channel.
Without loss of generality we may assume that jihj is a pure
-dependent state while [] U U is unitary. For a given QS, we
may write the corresponding Kraus representation of the channel by
choosing a particular basis jiiE in the E space: Ki() EhijUjiE. In order
for the QS to be valid, these Kraus operators should P correspond to a legit-
imate Kraus representation of the channel i Ki Ki . Two
Kraus representation of a given quantum channel are equivalent if and only
if they are related by a unitary matrix u:
X
K i uij Kj , (143)
j
where the parameter s has the interpretation of FQ jihj for the partic-
ular QS at 0 and the constraints imposed in Equation (143) are necessary
and sufficient for the QS required transformation U and the state ji
to exist.
The above optimization problem may not always be easy to solve. Still,
its relaxed version:
X _ X _
Ki 0 { K_i 0 k ,
min k
h
Ki 0 { K i 0 0, (145)
i i
where kk is the operator norm, can always be cast in the form of an explicit
semidefinite program, which can be easily solved numerically (Demkowicz-
Dobrza nski et al., 2012). Numerical solution of the semidefinite program
provides a form of the optimal h which may then be taken as an ansatz
for further analytical optimization.
Plugging in the Kraus operators Ki() KiU representing the lossy
interferometer, see Equation (130), and following the above described
procedure one obtains
4
FQS q q2
1a 1b (146)
+a b
1 a 4 b 4
hopt diag , , + , (147)
8 1 a a 1 b a
where FQS a b
. This indeed reproduces the bounds given in (140).
a b
In order to provide the reader with a simple intuition concerning the QS
method, we shall present an elementary construction of the QS for lossy
interferometer in the special case of a b 1/2. In this case, the bound
416 Rafal Demkowicz-Dobrza
nski et al.
p
(140) yields 1= N which implies that using optimal entangled probe
state at the input under 50% losses cannot beat the precision which can be
obtained by uncorrelated probes in the ideal lossless scenario.
Consider the action of the single photon lossy channel on the pure
input state ji jai + jbi:
1 1
jihj j ih j + jvacihvacj (148)
2 2
1 1
jiji p ei j0ijai + j1ijbi + p ei j0ijbi + j1ijai : (149)
2 2
The map realizing the QS consists now of two steps. First the controlled
NOT operation is performed with the auxiliary system being the target
qubit, this transform the above state to: p12 j0iei jai + jbi + p12 j1i
ei jbi + jai. The second step is the measurement of the auxiliary system.
If the result j0i is measured (probability 1/2), the system is left in the correct
state j i and the map leaves it unchanged, if the j1i is measured the state of
the photon is not the desired one, in which case the map returns the jvaci
state. This map is therefore a proper QS of the desired lossy interferometer
transformation for a b 1/2. Since the auxiliary state employed in this
p
construction was ji ei j0i + j1i= 2 for which FQ(jihj N) N this
p
leads to the anticipated result 1= N .
over purifications method of Escher et al. (2011) and most recently using the
calculus of variations approach of Knysh et al. (2014).
2 p
Thanks to the fact that jhxj0ij2 ex = , the reduced state
Z1
dx E xj jx E (151)
1
indeed coincides with the correct output phase diffused state (125). There-
fore, this is a legitimate purification of the interferometer output state in
presence of phase diffusion.
418 Rafal Demkowicz-Dobrza
nski et al.
E
Consider now another purification eiH^ E generated by a
local ( 0 + ) rotation of the mirror modes, i.e., a unitary operation on
the system E. We look for a transformation of the above form which hope-
fully erases as much information
E on the estimated phase as possible, so that
QFI for the purified state will be minimized leading to the best bound
on the QFI of . Choosing H^ E ^
p E we obtain the following upper
bound on the QFI
FQ min FQ ei^p E ji
p 2 4 2 Jz
min 22 + 4 1 2 2 Jz ,
1 + 4 2 Jz
and thus a lower limit on the precision
r r
1 1
+ 2 + 2, (152)
4 Jz N
where we plugged in 2Jz N2/4 corresponding to the NOON state
which maximized the variance for N-photon states. See also an alternative
derivation of the above result that has been proposed recently in
Macieszczak (2014). Crucially, the above result proves that the phase diffu-
p
sion constrains the error to approach a constant value as N ! 1, which
p
does not vanish in the asymptotic limit, what contrasts the 1= N behavior
characteristic for uncorrelated noise models. Note also that due to the cor-
related character of the noise, the bound (152) predicts that it may be more
beneficial to perform the estimation procedure on a group of k particles and
p
then repeat the procedure independently times obtaining 1= reduction
in estimation error, rather than employing N k in a single experimental
shot (Knysh et al., 2014).
Only very recently, the exact ultimate quantum limit for the N-photon
input states has been derived in (Knysh et al., 2014)
r
2
+ 2 , (153)
N
showing that the previous bound was not tight, with the second term fol-
lowing the HL-like asymptotic scaling of the noiseless decoherence-free
Bayesian scenario stated in Equation (107). In fact, as proven by Knysh
et al. (2014), the optimal states of the noiseless Bayesian scenario, i.e., the
Quantum Limits in Optical Interferometry 419
sine states (105), attain the above correct quantum limit. In Section 6.3.2, we
show that within the Bayesian approach, with the phase-diffusion effects
incorporated, the sine states are always the optimal inputs.
where Amax max 1nN fAn, n1 g is the largest element of the matrix A,
contained within its off-diagonal entries (155). The bound yields exactly the
same formula as the QFI bound (140), proving that in this case the Bayesian
and QFI approaches are equivalent:
s s!
p p 1 1 a 1 b 1
hC i hC imin + p , (157)
2 a b N
where represents the fact that Bayesian cost approximated the variance
only in the limit of large N. The fact that both approaches lead to the same
ultimate bounds on precision suggests that the optimal input states may be
approximated for N ! 1 up to an arbitrary good precision with states
manifesting only local finite-number of particle correlations and may in
particular be efficiently simulated with the concept of matrix-product
states ( Jarzyna & Demkowicz-Dobrza nski, 2013, 2015).
likewise assuming no prior knowledge and the natural cost function introduced
in Equation (93) the formula for the average cost reads
n
o
hC i Tr C 2 cT Ac, (158)
where this time one may think of the effective state, as of the input state in
which is firstly averaged over the Gaussian distribution dictated by the evo-
lution (124) and then over the cost function in accordance with
Equation (93). The optimal seed element of the covariant POVM is identical
as in the decoherence-free case jeN i heN j and the matrix A possesses
again only nonzero entries on its first off-diagonals, but this time all of them
are equal to e 2 . As a result, the minimal average cost (158) may be evaluated
analytically following exactly the calculation of Berry and Wiseman (2000)
for the noiseless scenario, which leads then to max 2e 2 cos and
N +2
hence
N !1
2
hC imin 2 1 e 2 cos 2 1e 2 + e 2 2 : (159)
N +2 N
The optimal input states are the same as in the decoherence-free case, i.e.,
they are the N-photon sine states of Equation (105). Note that in contrast to
the photonic loss which is an example of an uncorrelated noise, the minimal
average cost (159) does not asymptotically coincide with the QFI-based
precision limit (153) unless 1.
the fundamental bounds derived above. We will not discuss the phase-
diffusion noise, since the estimation uncertainty is finite in the asympto-
tic limit, and the issue of saturating the asymptotic bound becomes
trivial as practically all states lead to the same asymptotic precision value,
while saturating the bound for finite N requires the use of experimentally
inaccessible sine states.
Hence the bounds on precision derived in Sections 6.2.2 (losses) and 6.2.1
(imperfect visibility) are valid also under replacement of N by hN i. Still, one
may come across claims of precisions going beyond the above mentioned
bounds typically by a factor of two (Aspachs, Calsamiglia, Munoz
Tapia, & Bagan, 2009; Joo et al., 2011). This is only possible, however, if
classical reference beam required to perform, e.g., the homodyne detection
is not treated as a resource. As discussed in detail in Section 5.3.1, we take the
position that such reference beams should be treated in the same way as the
light traveling through the interferometer and as such also counted as a
resource.
Quantum Limits in Optical Interferometry 423
h J^z i
cosh J^z iin sinh J^x iin ,
2 Jz hN i (162)
f + cos 2 2 Jz jin + sin 2 2 Jx jin +
2 4
2 sincos cov Jx , Jz jin :
where f() (1 )/ for the loss model and f() (1 2)/2 in the case of
local dephasing model. The above expressions have a clear intuitive inter-
pretation. The signal h J^z i is rescaled by a factor compared with the
decoherence-free case, while the variance apart from the analogous rescaling
is enlarged by an additional noise contribution f()hN i/4 due to lost or
dephased photons.
424 Rafal Demkowicz-Dobrza
nski et al.
Figure 12 The phase estimation precision of an interferometer with equal losses in both
arms ( 0.9). The performance of the optimal N-photon input states (120) is shown
(solid black) that indeed saturate the asymptotic quantum limit (140) (dotted):
p
1 =N. The NOON states (solid grey) achieve nearly optimal precision only
for low N ( 10) and rapidly diverge becoming out-performed by classical strategies.
For comparison, the precision attained for an indefinite photon-number scheme is pres-
ented, i.e., a coherent state and squeezed vacuum optimally mixed on a beam-splitter
(Caves, 1981) (dashed), which in the presence of loss also saturates the asymptotic quan-
tum limit (140).
7. CONCLUSIONS
In this review, we have showed how the tools of quantum estimation
theory can be applied in order to derive fundamental bounds on achievable pre-
cision in quantum-enhanced optical interferometric experiments. The main
message to be conveyed is the fact that while the power of quantum enhance-
ment is seriously reduced by the presence of decoherence, and, in general, the
Heisenberg scaling cannot be reached, nonclassical states of light offer a notice-
able improvement in interferometric precision and simple experimental
schemes may approach arbitrary close the fundamental quantum bounds. It
is also worth noting that in the presence of uncorrelated decoherence the
Bayesian approaches coincide asymptotically with the QFI approaches easing
the tension between this two often competing ways of statistical analysis.
We would also like to mention an inspiring alternative approach to the
derivation of limits on precision of phase estimation, where the results are
derived making use of information theoretic concepts such as rate-distortion
theory (Nair, 2012) or entropic uncertainty relations (Hall & Wiseman,
2012). Even though the bounds derived in this way are weaker than the bou-
nds presented in this review and obtained via Bayesian or QFI approaches,
they carry a conceptual appeal encouraging to look for deeper connections
between quantum estimation and communication theories.
Let us also point out, that while we have focused our discussion on opti-
cal interferometry using the paradigmatic MZ model, the same methods can
be applied to address the problems of fundamental precision bounds in
atomic interferometry (Cronin et al., 2009), magnetometry (Budker &
Romalis, 2007), frequency stabilization in atomic clocks (Diddams et al.,
2004) as well as the limits on resolution of quantum-enhanced lithographic
protocols (Boto et al., 2000). All these setups can be cast into a common
mathematical framework, see Section 3.6, but the resulting bounds will
depend strongly on the nature of dominant decoherence effects and the rel-
evant resource limitations such as: total experimental time, light power,
number of atoms, etc., as well as on the chosen figure of merit. In particular,
it is not excluded that in some atomic metrological scenarios one may still
Quantum Limits in Optical Interferometry 427
p
obtain a better than 1= N of precision if decoherence is of a special form
allowing for use of the decoherence-free subspaces (Dorner, 2012; Jeske,
Cole, & Huelga, 2013) or when its impact may be significantly reduced
by considering short evolution times, in which the SQL-like scaling bounds
may in principle be circumvented by: adjusting decoherence geometry
(Arrad, Vinkler, Aharonov, & Retzker, 2014; Chaves, Brask,
Markiewicz, Koody nski, & Acin, 2013; D ur, Skotiniotis, Fr
owis, &
Kraus, 2014; Kessler, Lovchinsky, Sushkov, & Lukin, 2013) or by consid-
ering non-Markovian short-time behavior (Chin, Huelga, & Plenio, 2012;
Matsuzaki, Benjamin, & Fitzsimons, 2011).
We should also note that application of the tools presented in this review
to a proper analysis of fundamental limits to the operation of quantum-
enhanced atomic clocks (Andre`, Srensen, & Lukin, 2004; Leibfried
et al., 2004) is not that direct as it requires taking into account precise fre-
quency noise characteristic of the local oscillator, allowing to determine the
optimal stationary operation regime of the clock (Macieszczak,
Demkowicz-Dobrza nski, & Fraas, 2013) ideally in terms of the Allan vari-
ance taken as a figure of merit (Fraas, 2013). A deeper theoretical insight into
this problem is still required to yield computable fundamental bounds.
The applicability of the tools presented has also been restricted to single
phase parameter estimation. A more general approach may be taken, were
multiple-phases (Humphreys, Barbieri, Datta, & Walmsley, 2013) or the
phase as well as the decoherence strength itself are the quantities to be esti-
mated (Crowley, Datta, Barbieri, & Walmsley, 2014; Knysh & Durkin,
2013). This poses an additional theoretical challenge as then the multi-
parameter quantum estimation theory needs to be applied, while most of
the tools discussed in this review are applicable only to single-parameter
estimation. Developing nontrivial multiparameter fundamental bounds for
quantum metrology is therefore still an open field for research.
ACKNOWLEDGMENTS
This research work supported by the FP7 IP project SIQS co-financed by the Polish Ministry
of Science and Higher Education, Polish NCBiR under the ERA-NET CHIST-ERA
project QUASAR, and Foundation for Polish Science TEAM project.
REFERENCES
Adesso, G. (2006). Entanglelement of Gaussian states.. Universita degli Studi di Salerno,
Unpublished doctoral dissertation.
Afek, I., Ambar, O., & Silberberg, Y. (2010). High-NOON states by mixing quantum and
classical light. Science, 328(5980), 879881. http://dx.doi.org/10.1126/science.1188172.
428 Rafal Demkowicz-Dobrza
nski et al.
Alipour, S. (2014). Dynamical and kinematic bounds for quantum metrology in open systems
(ArXiv e-prints arXiv:1403.8033).
Alipour, S., Mehboudi, M., & Rezakhani, A. T. (2014). Quantum metrology in open sys-
tems: Dissipative Cramer-Rao bound. Physical Review Letters, 112, 120405. http://dx.
doi.org/10.1103/PhysRevLett.112.120405.
Anderson, B. P., & Kasevich, M. A. (1998). Macroscopic quantum interference from atomic
tunnel arrays. Science, 282(5394), 16861689. http://dx.doi.org/10.1126/
science.282.5394.1686.
Andre`, A., Srensen, A. S., & Lukin, M. D. (2004). Stability of atomic clocks based on
entangled atoms. Physical Review Letters, 92, 230801.
Anisimov, P. M., Raterman, G. M., Chiruvelli, A., Plick, W. N., Huver, S. D., Lee, H., &
Dowling, J. P. (2010). Quantum metrology with two-mode squeezed vacuum: Parity
detection beats the Heisenberg limit. Physical Review Letters, 104, 103602. http://dx.
doi.org/10.1103/PhysRevLett.104.103602.
Arrad, G., Vinkler, Y., Aharonov, D., & Retzker, A. (2014). Increasing sensing resolution
with error correction. Physical Review Letters, 112, 150801. http://dx.doi.org/10.1103/
PhysRevLett.112.150801.
Aspachs, M., Calsamiglia, J., Munoz Tapia, R., & Bagan, E. (2009). Phase estimation for
thermal Gaussian states. Physical Review A, 79, 033834. http://dx.doi.org/10.1103/
PhysRevA.79.033834.
Bachor, H.-A., & Ralph, T. (2004). A guide to experiments in quantum optics. Weinheim:
Wiley-VCH.
Banaszek, K., Horodecki, P., Karpinski, M., & Radzewicz, C. (2013). Quantum mechanical
which-way experiment with an internal degree of freedom. Nature Communications, 4, 2594.
Banaszek, K., Radzewicz, C., Wodkiewicz, K., & Krasinski, J. S. (1999). Direct measure-
ment of the Wigner function by photon counting. Physical Review A, 60, 674677.
http://dx.doi.org/10.1103/PhysRevA.60.674.
Banaszek, K., & Wodkiewicz, K. (1998). Nonlocality of the Einstein-Podolsky-Rosen state
in the Wigner representation. Physical Review A, 58, 43454347. http://dx.doi.org/
10.1103/PhysRevA.58.4345.
Barnett, S. M., & Pegg, D. T. (1992). Limiting procedures for the optical phase operator.
Journal of Modern Optics, 39(10), 21212129. http://dx.doi.org/10.1080/
09500349214552141.
Bartlett, S. D., Rudolph, T., & Spekkens, R. W. (2007). Reference frames, superselection
rules, and quantum information. Reviews of Modern Physics, 79(2), 555. http://dx.doi.org/
10.1103/RevModPhys.79.555.
Benatti, F., & Braun, D. (2013). Sub-shot-noise sensitivities without entanglement. Physical
Review A, 87, 012340. http://dx.doi.org/10.1103/PhysRevA.87.012340.
Benatti, F., Floreanini, R., & Marzolino, U. (2010). Sub-shot-noise quantum metrology
with entangled identical particles. Annals of Physics, 325(4), 924935. http://dx.doi.
org/10.1016/j.aop.2010.01.005.
Bengtsson, I., & Zyczkowski, K. (2006). Geometry of quantum states: An introduction to quantum
entanglement. Cambridge: Cambridge Univeristy Press.
Berry, D. W., Hall, M. J. W., Zwierz, M., & Wiseman, H. M. (2012). Optimal Heisenberg-
style bounds for the average performance of arbitrary phase estimates. Physical Review A,
86, 053813. http://dx.doi.org/10.1103/PhysRevA.86.053813.
Berry, D. W., Higgins, B. L., Bartlett, S. D., Mitchell, M. W., Pryde, G. J., & Wiseman, H. M.
(2009). How to perform the most accurate possible phase measurements. Physical Review A,
80(5), 052114. http://dx.doi.org/10.1103/PhysRevA.80.052114.
Berry, D. W., & Wiseman, H. M. (2000). Optimal states and almost optimal adaptive mea-
surements for quantum interferometry. Physical Review Letters, 85(24), 50985101.
http://dx.doi.org/10.1103/PhysRevLett.85.5098.
Quantum Limits in Optical Interferometry 429
Bollinger, J. J., Itano, W. M., Wineland, D. J., & Heinzen, D. J. (1996). Optimal frequency
measurements with maximally correlated states. Physical Review A, 54(6),
R4649R4652. http://dx.doi.org/10.1103/PhysRevA.54.R4649.
Bondurant, R. S., & Shapiro, J. H. (1984). Squeezed states in phase-sensing inter-
ferometers. Physical Review D, 30, 25482556. http://dx.doi.org/10.1103/
PhysRevD.30.2548.
Boto, A. N., Kok, P., Abrams, D. S., Braunstein, S. L., Williams, C. P., & Dowling, J. P.
(2000). Quantum interferometric optical lithography: Exploiting entanglement to beat
the diffraction limit. Physical Review Letters, 85, 27332736. http://dx.doi.org/10.1103/
PhysRevLett.85.2733.
Braunstein, S. L. (1992). Quantum limits on precision measurements of phase. Physical Review
Letters, 69, 35983601. http://dx.doi.org/10.1103/PhysRevLett.69.3598.
Braunstein, S. L., & Caves, C. M. (1994). Statistical distance and the geometry of quantum
states. Physical Review Letters, 72(22), 34393443. http://dx.doi.org/10.1103/
PhysRevLett.72.3439.
Budker, D., & Romalis, M. (2007). Optical magnetometry. Nature Physics, 3, 227234.
http://dx.doi.org/10.1038/nphys566.
Buzek, V., Derka, R., & Massar, S. (1999). Optimal quantum clocks. Physical Review Letters,
82, 22072210. http://dx.doi.org/10.1103/PhysRevLett.82.2207.
Buzek, V., & Knight, P. L. (1995). I: Quantum interference, superposition states of light, and
nonclassical effects. In E. Wolf (Ed.), Annals of Physics: Vol. 34 (pp. 1158). Amsterdam:
Elsevier, http://dx.doi.org/10.1016/S0079-6638(08)70324-X.
Caves, C. M. (1981). Quantum-mechanical noise in an interferometer. Physical Review D,
23(8), 16931708. http://dx.doi.org/10.1103/PhysRevD.23.1693.
Chaves, R., Brask, J. B., Markiewicz, M., Koody nski, J., & Acin, A. (2013). Noisy metrol-
ogy beyond the standard quantum limit. Physical Review Letters, 111, 120401. http://dx.
doi.org/10.1103/PhysRevLett.111.120401.
Chekhova, M. (2011). Chapter 4Polarization and spectral properties of biphotons.
In E. Wolf (Ed.), Progress in optics: Vol. 56 (pp. 187226). Amsterdam: Elsevier.
http://dx.doi.org/10.1016/B978-0-444-53886-4.00004-6.
Chin, A. W., Huelga, S. F., & Plenio, M. B. (2012). Quantum metrology in non-Markovian
environments. Physical Review Letters, 109(23), 233601. http://dx.doi.org/10.1103/
PhysRevLett.109.233601.
Chiribella, G., DAriano, G. M., & Sacchi, M. F. (2005). Optimal estimation of group trans-
formations using entanglement. Physical Review A, 72, 042338. http://dx.doi.org/
10.1103/PhysRevA.72.042338.
Choi, M.-D. (1975). Completely positive linear maps on complex matrices. Linear Algebra and
its Applications, 10, 285290. http://dx.doi.org/10.1016/0024-3795(75)90075-0.
Chwede nczuk, J., Hyllus, P., Piazza, F., & Smerzi, A. (2012). Sub-shot-noise interferometry
from measurements of the one-body density. New Journal of Physics, 14(9), 093001.
http://dx.doi.org/10.1088/1367-2630/14/9/093001.
Chwede nczuk, J., Piazza, F., & Smerzi, A. (2011). Phase estimation from atom position mea-
surements. New Journal of Physics, 13(6), 065023. http://dx.doi.org/10.1088/1367-
2630/13/6/065023.
Cronin, A. D., Schmiedmayer, J., & Pritchard, D. E. (2009). Optics and interferometry with
atoms and molecules. Reviews of Modern Physics, 81, 10511129. http://dx.doi.org/
10.1103/RevModPhys.81.1051.
Crowley, P. J. D., Datta, A., Barbieri, M., & Walmsley, I. A. (2014). Tradeoff in simulta-
neous quantum-limited phase and loss estimation in interferometry. Physical Review A,
89, 023845. http://dx.doi.org/10.1103/PhysRevA.89.023845.
DAriano, G. M., Macchiavello, C., & Paris, M. G. (1995). Optimized phase detection. Phys-
ics Letters A, 198(4), 286294. http://dx.doi.org/10.1016/0375-9601(95)00009-R.
430 Rafal Demkowicz-Dobrza
nski et al.
Datta, A., Zhang, L., Thomas-Peter, N., Dorner, U., Smith, B. J., & Walmsley, I. A. (2011).
Quantum metrology with imperfect states and detectors. Physical Review A, 83, 063836.
http://dx.doi.org/10.1103/PhysRevA.83.063836.
Demkowicz-Dobrza nski, R. (2010). Multi-pass classical vs. quantum strategies in lossy phase
estimation. Laser Physics, 20, 11971202. http://dx.doi.org/10.1134/S1054660X10090306.
Demkowicz-Dobrza nski, R., Banaszek, K., & Schnabel, R. (2013). Fundamental quantum
interferometry bound for the squeezed-light-enhanced gravitational wave detector GEO
600. Physical Review A 88. http://dx.doi.org/10.1103/PhysRevA.88.041802, 041802(R).
Demkowicz-Dobrza nski, R., Dorner, U., Smith, B. J., Lundeen, J. S., Wasilewski, W.,
Banaszek, K., et al. (2009). Quantum phase estimation with lossy interferometers. Phys-
ical Review A, 80, 013825. http://dx.doi.org/10.1103/PhysRevA.80.013825.
Demkowicz-Dobrza nski, R., Koody nski, J., & Guta, M. (2012). The elusive Heisenberg
limit in quantum enhanced metrology. Nature Communications, 3, 1063. http://dx.doi.
org/10.1038/ncomms2067.
Demkowicz-Dobrza nski, R., & Maccone, L. (2014). Using Entanglement Against Noise in
Quantum Metrology. Physical Review Letters, 113, 250801.
Diddams, S. A., Bergquist, J. C., Jefferts, S. R., & Oates, C. W. (2004). Standards of time and
frequency at the outset of the 21st century. Science, 306(5700), 13181324. http://dx.doi.
org/10.1126/science.1102330.
Dorner, U. (2012). Quantum frequency estimation with trapped ions and atoms. New Journal
of Physics, 14(4), 043011. http://dx.doi.org/10.1088/1367-2630/14/4/043011.
Dorner, U., Demkowicz-Dobrza nski, R., Smith, B. J., Lundeen, J. S., Wasilewski, W.,
Banaszek, K., et al. (2009). Optimal quantum phase estimation. Physical Review Letters,
102, 040403. http://dx.doi.org/10.1103/PhysRevLett.102.040403.
Dowling, J. P. (1998). Correlated input-port, matter-wave interferometer: Quantum-noise
limits to the atom-laser gyroscope. Physical Review A, 57, 4736. http://dx.doi.org/
10.1103/PhysRevA.57.4736.
Dur, W., Skotiniotis, M., Fr owis, F., & Kraus, B. (2014). Improved quantum metrology
using quantum error correction. Physical Review Letters, 112, 080801. http://dx.doi.
org/10.1103/PhysRevLett.112.080801.
Escher, B. M., Davidovich, L., Zagury, N., & de Matos Filho, R. L. (2012). Quantum met-
rological limits via a variational approach. Physical Review Letters, 109, 190404. http://dx.
doi.org/10.1103/PhysRevLett.109.190404.
Escher, B. M., de Matos Filho, R. L., & Davidovich, L. (2011). General framework for
estimating the ultimate precision limit in noisy quantum-enhanced metrology. Nature
Physics, 7, 406411. http://dx.doi.org/10.1038/nphys1958.
Fox, M. (2006). Quantum optics. Oxford: Oxford University Press.
Fraas, M. (2013). An analysis of the stationary operation of atomic clocks (ArXiv e-prints
arXiv:1303.6083).
Fr
owis, F., Skotiniotis, M., Kraus, B., and D ur, W. (2014). Optimal quantum states for
frequency estimation (ArXiv e-prints arXiv:1402.6946).
Fujiwara, A., & Imai, H. (2008). A fibre bundle over manifolds of quantum channels and its
application to quantum statistics. Journal of Physics A: Mathematical and Theoretical, 41,
255304. http://dx.doi.org/10.1088/1751-8113/41/25/255304.
Genoni, M. G., Olivares, S., & Paris, M. G. A. (2011). Optical phase estimation in the pres-
ence of phase diffusion. Physical Review Letters, 106, 153603. http://dx.doi.org/10.1103/
PhysRevLett.106.153603.
Gilbert, G., Hamrick, M., & Weinstein, Y. S. (2008). Use of maximally entangled N-photon
states for practical quantum interferometry. Journal of the Optical Society of America B:
Optical Physics, 25, 1336. http://dx.doi.org/10.1364/JOSAB.25.001336.
Giovannetti, V., Lloyd, S., & Maccone, L. (2004). Quantum-enhanced measurements:
Beating the standard quantum limit. Science, 306, 1330. http://dx.doi.org/10.1126/
science.1104149.
Quantum Limits in Optical Interferometry 431
Giovannetti, V., Lloyd, S., & Maccone, L. (2006). Quantum metrology. Physical Review
Letters, 96, 010401. http://dx.doi.org/10.1103/PhysRevLett.96.010401.
Giovannetti, V., Lloyd, S., & Maccone, L. (2011). Advances in quantum metrology. Nature
Photonics, 5, 222229. http://dx.doi.org/10.1038/nphoton.2011.35.
Giovannetti, V., & Maccone, L. (2012). Sub-Heisenberg estimation strategies are ineffective.
Physical Review Letters, 108, 210404. http://dx.doi.org/10.1103/PhysRevLett.108.210404.
Glauber, R. J. (1963). Coherent and incoherent states of the radiation field. Physics Review,
131, 2766. http://dx.doi.org/10.1103/PhysRev.131.2766.
Gross, C., Zibold, T., Nicklas, E., Esteve, J., & Oberthaler, M. K. (2010). Nonlinear atom
interferometer surpasses classical precision limit. Nature, 464, 11651169. http://dx.doi.
org/10.1038/nature08919.
Hall, M. J. W., & Wiseman, H. M. (2012). Heisenberg-style bounds for arbitrary estimates of
shift parameters including prior information. New Journal of Physics, 14, 033040. http://
dx.doi.org/10.1088/1367-2630/14/3/033040.
Hariharan, P. (2003). Optical interferometry. Amsterdam: Elsevier.
Helstrom, C. W. (1976). Quantum detection and estimation theory. London: Academic press.
Higgins, B. L., Berry, D. W., Bartlett, S. D., Wiseman, H. M., & Pryde, G. J. (2007).
Entanglement-free Heisenberg-limited phase estimation. Nature, 450, 393396.
http://dx.doi.org/10.1038/nature06257.
Hofheinz, M., Weig, E. M., Ansmann, M., Bialczak, R. C., Lucero, E., Neeley, M., et al.
(2008). Generation of Fock states in a superconducting quantum circuit. Nature, 454,
310314. http://dx.doi.org/10.1038/nature07136.
Hofmann, H. F. (2009). All path-symmetric pure states achieve their maximal phase sensi-
tivity in conventional two-path interferometry. Physical Review A, 79(3), 033822. http://
dx.doi.org/10.1103/PhysRevA.79.033822.
Holevo, A. S. (1982). Probabilistic and statistical aspects of quantum theory. Amsterdam: North
Holland.
Holland, M. J., & Burnett, K. (1993). Interferometric detection of optical phase shifts at the
Heisenberg limit. Physical Review Letters, 71(9), 13551358. http://dx.doi.org/10.1103/
PhysRevLett.71.1355.
Horodecki, R., Horodecki, P., Horodecki, M., & Horodecki, K. (2009). Quantum entanglement.
Reviews of Modern Physics, 81, 865942. http://dx.doi.org/10.1103/RevModPhys.81.865.
Horrom, T., Singh, R., Dowling, J. P., & Mikhailov, E. E. (2012). Quantum-enhanced
magnetometer with low-frequency squeezing. Physical Review A, 86, 023803. http://
dx.doi.org/10.1103/PhysRevA.86.023803.
Hradil, Z., Myska, R., Perina, J., Zawisky, M., Hasegawa, Y., & Rauch, H. (1996). Quan-
tum phase in interferometry. Physical Review Letters, 76, 42954298. http://dx.doi.org/
10.1103/PhysRevLett.76.4295.
Huelga, S. F., Macchiavello, C., Pellizzari, T., Ekert, A. K., Plenio, M. B., & Cirac, J. I.
(1997). Improvement of frequency standards with quantum entanglement. Physical
Review Letters, 79(20), 38653868. http://dx.doi.org/10.1103/PhysRevLett.79.3865.
Humphreys, P. C., Barbieri, M., Datta, A., & Walmsley, I. A. (2013). Quantum enhanced
multiple phase estimation. Physical Review Letters, 111, 070403. http://dx.doi.org/
10.1103/PhysRevLett.111.070403.
Huver, S. D., Wildfeuer, C. F., & Dowling, J. P. (2008). Entangled Fock states for robust
quantum optical metrology, imaging, and sensing. Physical Review A, 78, 063828.
http://dx.doi.org/10.1103/PhysRevA.78.063828.
Hyllus, P., Laskowski, W., Krischek, R., Schwemmer, C., Wieczorek, W., Weinfurter, H.,
et al. (2012). Fisher information and multiparticle entanglement. Physical Review A, 85,
022321. http://dx.doi.org/10.1103/PhysRevA.85.022321.
Jamiokowski, A. (1972). Linear transformations which preserve trace and positive semi-
definiteness of operators. Reports on Mathematical Physics, 3, 275278. http://dx.doi.
org/10.1016/0034-4877(72)90011-0.
432 Rafal Demkowicz-Dobrza
nski et al.
Jarzyna, M., & Demkowicz-Dobrza nski, R. (2012). Quantum interferometry with and
without an external phase reference. Physical Review A, 85. http://dx.doi.org/
10.1103/PhysRevA.85.011801, 011801(R).
Jarzyna, M., & Demkowicz-Dobrza nski, R. (2013). Matrix product states for quantum
metrology. Physical Review Letters, 110, 240405. http://dx.doi.org/10.1103/
PhysRevLett.110.240405.
Jarzyna, M., & Demkowicz-Dobrzanski, R. (2015). True precision limits in quantum
metrology. New Journal of Physics, 17, 013010.
Jeske, J., Cole, J. H., & Huelga, S. (2013). Quantum metrology in the presence of spatially
correlated noise: Restoring Heisenberg scaling (ArXiv e-prints arXiv:1307.6301).
Joo, J., Munro, W. J., & Spiller, T. P. (2011). Quantum metrology with entangled coherent
states. Physical Review Letters, 107, 083601. http://dx.doi.org/10.1103/
PhysRevLett.107.083601.
Kacprowicz, M., Demkowicz-Dobrza nski, R., Wasilewski, W., & Banaszek, K. (2010).
Experimental quantum enhanced phase-estimation in the presence of loss. Nature
Photonics, 4, 357360. http://dx.doi.org/10.1038/nphoton.2010.39.
Kaftal, T., & Demkowicz-Dobrza nski, R. (2014). Usefulness of an enhanced Kitaev phase-
estimation algorithm in quantum metrology and computation. Physical Review A, 90,
062313.
Kay, S. M. (1993). Fundamentals of statistical signal processing: Estimation theory (S. E. Alan V.
Oppenheim, Ed.). Prentice Hall: Upper Saddle River.
Kessler, E. M., Lovchinsky, I., Sushkov, A. O., & Lukin, M. D. (2013). Quantum Error
Correction for Metrology. Physical Review Letters, 112, 150802.
Killoran, N., Cramer, M., & Plenio, M. B. (2014). Extracting entanglement from identical
particles. Physical Review Letters, 112, 150501. http://dx.doi.org/10.1103/
PhysRevLett.112.150501.
Kitagawa, M., & Ueda, M. (1993). Squeezed spin states. Physical Review A, 47, 51385143.
http://dx.doi.org/10.1103/PhysRevA.47.5138.
Knysh, S., Smelyanskiy, V. N., & Durkin, G. A. (2011). Scaling laws for precision in quan-
tum interferometry and the bifurcation landscape of the optimal state. Physical Review A,
83, 021804. http://dx.doi.org/10.1103/PhysRevA.83.021804.
Knysh, S. I., Chen, E. H., & Durkin, G. A. (2014). True limits to precision via unique quan-
tum probe (ArXiv e-prints arXiv:1402.0495).
Knysh, S. I., & Durkin, G. A. (2013). Estimation of phase and diffusion: Combining quantum
statistics and classical noise (ArXiv e-prints arXiv:1307.0470).
Kok, P., Munro, W. J., Nemoto, K., Ralph, T. C., Dowling, J. P., & Milburn, G. J. (2007).
Linear optical quantum computing with photonic qubits. Reviews of Modern Physics, 79,
135174. http://dx.doi.org/10.1103/RevModPhys.79.135.
Koody nski, J., & Demkowicz-Dobrza nski, R. (2010). Phase estimation without a priori
phase knowledge in the presence of loss. Physical Review A, 82, 053804. http://dx.
doi.org/10.1103/PhysRevA.82.053804.
Koody nski, J., & Demkowicz-Dobrza nski, R. (2013). Efficient tools for quantum metrology
with uncorrelated noise. New Journal of Physics, 15, 073043. http://dx.doi.org/
10.1088/1367-2630/15/7/073043.
Krischek, R., Schwemmer, C., Wieczorek, W., Weinfurter, H., Hyllus, P., Pezze, L., et al.
(2011). Useful multiparticle entanglement and sub-shot-noise sensitivity in experimental
phase estimation. Physical Review Letters, 107, 080504. http://dx.doi.org/10.1103/
PhysRevLett.107.080504.
Lee, H., Kok, P., & Dowling, J. P. (2002). A quantum Rosetta stone for interferometry.
Journal of Modern Optics, 49(1415), 23252338. http://dx.doi.org/10.1080/
0950034021000011536.
Lehmann, E. L., & Casella, G. (1998). Theory of point estimation (Vol. 31). New York:
Springer.
Quantum Limits in Optical Interferometry 433
Leibfried, D., Barrett, M. D., Schaetz, T., Britton, J., Chiaverini, J., Itano, W. M., et al.
(2004). Toward Heisenberg-limited spectroscopy with multiparticle entangled states.
Science, 304, 14761478.
Leonhardt, U. (1997). Measuring the quantum state of light (Vol. 22). Cambridge:
Cambridge Univeristy Press.
LIGO Collaboration. (2011). A gravitational wave observatory operating beyond the quantum
shot-noise limit. Nature Physics, 7, 962965. http://dx.doi.org/10.1038/nphys2083.
LIGO Collaboration. (2013). Enhanced sensitivity of the LIGO gravitational wave detector
by using squeezed states of light. Nature Photonics, 7(8), 613619. http://dx.doi.org/
10.1038/nphoton.2013.177.
Loudon, R. (2000). The quantum theory of light. Oxford: Oxford University Press.
Luis, A., & Perina, J. (1996). Optimum phase-shift estimation and the quantum description of
the phase difference. Physical Review A, 54, 45644570. http://dx.doi.org/10.1103/
PhysRevA.54.4564.
Lynch, R. (1995). The quantum phase problem: A critical review. Physics Reports, 256(6),
367436. http://dx.doi.org/10.1016/0370-1573(94)00095-K.
Ma, J., Wang, X., Sun, C., & Nori, F. (2011). Quantum spin squeezing. Physics Reports,
509(2), 89165. http://dx.doi.org/10.1016/j.physrep.2011.08.003.
Macieszczak, K. (2014). Upper bounds on the quantum Fisher information in the presence of
general dephasing (ArXiv e-prints arXiv:1403.0955).
Macieszczak, K., Demkowicz-Dobrza nski, R., & Fraas, M. (2013). Bayesian quantum
frequency estimation in presence of collective dephasing. New Journal of Physics, 16,
113002.
Mandel, L., & Wolf, E. (2000). Optical coherence and quantum optics. Cambridge: Cambridge
Univeristy Press.
Matsumoto, K. (2010). On metric of quantum channel spaces., 1006.0300v1 (ArXiv
e-prints).
Matsuzaki, Y., Benjamin, S. C., & Fitzsimons, J. (2011). Magnetic field sensing beyond the
standard quantum limit under the effect of decoherence. Physical Review A, 84, 012103.
http://dx.doi.org/10.1103/PhysRevA.84.012103.
Mitchell, M. W., Lundeen, J. S., & Steinberg, A. M. (2004). Super-resolving phase measure-
ments with a multi-photon entangled state. Nature, 429, 161164. http://dx.doi.org/
10.1038/nature02493.
Mlmer, K. (1997). Optical coherence: A convenient fiction. Physical Review A, 55,
31953203. http://dx.doi.org/10.1103/PhysRevA.55.3195.
Nagaoka, H. (2005). On Fisher information of quantum statistical models. In M. Hayashi
(Ed.), (pp. 113125). Singapore: World Scientific, (Chapter 9).
Nagata, T., Okamoto, R., OBrien, J. L., Sasaki, K., & Takeuchi, S. (2007). Beating the stan-
dard quantum limit with four entangled photons. Science, 316, 726729. http://dx.doi.
org/10.1126/science.1138007.
Nair, R. (2012). Fundamental limits on the accuracy of optical phase estimation from rate-
distortion theory (ArXiv e-prints arXiv:1204.3761).
Nielsen, M. A., & Chuang, I. L. (2000). Quantum computing and quantum information.
Cambridge: Cambridge University Press.
Noh, J. W., Fouge`res, A., & Mandel, L. (1992). Operational approach to the phase of a
quantum field. Physical Review A, 45, 424442. http://dx.doi.org/10.1103/
PhysRevA.45.424.
Okamoto, R., Hofmann, H. F., Nagata, T., OBrien, J. L., Sasaki, K., & Takeuchi, S. (2008).
Beating the standard quantum limit: Phase super-sensitivity of N-photon interferome-
ters. New Journal of Physics, 10(7), 073033. http://dx.doi.org/10.1088/1367-2630/10/
7/073033.
Olivares, S., & Paris, M. G. A. (2007). Optimized interferometry with Gaussian states. Optics
and Spectroscopy, 103, 231236. http://dx.doi.org/10.1134/S0030400X07080103.
434 Rafal Demkowicz-Dobrza
nski et al.
Ono, T., & Hofmann, H. F. (2010). Effects of photon losses on phase estimation near the
Heisenberg limit using coherent light and squeezed vacuum. Physical Review A, 81,
033819. http://dx.doi.org/10.1103/PhysRevA.81.033819.
Pan, J.-W., Chen, Z.-B., Lu, C.-Y., Weinfurter, H., Zeilinger, A., & zukowski, M. (2012).
Multiphoton entanglement and interferometry. Reviews of Modern Physics, 84, 777838.
http://dx.doi.org/10.1103/RevModPhys.84.777.
Pezze, L., & Smerzi, A. (2008). Mach-Zehnder interferometry at the Heisenberg limit with
coherent and squeezed-vacuum light. Physical Review Letters, 100(7), 073601. http://dx.
doi.org/10.1103/PhysRevLett.100.073601.
Pezze, L., & Smerzi, A. (2009). Entanglement, nonlinear dynamics, and the Heisenberg limit.
Physical Review Letters, 102, 100401. http://dx.doi.org/10.1103/PhysRevLett.
102.100401.
Pinel, O., Fade, J., Braun, D., Jian, P., Treps, N., & Fabre, C. (2012). Ultimate sensitivity of
precision measurements with intense Gaussian quantum light: A multimodal approach.
Physical Review A, 85. http://dx.doi.org/10.1103/PhysRevA.85.010101, 010101(R).
Pinel, O., Jian, P., Fabre, C., Treps, N., & Braun, D. (2013). Quantum parameter estimation
using general single-mode Gaussian states. Physical Review A, 88. http://dx.doi.org/
10.1103/PhysRevA.88.040102, 040102(R).
Pitkin, M., Reid, S., Rowan, S., & Hough, J. (2011). Gravitational wave detection by inter-
ferometry (ground and space). Living Reviews in Relativity. 14(5). http://dx.doi.org/
10.12942/lrr-2011-5.
Reid, M. D., & Drummond, P. D. (1988, Jun). Quantum correlations of phase in non-
degenerate parametric oscillation. Physical Review Letters, 60, 27312733. http://dx.
doi.org/10.1103/PhysRevLett.60.2731.
Resch, K. J., Pregnell, K. L., Prevedel, R., Gilchrist, A., Pryde, G. J., OBrien, J. L., et al.
(2007). Time-reversal and super-resolving phase measurements. Physical Review Letters,
98(22), 223601. http://dx.doi.org/10.1103/PhysRevLett.98.223601.
Rivas, A., & Luis, A. (2012). Sub-Heisenberg estimation of non-random phase shifts. New
Journal of Physics, 14(9), 093052. http://dx.doi.org/10.1088/1367-2630/14/9/093052.
Rubin, M. A., & Kaushik, S. (2007). Loss-induced limits to phase measurement precision
with maximally entangled states. Physical Review A, 75, 053805. http://dx.doi.org/
10.1103/PhysRevA.75.053805.
Sanders, B. C., & Milburn, G. J. (1995). Optimal quantum measurements for phase estima-
tion. Physical Review Letters, 75, 29442947. http://dx.doi.org/10.1103/
PhysRevLett.75.2944.
Sarovar, M., & Milburn, G. J. (2006). Optimal estimation of one parameter quantum chan-
nels. Journal of Physics A: Mathematical and General, 39, 8487. http://dx.doi.org/
10.1088/0305-4470/39/26/015.
Sayrin, C., Dotsenko, I., Zhou, X., Peaudecerf, B., Rybarczyk, T., Gleyzes, S., et al. (2011).
Real-time quantum feedback prepares and stabilizes photon number states. Nature, 477,
7377. http://dx.doi.org/10.1038/nature10376.
Schleich, W. P. (2001). Quantum optics in phase space. Berlin: Wiley-VCH.
Schwinger, J. (1965). Quantum theory of angular momentum. New York: Academic Press.
Seshadreesan, K. P., Anisimov, P. M., Lee, H., & Dowling, J. P. (2011). Parity detection
achieves the Heisenberg limit in interferometry with coherent mixed with squeezed
vacuum light. New Journal of Physics, 13, 083026. http://dx.doi.org/10.1088/1367-
2630/13/8/083026.
Sewell, R. J., Koschorreck, M., Napolitano, M., Dubost, B., Behbood, N., & Mitchell, M. W.
(2012). Magnetic sensitivity beyond the projection noise limit by spin squeezing. Physical
Review Letters, 109, 253605. http://dx.doi.org/10.1103/PhysRevLett.109.253605.
Shaji, A., & Caves, C. M. (2007). Qubit metrology and decoherence. Physical Review A, 76,
032111. http://dx.doi.org/10.1103/PhysRevA.76.032111.
Shankar, R. (1994). Principles of quantum mechanics. New York: Plenum.
Quantum Limits in Optical Interferometry 435
Shin, Y., Saba, M., Pasquini, T. A., Ketterle, W., Pritchard, D. E., & Leanhardt, A. E. (2004).
Atom interferometry with Bose-Einstein condensates in a double-well potential. Physical
Review Letters, 92, 050405. http://dx.doi.org/10.1103/PhysRevLett.92.050405.
Spagnolo, N., Vitelli, C., Lucivero, V. G., Giovannetti, V., Maccone, L., & Sciarrino, F.
(2012). Phase estimation via quantum interferometry for noisy detectors. Physical Review
Letters, 108(23), 233602. http://dx.doi.org/10.1103/PhysRevLett.108.233602.
Summy, G. S., & Pegg, D. T. (1990). Phase optimized quantum states of light. Optics
Communication, 77(1), 7579. http://dx.doi.org/10.1016/0030-4018(90)90464-5.
Torres, J. P., Banaszek, K., & Walmsley, I. (2011). Chapter 5Engineering nonlinear optic
sources of photonic entanglement. In E. Wolf (Ed.), Progress in optics: Vol. 56,
(pp. 227331). Amsterdam: Elsevier. http://dx.doi.org/10.1016/B978-0-444-53886-
4.00005-8.
Toth, G. (2012). Multipartite entanglement and high-precision metrology. Physical Review A,
85, 022322. http://dx.doi.org/10.1103/PhysRevA.85.022322.
Toth, G., & Petz, D. (2013). Extremal properties of the variance and the quantum Fisher
information. Physical Review A, 87, 032324. http://dx.doi.org/10.1103/PhysRevA.87.
032324.
Tsang, M. (2012). Ziv-Zakai error bounds for quantum parameter estimation. Physical Review
Letters, 108, 230401. http://dx.doi.org/10.1103/PhysRevLett.108.230401.
van der Vaart, A. W. (2000). Asymptotic statistics. Cambridge: Cambridge Univeristy Press.
Vitelli, C., Spagnolo, N., Toffoli, L., Sciarrino, F., & De Martini, F. (2010). Enhanced
resolution of lossy interferometry by coherent amplification of single photons.
Physical Review Letters, 105(11), 113602. http://dx.doi.org/10.1103/PhysRevLett.105.
113602.
Walls, D. F., & Milburn, G. J. (1995). Quantum optics. Berlin: Springer Verlag.
Wasilewski, W., Jensen, K., Krauter, H., Renema, J. J., Balabas, M. V., & Polzik, E. S.
(2010). Quantum noise limited and entanglement-assisted magnetometry. Physical
Review Letters, 104, 133601. http://dx.doi.org/10.1103/PhysRevLett.104.133601.
Wigner, E. P. (1932). On the quantum correction for thermodynamic equilibrium. Physics
Review, 40, 749759. http://dx.doi.org/10.1103/PhysRev.40.749.
Wineland, D. J., Bollinger, J. J., Itano, W. M., & Heinzen, D. J. (1994). Squeezed atomic
states and projection noise in spectroscopy. Phys. Rev. A, 50, 6788. http://dx.doi.
org/10.1103/PhysRevA.50.67.
Wineland, D. J., Bollinger, J. J., Itano, W. M., Moore, F. L., & Heinzen, D. J. (1992). Spin
squeezing and reduced quantum noise in spectroscopy. Phys. Rev. A, 46,
R6797R6800. http://dx.doi.org/10.1103/PhysRevA.46.R6797.
Xiang, G.-Y., Higgins, B. L., Berry, D. W., Wiseman, H. M., & Pryde, G. J. (2010).
Entanglement-enhanced measurement of a completely unknown optical phase. Nature
Photonics, 5, 4347. http://dx.doi.org/10.1038/nphoton.2010.268.
Xiao, M., Wu, L.-A., & Kimble, H. J. (1987). Precision measurement beyond the shot-noise
limit. Physical Review Letters, 59(3), 278281. http://dx.doi.org/10.1103/PhysRevLett.
59.278.
Yu, S. (2013). Quantum Fisher information as the convex roof of variance (ArXiv e-prints
arXiv:1302.5311).
Yurke, B., McCall, S. L., & Klauder, J. R. (1986). Su(2) and SU(1,1) interferometers. Physical
Review A, 33(6), 40334054. http://dx.doi.org/10.1103/PhysRevA.33.4033.
Zhang, Y. R., Jin, G. R., Cao, J. P., Liu, W. M., & Fan, H. (2013). Unbounded quantum
Fisher information in two-path interferometry with finite photon number. Journal of
Physics A: Mathematical and Theoretical, 46, 035302. http://dx.doi.org/10.1088/1751-
8113/46/3/035302.
Zwierz, M., Perez-Delgado, C. A., & Kok, P. (2010). General optimality of the Heisenberg
limit for quantum metrology. Physical Review Letters, 105, 180402. http://dx.doi.org/
10.1103/PhysRevLett.105.180402.
This page intentionally left blank
CHAPTER FIVE
Contents
1. Introduction 437
2. Classical Electrodynamics 440
2.1 Electrodynamics of Moving Media 441
2.2 The Quasi-Stationary Approximation 446
2.3 Electrodynamics of Quasi-Stationary Media 449
3. Optical Force, Momentum, and Stress 452
3.1 Electromagnetic Force 452
3.2 Electromagnetic Momentum 455
3.3 Electromagnetic Stress 460
4. Momentum and Stress in Dispersive Media 467
4.1 Field and Material Contributions 468
4.2 Time-Averaged Energy and Momentum Propagation 470
5. Discussion 472
5.1 Conclusions 473
5.2 Application to Optical Manipulation 475
Appendix A. Scattering by a Cylinder 478
Appendix B. Scattering by a Sphere 481
Acknowledgments 485
References 485
1. INTRODUCTION
Light possesses energy and momentum. This statement should be obvi-
ous any time we feel the warmth of the sun or look at the tail of a comet. Light
leaves the sun, travels long distances through space, and does work on pon-
derable media at remote distances. Conservation principles ensure that the
energy and momentum are not lost only to be regained after the elapsed time
required for light to reach its destination. Therefore, energy and momentum
must be accounted for within light itself. Maxwell (1891) first applied elec-
tromagnetic wave theory to predict and calculate the radiation pressure of
light. This prediction was soon verified by the observation of light pressure
on reflectors in vacuum by Lebedew (1901) and Nichols and Hull (1903).
Poynting (1905) quantitatively validated the theory with detailed measure-
ments of radiation pressure. Such experiments were relegated to scientific val-
idation of the electromagnetic wave theory until the invention of the laser.
Ashkin (1970) first demonstrated the manipulation of small dielectric
particles by laser light when particles suspended in water were drawn into
the axis of a laser beam and accelerated in the direction of propagation.
Rough calculations for the radiation pressure were used for designing the
experiments which involved a laser focused upon spheres a few wavelengths
in diameter. In addition to radiation pressure, the particles experienced a
trapping force that pulled the particles into the laser beam. This force was
reversed when the relative index of refraction was inverted. Air bubbles,
with dielectric permittivity less than that of the surrounding water, were
observed to be pushed out of the beam, while the radiation pressure
remained in the beam propagation direction. The trapping force and radi-
ation pressure are commonly known as the gradient force and the scattering
force, respectively, due to the separation based on the ray optics approxima-
tion for large particles or the Rayleigh approximation for small particles.
Optical levitation (Ashkin & Dziedzic, 1971, 1976) and radiation pres-
sure on a liquid surface (Ashkin & Dziedzic, 1973) were among the many
experiments that followed. Lasers were also proposed as a means to trap
atoms (Ashkin, 1978; Ashkin & Gordon, 1979), and single atoms were even-
tually trapped and cooled by optical tweezers (Chu, Bjorkholm, Ashkin, &
Cable, 1986; Chu, Bjorkholm, Ashkin, Gordon, & Holberg, 1986; Chu,
Holberg, Bjorkholm, Cable, & Ashkin, 1985). Colloidal dielectric particles
were also trapped in three dimensions using a single highly focused laser
beam (Ashkin, Dziedzic, Bjorkholm, & Chu, 1986). Such optical tweezers
result when a the gradient force is larger than the scattering force. Optical
manipulation has since found many application in physics, biology, chem-
istry, and medicine (Ashkin, 1997, 2000), and optical tweezers have become
a necessary tool for many researchers in microbiology (Ashkin &
Dziedzic, 1987).
The description of optical manipulation was further expanded by Burns,
Fournier, and Golovchenko (1989) when optical binding was first observed.
Optical binding can be used to create optical matter, a lattice of dielectric
Macroscopic Theory of Optical Momentum 439
Abraham 2 2 r + E + J B + @ fG
12 E rE 12 H Ag
M G
@t
r PE + @@tP B + r M
B + E + J B
Amperian
440 Brandon A. Kemp
2. CLASSICAL ELECTRODYNAMICS
To address energy, momentum, and charge conservation in electro-
dynamics, we use the subsystem mathematical framework. This is necessary
Macroscopic Theory of Optical Momentum 441
@Wj r , t
j r ,t r Sj r , t (3a)
@t
@ G j r , t
fj r , t r Tj r , t (3b)
@t
where fj is the force density, j is the power density, T j is the momentum
flux or stress tensor, Sj is the power flux, G j is the momentum density, and
Wj is the energy density. This indicates that each j may represent any sub-
system such as electromagnetic, hydrostatic, thermodynamic, kinetic, etc.
Closing the overall system, such that
X
j r , t 0 (4a)
j
X
fj r ,t 0, (4b)
j
@j r , t
0 r Jj r ,t , (5)
@t
where Jj r ,t and j r , t denote a current density and charge density for a
given formulation, respectively.
Kemp (2011), and Penfield and Haus (1967) for additional information
regarding these and other, less used, formulations.
r ,t @
r H Dr , t J r ,t (6a)
@t
r ,t + @ B
r E r ,t 0 (6b)
@t
r ,t r , t
r D (6c)
r , t 0,
r B (6d)
combine the field and material contributions within the time and space
dependent field vectors E r ,t, H
r , t, D r ,t, which contain
r , t, and B
the response of the material via macroscopic constitutive relations. The free
charge density r ,t and free current density J r ,t are the only field-free
quantities present when considering the Minkowski formulation. In a linear,
isotropic medium, which is effectively dispersionless at frequencies of interest,
r , t EE
the constitutive relations D r , t and B
r ,t H r , t yield the
terms for the momentum continuity equation in the Minkowski formulation
1 r ,trE 1 H r ,t H
r ,tr
fM r , t E r ,t E
2 2 (7a)
r , t + J r , t B
+ r ,tE r , t
1 r ,t + B
r ,t H
r , t I
TM r ,t D r ,t E
2 (7b)
D r ,t B
r ,tE r , tH r ,t
G r ,t B
M r , t D r , t, (7c)
r ,t E
where 12 E r ,trE is the well-known Helmholtz force. The
corresponding energy continuity equations are defined by
M r , t J r ,t E r ,t (8a)
SM r ,t E r , t H r ,t (8b)
1
WM r , t D r , t Er , t + B
r ,t H
r ,t : (8c)
2
In Section 4, the consequences of dispersion on the momentum of the wave
are explored. A statement of charge conservation results by taking the
Macroscopic Theory of Optical Momentum 443
medium. Here, the subscript C indicates that the values involved within the
Chu formulation differ from similar terms in other common formulations.
The momentum continuity equation in the EH representation is defined by
the terms
feh r ,t e EC + h H C Jh E0 EC
C + Je o H (12a)
1
Teh r ,t E0 EC EC + 0 H C H C I E0 EC EC 0 H
C HC (12b)
2
eh r ,t E0 0 EC H
G C , (12c)
and the corresponding terms in the energy continuity equation are
eh r , t Je EC + Jh H
C (13a)
Seh r ,t EC HC (13b)
1
Weh r , t E0 EC EC + 0 H C H
C , (13c)
2
where the subscript eh denotes quantities resulting from the EH representa-
tion which are quadratic in the fields. In these equations, the exact response
of the material has been left open. That is, the material could be nonlinear,
anisotropic, or dispersive. The Chu formulation combines the bound and
free electric and magnetic charges to conserve the charge within the system.
Charge conservation can be expressed in terms of the total charge and
current densities
@e r , t
0 r Je r ,t (14a)
@t
@ r ,t
0 r Jh r ,t h : (14b)
@t
separate the electric EA r , t and magnetic BA r , t fields from the material
response and is, therefore, often referred to as the EB representation. The
Amperian formulation models the material responses by equivalent electric
current density Jeb r ,t and electric charge density eb r ,t. These quantities
for moving media are defined as
@ A r ,t v r , t
Jeb r , t P A r ,t 0 E0 M
@t (16a)
+ r M A r , t + PA r ,t v r , t + J A r , t
eb r ,t r PA r ,t 0 E0 M A r , t v r , t + A r , t, (16b)
@ r , t
0 r Jeb r ,t eb : (19)
@t
446 Brandon A. Kemp
1 fE H
hSi g, (21)
2
where * denotes complex conjugation. The incident and reflected electric
and magnetic fields are
E i xE
^ 0 eiki z + i t + Reikr zr t (22a)
ik z + t
i ^
H yH0 e i i
Reikr zr t , (22b)
Macroscopic Theory of Optical Momentum 447
p
where H0 E0 E0 =0 , ki i/c, and kr r/c. The boundary condition
z^ E + 0 at z vt yields the phase matching condition
v B
kiv i krv r and the reflection coefficient
+1
R , (23)
1
where v/c. Note that the Doppler effect is a consequence of phase
matching at a moving boundary
1+
r i : (24)
1
Since the incident region is vacuum, there is no ambiguity as to the appli-
cable energy and momentum equations. The incident and reflected power
flow are
1
i z^1 E0 H0 zhS
hSi i Ei H ^ ii (25a)
2 2
1
i z^1 E0 H0 jRj2 zhS
hSr i Ei H ^ r i: (25b)
2 2
The total electromagnetic power flow into a closed volume at the reflector
interface is
hSi i
hPelec i hSi i hSr i 4 : (26)
c 1 2
The change in stored electromagnetic energy is
v nE0 0 o 1 + 2
hPstored i E E + H H 2hSi i : (27)
2 2 2 1 2
Therefore, the mechanical work that must be done to the system to keep the
reflector at constant velocity is
1 +
hPmech i hPstored i hPelec i 2hSi i : (28)
1
Likewise, the force on the reflector is
hF elec i z^hTi i + hTr i + vhGi i + hGr i, (29)
where the z^ components of the momentum densities and the z^z^ compo-
nents of the stress tensors are
448 Brandon A. Kemp
1
hGi i i hS i i
E0 0 Ei H (30a)
2 c
1
hGr i E0 0 Er H r hS r i (30b)
2 c
1 nE0 0 o hSi i
hTi i Ei Ei + H i H i (30c)
2 2 2 c
1 nE0 0 o hSr i
hTr i Er Er + H r H r : (30d)
2 2 2 c
The mechanical force required to keep the reflector from accelerating is
hSi i 1 +
hF mech i hF elec i z2
^ : (31)
c 1
Obviously, v hF mech i hPmech i so that the mechanical force and work are
required to keep the reflector at constant velocity. The case of v ! 0
produces the well known result hF elec i z2hS ^ i i=c, as is obtained from
the stationary reflector case treated in most electromagnetic textbooks
(Kong, 2005).
Since hPmechi is proportional to v, it will go to zero under a stationary
material approximation, but lim v!0 hF mech i z2hS^ i i=c. If the mechanical
balance force is removed, the reflector will accelerate from v 0 and the
kinetic energy of the mirror will be increased. However, this problem is typ-
ically treated using a stationary media approximation such that the Doppler
shift is ignored. It is equivalent to taking a quasi-stationary approximation
where v ! 0 while m ! 1 such that the kinetic momentum of the material
mv remains a nonzero number and the kinetic energy of the material 12 mv v
goes to zero.
e0 , m0
Ei
ki
x
n
z
y
z=0
Figure 1 A plane wave incident upon a perfect electrical conductor. The Fresnel reflec-
tion coefficient of the reflector in the stationary frame is R 1.
EC EA E, and likewise for all other fields and sources. Because of this,
we may simply state the polarization and magnetization in terms of the field
vectors such that
P r , t E0 Er , t
r ,t D (35a)
r ,t B
0 M r , t H
r , t, (35b)
0
which are general enough to include causal (i.e., lossy and dispersive) and
anisotropic media.
Obviously, the Minkowski, Chu, and Amperian force densities given in
Table 1 correspond to the corresponding formulations for stationary media.
Therefore, the stationary force densities in Table 1 may be used only when
the velocity of the material can be safely ignored, such as when relativistic
effects are insignificant. Two other equations, the Einstein-Laub and the
Abraham force densities, are also given in Table 1. Both have been discussed
extensively (see for example reviews by Pfeifer et al. (2007) and Milonni and
Boyd (2010). Here, Equation (35) are used to rearrange terms in the formu-
lations given in Section 2.1 to arrive at the Einstein-Laub and the Abraham
force densities. These mathematical rearrangements are exact and consistent
with Maxwells equations. However, such excercises shouldnt be taken as
basis for proof or interpretation (Grzegorczyk & Kemp, 2008).
Einstein and Laub (1908) proposed a formulation which models both
dielectric and magnetic response as effectively bound dipoles. This differs
from the Chu formulation, which models media as effectively bound electric
and magnetic charges, and the Amperian formulation, which models media
as bound electric charges and infinitesimal current loops. However, at the
macroscopic level, these microscopic viewpoints are effectively lost, and
the real difference is in which field terms give W, S, and T defining
G,
the electromagnetic subsystem for the formulation. Using vector calculus
identities to write the force on bound electric charges and bound magnetic
charges
E P rE r PE
r P (36a)
H
0 r M 0 M rH H
0 r M (36b)
allows for substitution into the Chu force given in Table 1. The tensor terms
are then moved to the other side of the momentum continuity equation
such that
Macroscopic Theory of Optical Momentum 451
@ P 0 @ M E0 E
+ P rE + 0 M r H + J + 0 H
@t @t
@
1
E0 0 E H r E0 E E + 0 H H I D E B H ,
@t 2
(37)
where the relations in Equation (35) have been used. Equation (37) is the
momentum continuity equation for the Einstein-Laub formulation. The
energy continuity equation is reported as identical to Chu (Mansuripur
et al., 2013).
Likewise, terms can be rearranged from the Minkowski formulation to
arrive at the Abraham formulation. This is accomplished by simply adding
the term known as the Abraham force @G M G A =@t to the Minkowski
force density and subtracting the same term from the momentum density
to arrive at the momentum continuity equation
1 2 1 2 @
E rE H r + E + J B + G
M GA
2 2
@t (38)
@
1
E0 0 E H r D E + B H I DE B H ,
@t 2
momentum densities and stress tensors in Table 2. The form of the energy
and momentum continuity equations are given in Equations (3a) and (3b).
The Minkowski, Chu, and Amperian energy terms have been given in
Section 2.1. The Einstein-Laub energy continuity terms are identical to
the Chu formulations, and the Abraham energy continuity terms are
identical to the Minkowski formulation. In general, torque density can be
calculated as r f, with the exception of the Einstein-Laub torque den-
sity which is supplemented by the terms P E + 0 M H (Einstein &
Laub, 1908; Mansuripur et al., 2013).
Z I
@ T r , t:
F t dV Gr , t dA (39)
V @t A
(a) (b)
e 0, m0 e b, mb
e,m e,m
Figure 2 Diagram for illustration of total force calculation. (a) The total force on an
object in vacuum is computed by volume integration of a force density within a region
which completely encloses the material object. The volume is enclosed by the surface
depicted by the dashed line. (b) The total force on an object embedded is computed in a
similar way by considering a thin vacuum region between the two materials.
454 Brandon A. Kemp
will produce identical results for the total force on an object at all points in
time. For example, the Chu, Einstein-Laub, and Abraham force densities are
equivalent in terms of total force. Since the Amperian formulation differs
only in the modeling of magnetic media, it will also produce identical results
for total force for dielectric materials.
As a second case, consider an object which is submerged or embedded
inside another medium as depicted in Figure 2b. In this case, the total force
will still be the same for two formulations if their momentum densities are
identical, but it is important to clarify how the force is computed. If the force
is computed by integrating the force density fr , t over a volume, there will
generally be both volume and surface forces. The surface forces arise due to
material discontinuities at the boundaries (Kemp, Grzegorczyk, & Kong,
2006a; Mansuripur, 2004). Therefore, there may be initial ambiguity as
to which material, the embedded object or the embedding medium, to
which the surface forces should be assigned. To alleviate this ambiguity, ima-
gine there is a thin layer of vacuum in the boundary. A volume integration
procedure can be applied, as before, to the embedded object. Since the sur-
face of the integration is in the vacuum region, the tensor reduces to the
vacuum stress tensor and the force separation for the two media is unambig-
uous. Allowing the vacuum region to vanish, or mathematically taking the
limit of the gap to approach zero, yields the original field problem with the
forces correctly separated (Kemp & Grzegorczyk, 2011). In this case, how-
ever, there will generally be forces on both the embedded object and the
surrounding medium. The conclusion remains the same; any formulations
which share a common momentum density will produce identical results for
the total electromagnetic force on a material object at any point in time.
the five leading formulations reviewed. First, the Abraham momentum den-
sity, G A E0 0 E H
shows up in the Chu, Einstein-Laub, and Abraham
formulations. Second, the Livens momentum density G L E0 E B (see
references by Livens (1918), Nelson (1991), and Scalora et al. (2006)) is
included in the Amperian subsystem. Third, the Minkowski momentum
density G M D B completes the Minkowski momentum continuity
equation. For each form, the momentum of an electromagnetic pulse trans-
mitted into a medium can be written in terms of vacuum quantities. In this
way, the various forms can be compared and contrasted (Kemp, 2011).
For comparison, consider a plane wave pulse normally incident upon a lin-
p
ear, transparent material with index of refraction n c E. For simplicity, the
material is considered to be nondispersive in the frequency range of interest and
the quasi-stationary approximation is applied. The incident pulse is described
by the electric field E xE
^ 0 z, tcosk0 z t. Here, E 0(z, t) is an envelope
function for the harmonic wave cos k0 z t and k0 is the wavenumber in the
incident vacuum region. The incident energy density Wi z E0 E 20 z, t and
momentum density G i z, t E0 0 Ei z, t H
i z, t zW
^ i z, t=c are
unambiguous since the wave is incident from vacuum. The incident optical
energy is defined as the volume integral over the entire incident pulse. The
one-dimensional pulse varies only in z so that the incident energy is
ZZZ Z
Ei dVWi z dzWi z (44)
V
2
and has units (J/m ). Similarly, the unambiguous incident momentum is
determined by integrating the incident momentum density at an instant
prior to the field interaction with the material. It is given by
ZZZ Z
Wi z Ei
pi dV G i z z^ dz z^ , (45)
V c c
and has units (N s/m2).
We may also arrive at quantities for the reflected energy density Wr(z)
R Wi(z) and reflected momentum density Gr(z) R2Wi(z)/c, where R2 is
2
the reflectivity (Kong, 2005). The reflected energy in terms of the incident
energy is
Er R2 Ei , (46)
and the reflected momentum in terms of the incident momentum is
pr R2 pi : (47)
Macroscopic Theory of Optical Momentum 457
t z^0 T 2 G
A z E0 0 Et H
G i z z^Wi z 1 R2 , (50)
c
where 0 is the wave impedance of free space, is the wave impedance of the
material, T is the transmission coefficient, and the relation (0/)T2 1 R2
resulting from the boundary conditions has been used. The Livens momen-
tum density is given by
L z E0 Et Bt E0 0 Et H
G t + E0 0 Et M
t
Wi z t Et , (51)
z^ 1 R2 E0 0 M
c
where the last term is the hidden momentum as coined by Shockley and
James (1967). Similarly, the transmitted Minkowski momentum density is
M z EEt H
t zn i z zn Wi z
G ^ 2 0 T 2G ^2 1 R2 : (52)
c
The spatial length of the transmitted pulse is decreased proportional to
the factor n1 due to the change in velocity. The Abraham momentum
of the transmitted pulse is
Z
pA dzG A z, t z^1 Ei 1 R2 z^ 1 Eexc : (53)
z nc nc
The Livens momentum of the transmitted pulse reduces to
Z Z
1 t Et ,
pL dzGL z,t z^ Eexc dzE0 0 M (54)
z nc z
which differs from the Abraham form by the hidden momentum. Finally,
the Minkowski momentum density yields the momentum
458 Brandon A. Kemp
Z
M z,t z^ 1 n2 Ei n
pM dzG 1 R2 z^ Eexc (55)
z n c c
transmitted into the medium. Consequently, both the direction and magni-
tude of the momentum transferred to the surface depend upon this choice
according to Loudon (2004). In more exact mathematical terms, this means
that the subsystem required to close the overall system depends upon which
formulation of electromagnetics is applied.
Ei 1
pm 1 : (56)
c n
Momentum conservation requires that the electromagnetic contribution to
the total momentum be the difference between the total momentum of the
incident pulse and the material momentum given by Equation (56).
Macroscopic Theory of Optical Momentum 459
(a)
n = c me
d
(b)
Figure 3 Diagram of Balazs thought experiment for determining the kinetic momen-
tum of light. An electromagnetic wave (a) traveling through an impedance-matched
magnetodielectric slab is delayed by a length L (n 1)d in comparison to (b) an iden-
tical, unimpeded pulse traveling through vacuum.
thought experiment does not imply that the Amperian formulation violates
momentum conservation. It does mean, however, that the Amperian for-
mulation is not considered as a possible choice for the kinetic subsystem
of light. It must be concluded that the macroscopic formulation of electro-
dynamics presented as the Amperian formulation (i.e., EB representation)
does not provide the field only contribution to energy and momentum.
Although, microscopic reasoning can be applied to justify the inclusion of
the hidden momentum model (Griffiths & Hnizdo, 2013), the hidden
momentum term cannot be simply added to the momentum density in
the Amperian formulation without being accounted for equally in either
the force density term or stress tensor term of the formulation. This math-
ematical rigour is necessary to ensure that the macroscopic momentum con-
tinuity equation is consistent with Maxwells equations. As previously
demonstrated, such mathematical rearrangements defines a new subsystem,
but this author is unaware of such a formally proposed formulation and
interpretation. Such a subsystem would effectively include the Abraham
momentum density with a modified version of the Amperian force density
and/or stress tensor. It should also be reiterated that the results of this section
make no claims regarding the microscopic nature of matter response.
Although the different formulations provide conceptual views of material
response (e.g., effective dipoles, bound charges, or current loops), at the
macroscopic level these details are lost (Kong, 2005; Penfield & Haus,
1967), and the important features of the macroscopic theories pertain to
which macroscopic fields appear in the continuity equations.
E0 xE
^ i eikz + Reikz (58a)
0 y^Hi eikz Reikz :
H (58b)
The excitation power flow is the power available to the medium
1
0 1 Ei Hi 1 jRj2 1 jRj2 hSi i,
hSexc i n^ E0 H (59)
2 2
where hSi i 12 fEi Hi g is the power flow of the incident wave. The reflec-
tion coefficient is R (n 1)/(n + 1). Therefore, the excitation power flow
can be written as
" #
n + 12 n 12 4n
hSexc i hSi i < Si > njT j2 hSi i, (60)
n + 1 n + 1 n + 12
1 1 hSexc i
hTexc i n + : (63)
2 n c
Inside the dielectric, both the Chu and Einstein-Laub formulations yield the
electromagnetic stress
1 nE0 2 0 2 o 1 1 1 E t Ht
hTeh i jEt j + jH t j n+
2 2
2 2 2 n c
(64)
1 1 hSexc i
n+ :
2 n c
462 Brandon A. Kemp
The electromagnetic stress appears as the average of n/c and 1/nc times the
excitation power flow. In this form, the transmitted momentum flow may
be written as the average of the transmitted Abraham and Minkowski
momenta (Mansuripur, 2004). However, it is noted that the transmitted
momentum flow given in Equation (64) is equal to the excitation momen-
tum in Equation (63). Therefore, the EH representation of the stress tensor
in media yields a momentum flow which is equal to the excitation momen-
tum flow at normal incidence. Both the Chu and Einstein-Laub formula-
tions yield zero surface pressure in this case as the net momentum flow is
conserved.
As a second example, consider an impedance-matched material such that
p p p
n c E and =E 0 =E0 . The reflectivity is zero in this case. The
excitation momentum flow is simply given by the incident momentum flow
1 nE0 2 0 2 o 1 Ei Hi hSi i
hTexc i jE0 j + jH 0 j : (65)
2 2 2 2 c c
The transmitted momentum flow given by either the Chu or Einstein-Laub
formulations is
1 nE0 2 0 2 o 1 Et Ht hSi i
hTeh i jE t j + jH t j : (66)
2 2 2 2 c c
In this case, the EH representation yields zero force on the surface since the
momentum flow is conserved.
p
in the dielectric fluid is v c/n, where n c 0 E. The electric and magnetic
fields are
1 nE0 2 0 2 o 1 1 hS0 i
he zi E + H n+ , (68)
2 2 0 2 0 2 n c
where hS0 i 12 E0 H0 .
The material stress is determined by closing the system. Inside the fluid,
the electromagnetic force is balanced by an equal and opposite force of the
material fe + fm 0. The material stress is written as fm fe rm . Con-
sidering the one-dimensional problem, the material stress at t 0 is
Z L Z L
@P
m z dzf e z^ dz 0 H z: ^ (69)
0 0 @t
Because the ramp is approximately linear, the @ P=@t term is constant over
the span [0, L] and can be approximated as P0v/L, where P0 E0(n2 1)E0.
The 0 H term is linear, so the integration yields 1 0 H
L. The material stress
2
reduces to
1
P0 0 1 S0
m z v H0 L n : (70)
L 2 2 n c
Taking the time-average in the region z < vt (i.e., once the time-harmonic
plane wave has been well established) yields
1 1 hS0 i
hm zi n : (71)
2 n c
464 Brandon A. Kemp
This gives the material response contribution to the average stress of a plane
wave. Adding the material and electromagnetic components gives the total
stress or momentum flow associated with the plane wave h(z)i nhS0i/c,
which is consistent with the Minkowski momentum. For comparison, eval-
uation of the Minkowski stress tensor in the region z < vt yields the total
stress associated with the fields and material response
1 nE o hS0 i
hMin zi E02 + 0 H02 n : (72)
2 2 2 c
2.5 |E1| = {1 + R }
|H1| = {1 R }
2
Fields
1.5
0.5
0
0 20 40 60 80 100 120 140 160 180
f ()
Figure 4 Normalized fields in a dielectric fluid at the surface of a perfect reflector as a
function of the phase of the mirror reflection coefficient (Rmirror ei. The electric field
(solid line) and magnetic field (dashed lines) are out of phase and are plotted as
jE1j {1 + R} and jH1j {1 R}, respectively.
n cos kd + =2 + i sin kd + =2
R , (74)
n cos kd + =2 i sin kd + =2
where d 0 for the submerged mirror case. The electric and magnetic fields
form out of phase interference patterns. Figure 4 shows that the electric field
is zero at the surface of a PEC ( ) while the magnetic field is zero at the
surface of a PMC ( 0).
The electromagnetic stress is determined from the Tzz component of the
time-averaged Chu or Einstein-Laub stress tensor. Substitution of the fields
in the dielectric region z < 0 yields an equation for the time-averaged elec-
tromagnetic stress in the dielectric fluid in terms of the index of refraction
1 1 i2kz hS0 i
he zi n + n Re : (75)
n n c
where T m represents the local material stress, and the momentum densities
vanish under time averaging. Since the fluid is nonmagnetic and the wave is
normally incident, the force reduces to a single term
1 1
h fe zi iP1 0 H
1 E E0 0 iE1 H
1 : (77)
2 2
The electromagnetic force acts only only in the
z^ direction, allowing the
use of scalar quantities. The material stress is the integral of the electromag-
netic force
Z
hm zi h fe zidz + h0 i: (78)
Here, the first term is the stress due to the interference of the standing wave
pattern and h0i is the stress due to the individual plane waves. The constant
h0i is twice the value given in Equation (71) since there are counter prop-
agating plane waves present, each contributing equally to the material
response stress. Therefore,
1 hS0 i
h0 i n , (79)
n c
and the time-average material stress reduces to
1 1 i2kz hS0 i
hm zi n + n Re : (80)
n n c
The total stress in the fluid is the sum of the electromagnetic stress in
Equation (75) and the material response stress in Equation (80). The sum,
hS0 i
htotal i he zi + hm zi 2n , (81)
c
gives twice the Minkowski momentum as the observed pressure on the mir-
ror with respect to the submerging fluid regardless of the phase of the reflec-
tion coefficient used. The contributions are plotted in Figure 5 for n 2. In
the case of the PEC reflector ( ) as used in the JRL experiment, the
material stress does not contribute. This is because the fields are zero at
the surface as shown in Figure 4, yielding zero material energy and momen-
tum contributions at z 0. Therefore, any formulation will directly yield
identical results for the pressure on the mirror, which means the JRL exper-
iment does not actually provide a true test for the momentum of light.
Macroscopic Theory of Optical Momentum 467
5
PMC (f = 0) PEC (f = 180)
4
Stress (N/m2)
3
EM
Material
2 Total
0
0 20 40 60 80 100 120 140 160 180
f ()
Figure 5 Stress in a dielectric fluid at the surface of a perfect reflector as a function of
the phase of the mirror reflection coefficient (Rmirror ei). The electromagnetic pres-
sure (triangles) and material pressure (circles) sum to give a constant value 2nhS0i/c for
the total stress regardless of reflector phase. The incident momentum is normalized
hS0i/c 1 and the index of refraction is n 2.
The PMC case ( 0) is different. Figure 4 shows that the electric field is
nonzero at the mirror surface, and the material contributes to both the local
energy and momentum of the wave. Adding these contributions restores
the Minkowski pressure to the reflector (Kemp, 2012; Kemp &
Grzegorczyk, 2011).
@ @
+ e + e0 P E0 2ep E
2
(83a)
@t2 @t
2
@ @ F2 H,
+ m + m0 M
2
(83b)
@t2 @t mp
where the parameters of the equations have their usual meanings (Cui &
Kong, 2004). To derive the energy of the electromagnetic wave, the mate-
rial Equations (83a) and (83b) are dot multiplied by Je and Jh , respectively.
The resulting equations
1 @ @ P @ P 2 e @ P @ P
Je E + P P + (84a)
2E0 2ep @t @t @t e0
E0 2ep @t @t
@ @ M @ M M + m 0 @ M @ M
Jh H
0
+ m0 M
2
2Fmp @t @t
2 @t Fmp @t
2 @t
(84b)
are then added to the energy conservation equation of the electromagnetic
subsystem given by the Chu formulation. The resulting energy
conservation equation for the electromagnetic wave is in the form of
energy density W, and energy dissi-
Equation (3a) with the energy flow S,
pation given by
Macroscopic Theory of Optical Momentum 469
S E H (85a)
E0 0 1 @ P @ P 2
W E E + HH+ + e0 P P
2 2 2E0 2ep @t @t
(85b)
0 @M @M 2
+ + m0 M M
2F2mp @t @t
e @ P @ P m 0 @ M @M
+ : (85c)
E0 ep @t @t Fmp @t
2 2 @t
The energy flow S in Equation (85) retains its free-space form even in the
presence of a lossy, dispersive material. Also, the energy density W contains
contributions from the potential energy and kinetic energy of the electric
and magnetic dipoles. Furthermore, the form of Equation (85b) has been
regarded as significant since the energy density remains positive in negative
index of refraction media (Cui & Kong, 2004). Furthermore, the energy dis-
sipation term depends upon the damping factors e and m in
Equation (83). Therefore, 0 in the limiting case of a lossless material,
which indicates the energy of the electromagnetic wave is conserved.
The wave momentum can be derived by a similar method. The material
dispersion Equation (83) are dot multiplied by the dyads rP and 0 rM,
respectively. The resulting vector equations are then added to the electro-
magnetic continuity equation given by the Chu formulation to yield
eh
@G
r Teh + + f + E rP + 0 H M
@t eh
2
@ 2 P 1 @ M 2 0
rP + 2
P r M + M
@t2 e0
E0 2ep 0
@t2 m0
F2mp
@ P e
rP + r M @ M m 0 :
@t E0 2ep @t F2mp
(86)
Vector calculus allows terms to be rewritten as
2
"
#
@ P 1 1 @ P @
P
rP + e0 P
2
r e0 P P
2
@t2 E0 2ep 2E0 2ep @t @t
" #
@ 1 @
P
rP
@t E0 2ep @t
(87a)
470 Brandon A. Kemp
"
#
2
@ M @ M @ M M
rM + m0 M
2 0
r 0
m0 M
2
@t 2 F2mp 2F2mp @t @t
" #
@ 0 @ M
rM
@t F2mp @t
(87b)
f eh + E rP + 0 H M r P E + 0 M H I P E 0 M H
@
+ D B E0 0 E H :
@t
(87c)
By combining Equation (86) and Equation (87), the momentum continuity
equation for the wave can be expressed in the form of Equation (3b) with the
stress tensor, momentum density, and force density
"
1 1 1 @ P @ P 2
T D E + B HI D E B H + P P
2 2 E0 2ep @t @t e0
@M @M M + P E + M H
I (88a)
+ 02 2m0 M
Fmp @t @t 0
D B 1 rP @ P 0 rM
G @M (88b)
E0 ep
2 @t Fmp 2 @t
@ P m 0
f e 2 rP 2 rM @M : (88c)
E0 ep @t Fmp @t
The expressions in Equation (88) define the quantities for the momentum
continuity equation, which complete the wave subsystem along with the
energy expressiongs in Equation (85). The momentum density G contains
the Minkowski momentum D B plus material dispersion terms. Likewise,
the momentum flow T is the Minkowski stress tensor plus dispersive terms.
Note that the momentum dissipation term f depends upon the damping
factors e and m in Equation (83). Therefore, f 0 in the limiting case
of a lossless, unbounded material, which indicates that the momentum of
the electromagnetic wave is conserved.
The constitutive parameters follow directly from Equation (83) given eit
dependence (Kong, 2005). The complex permittivity and permeability,
!
2ep
E E0 1 2 (89a)
2e0 + i e
!
F2mp
0 1 2 , (89b)
2m0 + i m
are functions of the frequency and consist of real and imaginary parts den-
oted by E ER + iEI and R + iI. Likewise, the time-average of the
squared polarization and magnetization are
E20 4ep
2
jPj 2
jEj (90a)
2
e0 + e
2 2 2 2
F 2 4mp
j2
jM jH j2 : (90b)
2
m0 + m
2 2 2 2
zero upon cycle averaging. That is, h@W/@ti 0, and the resulting conser-
vation equation
1 2 + I jH
j2
hr Si EI jEj (93)
2
is generally regarded as the complex Poyntings theorem, where
g is the time average Poynting power.
1 fE H
hSi 2
Similarly, the average momentum density,
(
1 E0 2ep
hGi D B + k 2
jEj
2 e0 + e
2 2 2 2 2
) (94)
0 Fmp 2
+ k j2 ,
jH
2 2m0 2 + 2m 2
is obtained from Equation (88b). It is simple to show using Equation (89)
that the average momentum given in Equation (94) satisfies (Veselago, 1968)
@E 2 @
1
EE H
+ k + j
hGi jEj jH 2
(95)
2 2 @ @
when the medium is lossless. The time average stress tensor is
1 1
hT i D E + B H I D E B H (96)
2 2
since the dispersive terms in Equation (88a) tend to zero upon cycle aver-
aging. Since the average rate of change in momentum density is zero (i.e.,
h@ G=@ti 0), the momentum conservation theorem for a monochromatic
wave reduces to
1
hr T i fEI E B I H g
D (97)
2
where the tensor is the complex Minkowski tensor. The right-hand side of
Equation (97) is defined as the force density on free currents by Loudon,
Barnett, and Baxter (2005) and Kemp, Grzegorczyk, and Kong (2006b).
5. DISCUSSION
Several conclusions can be made from the preceding sections. Each
conclusion adds to the logical deduction of how optical manipulation
Macroscopic Theory of Optical Momentum 473
5.1 Conclusions
The conclusions are ordered by section number so that a logical flow is pres-
ented and reference to the supporting analysis is maintained.
Conclusions 1: A number of electromagnetic force density equations
persist in the optics literature. Each is tied to a particular formulation of elec-
trodynamics. The most commonly used force equations are considered
herein with the goal of modeling modern experiments involving the manip-
ulation of macroscopic media with light.
Conclusions 2.1: Unique formulations of electrodynamics differ in
how matter is modeled with each resulting in various forms for the electro-
magnetic energy and momentum continuity equations. Within media, the
force density, momentum density, and stress tensor are determined by which
representation of Maxwells theory is applied. When a medium is in motion,
the field vectors are also unique to a particular formulation. In vacuum, the
fields as well as the energy and momentum continuity equations are
unambiguous.
Conclusions 2.2: A quasi-stationary approximation can be applied that
allows for the study of optical momentum transfer to media while optical
energy is conserved. This approximation is common and greatly reduces
the complexity involved with modeling optical manipulation experiments.
Essentially, the Doppler shift of the electromagnetic wave due to material
motion is ignored such that calculations predict momentum exchange
due to reflection or scattering, but optical energy is conserved.
Conclusions 2.3: The most commonly used electromagnetic force
equations assume stationary media, and are each associated with a momen-
tum continuity equation. The sum of the force density, divergence of the
stress tensor, and rate of change of the momentum density must be zero.
Under the quasi-stationary approximation, the field variables are equivalent
among the formulations in vacuum and in media. Terms may be rearranged
between the force density, momentum density, and stress tensor. Such exer-
cises define new electromagnetic subsystems, but the resulting energy and
momentum continuity equations are left for interpretation.
Conclusions 3.1: Any formulations which share a common momen-
tum density will produce identical results for the total electromagnetic force
474 Brandon A. Kemp
on a material object at any point in time, and all formulations produce iden-
tical values for the total time-average electromagnetic force on a material
object. This statement is true as long as the momentum continuity equations
are consistent with Maxwells equations. Additionally, it is impossible to
infer a time-domain force from a time-averaged force.
Conclusions 3.2: The total calculated electromagnetic force on an
object due to an optical pulse of finite extent is due to the time rate of change
of the momentum contained within the pulse. The momentum contained
within a pulse depends upon the momentum density that appears in the
momentum continuity equation. The Abraham momentum gives the
kinetic momentum of light, which is responsible for center of mass transla-
tions of media. Any subsystem that is to be considered as consisting of field
only quantities must include the Abraham momentum density.
Conclusions 3.3: The total average electromagnetic force on an object
due to time-harmonic fields is due to the divergence of the average stress
contained within the fields. The Chu and Einstein-Laub formulations
include the Abraham momentum density in the momentum continuity
equations. Both predict conservation of momentum flow due to normal
incident transmission and reflection at a boundary, both reveal that a plane
electromagnetic wave includes a nonzero average material response stress in
addition to the electromagnetic stress, and both predict an additional mate-
rial contribution to radiation pressure when an object is submerged in a
dielectric fluid. In regards to the last point, the observed optical force on
a submerged object is due to the electromagnetic force on the object and
the electromagnetic force on the submerging fluid. The pressure difference
at the boundary yields the Minkwoski momentum imparted to the sub-
merged object with respect to the submerging fluid, which is key to reduc-
ing the complexity of modeling optical manipulation experiments.
Conclusions 4.1: Causality requirements for media to be lossy and dis-
persive necessitate the inclusion of material dynamics in the wave energy and
momentum continuity equations. Starting with the Chu formulation
and modeling the material response as harmonic oscillators, the energy
and momentum response of the material can be added to the electromagnetic
energy and momentum continuity equation. The resulting wave energy den-
sity, power flux, momentum density, and stress tensor take the Minkowski
forms plus additional material dynamics terms.
Conclusions 4.2: Under time-harmonic excitation, the time-averaged
stress tensor for the electromagnetic wave, which includes both electromag-
netic and material response, reduces to the time-average Minkowski stress
Macroscopic Theory of Optical Momentum 475
tensor. Therefore, Conclusions 3.2 are valid for dispersive media when the
optical excitation has a narrow bandwidth. Otherwise, the conclusions hold,
but the Minkwoski stress tensor must be supplemented with additional terms
due to material dispersion.
where k2y k2b k2x , k2b 2 b Eb , and w is the beam waist. The electric field
in Equation (98) satisfies the wave equation (Kong, 2005) and is
implemented as a discrete sum of plane waves. The spectrum is actually lim-
ited, and the sum is implemented here within the domain jkx j NA c ,
where NA is the numerical aperture.
Figure 6 illustrates a model of the original experiments by Ashkin (1970).
The arrows depict the observed force on either a polystyrene particle or air
bubble in water due to a Gaussian beam described in Equation (98). The
divergence of the Minkowski stress tensor in the background water is
applied to calculate the time-average force on the 2D particle (i.e., cylinder)
at different points. This approach is consistent with the results of Sections 3.3
and 4.2. The forces in Figure 6 agree with the observed forces by Ashkin
(1970); radiation pressure on both particles is accompanied by a gradient
force that attracts the higher index of refraction particles (polystyrene) into
the center of the beam and repels the lower refractive index air bubbles
out of the center of the beam. Since the air bubbles have a polarization
of P 0, it is actually the kinetic force that pulls the water into the high field
intensity region of the beam. However, the pressure gradient that is
476 Brandon A. Kemp
Figure 6 Force per unit length (represented by the arrows) on a single infinite cylinder
due to an incident Gaussian beam with NA 1, E0 1, w 0.50, and wavelength
0 1064 nm. The background medium is water Eb 1.69E0 and the radius of the par-
ticle is a 0.50. The particle is (a) polystyrene and (b) air.
such that the so-called gradient force overcomes the scattering force. The
parameters match closely the experimental conditions used in the original
experiments to trap silica beads. Therefore, the modeling approach again
matches the experimental outcome as the optical force is seen to produce
a stable trap near the beam focus.
Optical manipulation of colloidal particles is typically achieved by crea-
tion of an optical intensity gradient as seen with optical tweezers. One such
configuration consists of three lasers with 2/3 incident angle separation
(Fournier et al., 2004). Such a configuration results in a zero net incident
momentum since the vector sum of the incident momentum vectors is zero,
which allows for isolation of the gradient force. This interference is repre-
sented by the background of Figure 8, where white denotes regions of
high field intensity. The force on a 2D polystyrene particle in water is cal-
culated via the divergence of the Minkowski stress tensor as prescribed in
Sections 3.3 and 4.2. Each arrow represents the direction and relative
magnitude of force on a dielectric cylinder placed at the tail of the arrow.
The force on the small particle with radius a 0.150 is seen to closely follow
the high intensity gradient of the incident field with stable optical traps
occurring in the high intensity regions. However, larger particles with radius
a 0.300 are repelled from the high intensity regions and find stable trap-
ping positions in the dark regions of the incident optical interference pattern.
Obviously, this result contrasts the simple description given by the gradient
force obtained in the Rayleigh approximation. The example demonstrates
that such an approximation yields results which are both quantitatively and
qualitatively incorrect for particles that are on the order of a wavelength. It is
also interesting to study the intermediate particle size regime. For example,
the force on a particle with radius a 0.2250 is also plotted over the back-
ground of the incident three plane wave interference pattern in Figure 8. An
interesting feature of this plot is that the force field appears to be non-
conservative, and a prediction of stable optical trapping in these vortices
has been reported by Grzegorczyk and Kong (2007).
A concluding example demonstrates the dependence of radiation pressure
on the background fluid. Consider a spherical particle surrounded by either
vacuum or a negative refractive index medium with negative vacuum param-
eters (E0, 0). The radiation pressure is reversed according to the theory of
Kemp et al. (2007), which is reproduced in Section 4. Figure 9 shows the
force on a sphere as a function of radius a, which is varied from the Rayleigh
regime to the Mie regime. The fields are calculated by the analytical approach
detailed in Appendix B. The results show that the force on the sphere
478 Brandon A. Kemp
Figure 8 Force per unit length (represented by the arrows) on a single infinite cylinder
due to the interference of three plane waves (represented by the background pattern)
of equal amplitude Ei 1 V/m and wavelength 0 532 nm. The incident angles of
the plane waves are {/2, 7/6, 11/6}. The background medium is water Eb 1.69E0,
and the cylinder is polystyrene Ep 2.56E0 with radius (a) a 0.150, (b) a 0.300,
and (c) a 0.2250.
depends upon the refractive index of the background medium. The particle
in vacuum is pushed, while the radiation pressure is reversed when the back-
ground refractive index is negative as first predicted by Veselago (1968).
1012
2
1.5
0 (eb,mb) = (e0,m0)
0.5
1.5
2
50 100 150 200 250 300 350 400 450 500
a (nm)
Figure 9 Radiation force on a spherical particle with respect to a submerging back-
ground medium as a function of sphere radius a. The particle has is nonmagnetic with
E 4E0. The solid line gives the results for a background medium of vacuum (E0, 0), and
the dashed line depicts the force for a hypothetical inverted vacuum (E0, 0). The
incident field is a plane wave with amplitude E0 1 V/m and vacuum wavelength
0 1064 nm.
The cylinder is characterized by (p, Ep) and the background by (b, Eb).
The incident wave is assumed to be a plane wave with eit dependence.
Many other field distributions, such as a Gaussian beam, can be described
by a sum of plane waves. Therefore, the total solution for such incident fields
can be described as a superposition of solutions resulting from a plane-wave.
The incident, scattered, and internal fields are expanded in cylindrical waves
given by N n, M
n , RgN
n , and RgM
n . The solution is given by Tsang, Kong,
and Ding (2000).
The incident electric field is polarized in the z-direction^ and
propagates in the plane (kz 0). The magnetic fields are obtained from
Faradays law
r E
iH (A.1)
using the identities
rM n
n kN (A.2a)
n kM
rN n: (A.2b)
480 Brandon A. Kemp
The incident fields are regular at the origin and are given by
X
N
Einc zE
^ 0 eik i n kb ,
an RgN (A.3a)
nN
kb X N
inc
H n kb ,,
an RgM (A.3b)
ib nN
X
N
Escat n kb ,
asn N (A.4a)
nN
kb X N
scat
H n kb ,:
as M (A.4b)
ib nN n
The internal fields are also regular at the origin and can be computed from
X
N
Eint n kp ,
cn RgN (A.5a)
nN
kp X N
int
H n kp , ,
cn RgM (A.5b)
ip nN
n k, zkH
N ^ n1 kein (A.6a)
n 1
M n k, ^ i Hn k ein
(A.6b)
n
+ ^ kHn + 1 k Hn1 k ein
1
RgN n k, zkJ
^ n kein (A.6c)
Macroscopic Theory of Optical Momentum 481
n
RgM n k, ^ i Jn k ein
(A.6d)
n
+ ^ kJn + 1 k Jn k ein :
Here, Hn1 is the Hankel function of the first kind and Jn() is the Bessel
function. The coordinates (, , ) represent the point for field evaluation
(i.e., the observer position). The angle i is used to represent the incident
direction of the illuminating wave.
For a general dielectric and magnetic medium, the boundary conditions
give a 2 2 system of equations that must be solved for the unknown
coefficients,
eini
an in E0 (A.7a)
kb
b1 m22 b2 m12
asn (A.7b)
m11 m22 m21 m12
b1 m21 b2 m11
cn : (A.7c)
m12 m21 m22 m11
The matrix elements and right-hand side for the linear system are
Einc xE
^ 0 eikb z xE
^ 0 eikb rcos (B.1a)
X
N
eikrcos in 2n + 1jn krPn cos , (B.2)
n0
where jn(kr) is a spherical Bessel function and Pn cos is the Legendre func-
tion. The approximation is exact as N ! 1.
The incident, scattered, and internal waves are decomposed into TM to
r^ and TE to r^ waves by using scalar potentials e and m respectively.
The potentials are defined by
A r e (B.3a)
H ^ 1 @ e ^ @ e
rA (B.3b)
sin @ @
for TM waves and
Z r m (B.4a)
1 @ @
E r Z ^ m ^ m (B.4b)
sin @ @
for TE waves. The potentials satisfy the Helmholtz equation in spherical
coordinates
r2 + k2 0: (B.5)
The scaler potentials are
X
N
e An zn krPn1 cos cos (B.6a)
n1
Macroscopic Theory of Optical Momentum 483
X
N
m Bn zn krPn1 cos sin , (B.6b)
n1
where zn(kr) represents solutions to the Bessel equation. The solutions differ
in the three regions by the following.
Incident ( e, m)
zn kr jn kb r
kb in 2n + 1
An E0
b nn + 1
b
Bn An
kb
Scattered ( se , sm )
zn kr hn kb r
kb
A n E0 an
b
Bn E0 bn
hn kb r h1
n kb r is the spherical Hankel function of the first kind.
internal ( ie , im )
zn kr jn kp r
kp
An E0 cn
p
Bn E0 dn
The unknown coefficients an, bn, cn, dn are found using the boundary con-
ditions for the tangential electric field and tangential magnetic field.
The total fields are:
k XN
Er i An Pn1 cos krzn kr + 2z0n kr + krz00n kr cos (B.10a)
E n1
XN
k @ 1 1 0
E iAn P cos zn kr + zn kr (B.10b)
n1
E @ n kr
P 1 cos
+ Bn n zn kr cos (B.10c)
sin
484 Brandon A. Kemp
N
X
k Pn1 cos 1
E iAn zn kr + z0n kr (B.10d)
n1
E sin kr
@ 1
+ Bn Pn cos zn kr sin (B.10e)
@
k XN
Hr i Bn Pn1 cos krzn kr + 2z0n kr + krz00n kr sin (B.10f)
n1
XN
k @ 1 1 0
H iBn P cos zn kr + zn kr (B.10g)
n1
@ n kr
Pn1 cos
An zn kr sin (B.10h)
sin
XN
k Pn1 cos 1 0
H iBn zn kr + zn kr (B.10i)
n1
sin kr
@
+ An Pn1 cos zn kr cos: (B.10j)
@
The spherical Bessel and Hankel functions zn() are given in terms of the
Bessel and Hankel functions Zn() by
r
zn Z 1 : (B.11)
2 n + 2
Z^n zn (B.17a)
0
Z^n zn + z0n : (B.17b)
ACKNOWLEDGMENTS
The author is grateful to Masud Mansuripur helpful discussions and, in particular, for pointing
out discrepancies in the literature regarding the Abraham stress tensor. The author
acknowledges Cheyenne J. Sheppard for proofreading parts of the manuscript.
This work was sponsored in part by the National Science Foundation EECS Division of
Electrical, Communications, and Cyber Systems (award number ECCS-1150514). The
authors would like to thank the Arkansas Science and Technology Authority for funding
provided to the Center for Efficient and Sustainable Use of Resources (CESUR).
REFERENCES
Abraham, M. (1909). Zur Elektrodynamik der bewegter K orper. Rendiconti del Circolo Mate-
matico di Palermo, 28, 128.
Abraham, M. (1910). SullElettrodinamica di Minkowski. Rendiconti del Circolo Matematico di
Palermo, 30, 3346.
Ashkin, A. (1970). Acceleration and trapping of particles by radiation pressure. Physical
Review Letters, 24, 156159.
Ashkin, A. (1978). Trapping of atoms by resonance radiation pressure. Physical Review Letters,
40, 729732.
Ashkin, A. (1997). Optical trapping and manipulation of neutral particles using lasers.
Proceedings of the National Academy of Sciences, 94, 48534860.
486 Brandon A. Kemp
Tunneling of Electromagnetic
Waves in All-Dielectric Gradient
Metamaterials
Alexander B. Shvartsburg1,2,4, Yuri A. Obod3, Oleg D. Volpian3
1
Joint Institute for High Temperatures Russian Academy of Sciences, Moscow, Russia
2
Institute for Space Researches Russian Academy of Sciences, Moscow, Russia
3
R&D Company Fotron Auto, Moscow, Russia
4
Far Eastern Federal University, Vladivostok, Russia
Contents
1. Introduction: Nonlocal Dispersion of Waves in Transparent Gradient Dielectrics 490
2. Frustrated TIR from Gradient Dielectric Layers 495
2.1 Evanescent Waves in Uniform Photonic Barriers 496
2.2 TIR of Light from a Gradient Half-Space 498
2.3 Tunneling Through a Transparent Gradient Layer with >0 501
2.4 Propagation of Light in a Gradient Dielectric with <0 507
2.5 Evanescent Modes Near the Transition Point in Gradient Media 509
3. Interplay of Artificial and Natural Dispersion in Tunneling Phenomena 511
3.1 TIR from a Transition Plasma Layer 512
3.2 Evanescent Modes Beyond the Polariton Gap in a Transparent Dielectric 516
3.3 Resonant Magnetic Modes in Dielectric Inclusions 520
3.4 Tunneling Phenomena in Negative and ENZ Metamaterials 523
3.5 Energy Transfer by Reactive Modes in the Near Zone of the Source 527
4. Tunneling of Microwaves Through Gradient TLs 529
4.1 Microwave Analogy of a Gradient Photonic Barrier 531
4.2 Evanescent Voltage and Current Waves in a Gradient TL 536
4.3 Brewster Effect for the TE10 Mode Tunneling Through a Gradient Diaphragm
in a Waveguide 538
4.4 GHz Resonance in Periodic Arrays of Plasmonic Nanowires 543
4.5 Propagation of Microwaves Through a Subwavelength Narrowing in a
Waveguide 545
5. Conclusions: The Mystery of Photon TunnelingParadoxes and Perspectives 551
5.1 Tunneling and Nonlocality 551
5.2 Tunneling of Light as a Stochastic Process 554
5.3 Tunneling of Wave Pulses as a Quasistatic Process 556
Acknowledgments 558
References 558
The goal of this review is to put in order the scientific results, obtained
sometimes on a case-by-case basis, and to stress the underlying physical
principles of tunneling in transparent heterogeneous media. The tunnel
phenomena, sharing a common stage with quantum mechanics and classical
electrodynamics, present intriguing features that stimulate the ongoing
interest.
Glancing back on this race of researches one can recall the picturesque
and inspired phrase of Prof. T.W. Hansch from his Nobel Lecture
(2005): The most surprising result of these researches would be, if we find
nothing surprising.
tg s, p th x
tg r s, p , tg t s, p : (7)
th x tg s, p
s,p (6); in this limit the phases of transmitted waves t become inde-
pendent upon the thickness of barrier too.
Phenomena (ii) and (iii) characterize FTIR in uniform nondispersive layers
with sharp boundaries, where the tunneling regime can arise only for
oblique incidence above the critical angle. The drastic changes of this pic-
ture, stipulated by the gradual distribution of dielectric permittivity, are
examined below.
1 W 2 z
z n21 U 2 z, U 2 z 1 + , g > 1: (9)
g g
W1 z 1 + , W2 z exp : (10)
L L
The dimensionless quantity g and spatial scale L are the free parameters of
distributions (9) and (10). Supposing for simplicity, the refractive indices of
both media to be equal at the boundary z 0 n1 jz0 n0 and taking into
account the vanishing of spatial dependence of refractive index n1 in the
depth of medium z L , we find from (9)
s
1
n1 jzL n0 1 : (11)
g
f expik0 ysin
z p , 1 + z=L; (12)
W1 z
1
cos 2 c : (15)
g
here the values of index s can be both real and imaginary in the frequency
intervals > s and < s ,respectively, where
p
c g
s : (17)
2Ln0
This heterogeneity-induced dispersion of gradient layer results in the
difference in spatial structure of tunneling field : the analysis of func-
tions Ks (Dunster, 1990) shows that the spatial weakening of is monotonic
in the low frequency interval ( < s ), meanwhile in the high frequency
interval ( > s ) the weakening of is accompanied by its spatial
oscillations.
Now let us turn out to the consideration of the same model (9) with
another function W z W2 z (10). Proceeding in a similar fashion, we
500 Alexander B. Shvartsburg et al.
can introduce, by analogy with (12), the generating function and the new
variable :
z
The field described by this function f is distinguished from the field (16):
unlike the profile W1(z), the solution of master Equation (13) for W2(z) in a
case g > 0 is given by the Hankel functions of real argument
Hs1 p expz=L , the argument being independent upon the angle of
incidence ; however, the index of this function s can be either real or imag-
inary, subject to this angle. The TIR regime arises for the angles of incidence
> c , the value of critical angle c, determined in Equation (15), is valid for
both models W1(z) and W2(z). Comparing the distributions U2(z) (9), con-
nected with damping functions W1(z) and W2(z), one can see that the initial
values of n2(z) at the interface z 0, as well as their gradients at the interface
z 0, are equal; moreover, the values of n2(z) in the depth of media are
equal too (11). However, the curve W2(z) is decreasing more rapidly
than W1(z); owing to this steepness, the model U2(z) (9) with W2(z)
is approaching to the step-like variation of refractive index at the interface
z 0 and, similar to the step-like variation, the field in this model is
decaying monotonically. To the contrary, the model U2(z) with W1(z)
is sloping near by the interface z 0 more gently, and the spatial structure
of tunneling field ( > c ) in this case can be both monotonic and oscillating
subject to the correlation between the frequencies and s (17); thus the
models W1(z) and W2(z) relate to the dispersive and dispersiveless regimes of
TIR in the gradient dielectric layers, respectively. Moreover, the structure
of the fields for the waves incident under the critical angle c , deter-
mined in the models W1(z) and W2(z) by the Equation (13) under the con-
ditions p 0 and s 0, respectively, proves to be different. These examples
illustrate the dependence of evanescent fields, penetrating inside the gradient
media under the TIR conditions, upon the fine structure of refractive index
distribution n(z) in the gradient layer.
Tunneling of Electromagnetic Waves 501
1 L2 d d
Um 2
, y , L1 2 , L2 : (21)
1+y 2L1 4y 2y
Equation (2) governing the wave field inside the layer is reduced in
this geometry (@y 0) to an ordinary differential equation. Solutions of
this simplified equation with distribution U(z) (20) describe waves with
both real and imaginary wave numbers, corresponding to propagating
and tunneling regimes of propagation, respectively. The spectral ranges
related to these regimes, separated by some characteristic frequency
, dependent upon the shape and size of profile U(z) (Shvartsburg &
Petite, 2005):
p
2y 1 + y2
: (22)
no d
502 Alexander B. Shvartsburg et al.
1.0
0.9
0.8
U2
0.7 1
0.6
2
0.5
0.0 0.2 0.4 0.6 0.8 1.0
z/d
Figure 1 Profiles of dielectric permittivity U2(z) versus the normalized width of the
gradient barrier; curves 1 and 2 relate, respectively, to profile (20) with the value of
parameter y 0.577 and profile (109) with the values of parameters M 3.96, g 2.1.
Note the difference between the characteristic frequencies (22) and (15),
connected with the difference in distributions (20) and (9). Thus, the profile
of refractive index (20) provides the waveguide-like dispersion of a dielectric
layer without free carriers. The frequencies > ( < ) fall in the prop-
agating (tunneling) spectral ranges. Introducing the variable
z
U z1 dz1 , (23)
0
one can present the propagating wave field inside the layer as a result of
interference of forward and backward waves:
Aexpiq + Q expiq ne
p p , q ,
U z c (24)
p
ne n0 1 u2 , u :
Formula (24) describes the high frequency wave ( > , u < 1); in an
opposite case ( < , u > 1) the tunneling field inside the gradient layer
is presented by low frequency evanescent and anti-evanescent waves:
Tunneling of Electromagnetic Waves 503
Aexpp + Q expp ne p
t p , p , ne n0 u2 1: (25)
U z c
Pz Re Ex Hy* : (27)
4
Note that this energy flux is formed by evanescent and anti-
evanescent modes.
Let us consider the tunneling regime ( < , u > 1) of wave travers-
ing the nanostructure, containing m 1 similar adjacent nanofilms with
thickness d, deposited on a homogeneous half-space with refractive index n.
Attributing the number m 1 to the first layer at the far side of this
structure, one can find the explicit formulae for its complex reflection Rm
and transmission Tm coefficients using the standard continuity conditions
for the components (26) on the boundaries of each layer (Shvartsburg &
Maradudin, 2013):
504 Alexander B. Shvartsburg et al.
i + n e m
Rm 2 ; (28)
i + n e m
2
p !
p
1 + y2 + y
ne n0 u 1 , t th l 1 u2 , l ln p
2 2 2
,
1 + y2 y
ph p
im (29)
2un0 y 2i n ch l 1 u 2
Ym
p , Tm 1 tm :
1 + y2 i + ne m m1
2
The parameters m in (28) and (29) are linked by the chain of recursive
relations, obtained from the continuity conditions on the interfaces between
the m-th and m 1-th layers (m > 1):
ne m1 + t
m : (30)
ne 1 + tm1 t
The first term in this chain 1, related to the far layer in the structure, is:
n i ne t
1 2t
: (31)
nt i ne
2
The expressions for Rm and Tm coefficients for the case of propagating
waves ( > , u < 1) follows directly from (28) to (29) due to the
replacements
p p p
u2 1 ! i 1 u2 , t ! i tg l u2 1 ,
p
p
(32)
ch l 1 u2 ! cos l u2 1 :
(a)
2.2
2
1.8
1.6
1.4
1.2
n
1
0.8
0.6
0.4
0.2
0
0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420
z (nm)
(b)
0.5
0.45
0.4
0.35 1
0.3
|R|2 0.25
2
0.2
0.15
0.1
0.05
0
400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000
l (mm)
Figure 2 Reflectance spectra jRj2 of the periodical nanostructure, containing three gra-
dient films (20). (a) Geometry of nanostructure (thickness d 140 nm, the values of
refractive indices n0 2 and nmin 1.5). (b) Curves 1 and 2 present, respectively, the
spectra jRj2 calculated for the three-layer gradient structure, shown on (a), and for
the homogeneous layer with thickness D 3d, refractive index n0 2 and n 1.52.
1
0.95
2
0.9
0.85
0.8 1
0.75
0.7
0.65
0.6
0.55
|T|2
0.5
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100
l (nm)
Figure 3 Experimental (1) and theoretical (2) transmittance spectra for periodical nano-
structure, containing 11 layers (20) with width d 140 nm; the values of other param-
eters of this structure are: y 0.577, n 1.52, and n0 1.5.
506 Alexander B. Shvartsburg et al.
gradient dielectric structures without free carriers (Re > 0). The
wave energy is transmitted through these structures by means of
evanescent modes.
ii. Interference of evanescent and anti-evanescent modes in a stack of
gradient layers results in the formation of new reflection/transmission
spectra having nothing in common neither with the almost total
reflection nor with the exponentially small transmission of a plasma
layer in the FTIR regime ( < p ). Thus, considering the three-layer
gradient nanostructure, depicted on the Figure 2a, characterized by
the total thickness 3d and cutoff frequency (22), corresponding
to the wavelength c 2c= 1:32 m, one can compare its reflec-
tance with the reflectance of a homogeneous layer with the same
thickness 3d, the other parameters n0 and n being equal too
(Figure 2b). Unlike the calculated oscillating reflectance of the
homogeneous layer (curve 2), the experimentally measured reflec-
tance of the three-layer gradient nanostructure is characterized by
the narrow maximum nearby the blue end of the visible spectrum,
stipulated by HID, accompanied by horizontal plateau, presenting
almost constant small reflectance (jRj2 0:05) in a broad spectral range,
covering both short ( < c ) and long ( > c ) wavelengths. This
plateau corresponds to the broadband range of high transmittance
jT j2 1 jRj2 0:95, including, in particular, the tunneling waves
( > c ).
iii. Increasing the number of gradient nanofilms does not change essentially
the transmittance spectra. Thus, the transmittance spectra of a multi-
layer, containing 11 gradient nanofilms with profile of refractive index
(20), are depicted on Figure 3; curves 1 and 2 present, respectively, the
experimental results and theoretical values, calculated by means of for-
mulae (29)(31). The theoretical and experimental results prove to be
in a good agreement. The narrow minimum at the wavelengths near
450 nm, characterized a decrease of the transmission coefficient
by 1819
, arises as well, as at the Figure 2b, due to HID. The growth
of wavelengths in the interval 600 nm < < 2100 nm, brings the small
oscillations of transparency around the value jT j2 0:9; respectively,
the reflection coefficient keeps a weakly varying value jRj2 0:1 in this
visible and near IR range, resembling the plateau at the same range are
in Figure 2a.
Tunneling of Electromagnetic Waves 507
d2 z L
2
+ 2 n2 0, , , (33)
d L c
Till now the dependence q() remains arbitrary. Let us choose this
dependence, e.g., in a form:
q0 2 2
q 2 , a + b exp : (36)
508 Alexander B. Shvartsburg et al.
Figure 4 Square of refractive index in gradient layer n2 (35)(36) is plotted versus the
normalized coordinate ; the values of parameters in Equations (35)(37) are: a 1,
b 5, 2 4, q0 2, and 2 2.
Here, the real quantities a, b, and are the dimensionless free parameters
of spatial dependence of refractive index n(). The spatial distribution of
refractive index (35) inside the interval 0 can be written by substitution
of (36) to (35) in an explicit form, depicted in Figure 4b.
1 q20 2 2 1 2 2 2 2 2
n 2 2 +
2
, b exp : (37)
2
a + b2 q0
R :
a + b2 + q0
Tunneling of Electromagnetic Waves 509
n2 n20 + 2 z: (39)
Here, is some dimensionless constant, characterizing the maximum
perturbation of the background value of refractive index n0, (z) is an
unknown function, describing the spatial distribution of this perturbation.
Considering wave propagation in the y-direction, orthogonal to the
510 Alexander B. Shvartsburg et al.
z-direction, substituting the distribution (39) into the wave equation (2) and
seeking the generating function in a form
n0
f vexpik0 y, k0 , v k0 z, (40)
c
we find the equation governing the function f:
@ 2 f 2 v
+ f 0: (41)
@v2 n20
Unlike the approach, used in the previous sections, based on the search of
solutions determined by the given distribution of , we will seek the distribution
(z), related to the prescribed solution f(v). Thus, for the bell-like solutions
m
f 1 + a2 v 2 , m > 0, (42)
we obtain
2ma2 n20 1 1 + 2ma2 v2
v : (43)
2 1 + a2 v2 2
In the simple case m 1, the solution (42) has a shape of the well-known
Cauchy curve. The profile (z) resembles the symmetrical Mexican hat
with a maximum at v 0 and zero-crossing points v v0 and minima
at v vm
2
an0 ma2 n20 1 + 2m2
0 2m , v0 0, vm : (44)
42 1 + m
Thus, for each wavelength and field structure (42) one can design the
distribution (z), providing the prescribed structure f(v). Herein, the pertur-
bation of refractive index is not small and can exceed at some points the
background value n0. Far from the maximum at v 0 the perturbation
(z) is vanishing: 1 0. The bound mode fields have been found
to be more weakly localized than those of guided modes supported by
the normal waveguide because the evanescent tails of (42) obey, instead
of an exponential law, the power law:
f jvj jaj1 ! jzj2m : (45)
Substitution of the function (40) as the Ax component of the vector
potential to (1) yields the values of the transverse components Ex, Hz and the
longitudinal component of the modal field Hy. The former component is
Tunneling of Electromagnetic Waves 511
decaying at the periphery of the localized wave faster than Ex and Hz:
Hy jvj jaj1 ! jzj2m1 . This delayed damping of evanescent tails may
be useful for deep penetration of fields, e.g., for the excitation of electronic
states in the subsurface layers of heterogeneous dielectrics (Zayatz,
Smolyaninov, & Maradudin, 2005).
2p 4e2 Ne
1 , 2p : (46)
2 me
2
2L
1 2 : (47)
2T
512 Alexander B. Shvartsburg et al.
1 W 2 z
N z N0 1 + , g > 0, W z exp ,
g g L (48)
z=L,
we can examine the spatial structure of evanescent modes, formed, e.g., by
the low frequency TE-polarized wave, incident on the interface under an
angle . Substitution of distribution (48) to (46) brings the equation for
the generating function in a form of a Bessel equation
d2 1 d s2
+ + p 2 0,
2
(49)
d2 d
herein the values of variable and parameters p2 and s2 are:
z
1 p L 2
exp , p 2
,
L g c
2 (50)
p L 2 1 L
s
2
1+ 1 sin 2 :
c g c
The solution of Equation (49) in the case s2 > 0 is given by Bessel func-
tions Js(p).
Consider first the case of normal incidence on a plasma ( 0). To visu-
alize the salient features of these solutions, it is worthwhile to recall that the
Bessel functions with odd half-integer indices are expressed via the
514 Alexander B. Shvartsburg et al.
Y sin p
ez , l cos p, (53)
l p
exp sin p exp
Y exp cos p exp :
2 p
Figure 5 Penetration of electromagnetic field into the plasma transition layer with var-
iable electron density (48). Normalized values of electric e and magnetic h components
are plotted versus the normalized distance from the plasma boundary in a case
p 2.35. Curves 1 and 2 show the spatial distributions of e and h for the normal inci-
dence and curve 3 illustrates the decrease of e for the inclined incidence (50).
The influence of the boundaries of the confined plasma volume may lead
to plasma depolarization inside the volume. Such effects are important for
three-dimensionally bound particles of a crystalline material, containing
the conduction electrons, e.g., for n-type semiconductors with free carriers.
Thus, for the spheroid particle the electron plasma frequency inside the par-
ticle p is coordinate-independent; however, owing to the depolarization
factor D, is dependent upon the spheroid axial ratio, the frequency p is dis-
tinguished upon the value p, characterized an unbounded plasma
(Dresselhaus, 1955):
2p D 4e2 N0
2p , 2p : (54)
1 + D m
Here, is the dielectric permittivity per unit volume. Thus, the plasma fre-
quency values for two samples of the same material, distinguished only by
their shapes, may be considerably shifted.
The spatial variations of refractive indices in the tailored semiconductors
with free carriers can be formed by the differences in their plasma frequen-
cies, dependent, in their turn, from the effective masses of carriers me: thus,
the difference of plasma frequency p for GaAsme 0:0665m0 and
Alx Ga1x Asme 0:15m0 results in the disparity of the related refractive
516 Alexander B. Shvartsburg et al.
indices: n 3:20 (GaAs) and n 3:50 (Alx Ga1x As) (Wilson, Glytsis, &
Gaylord, 1993). The specially designed periodical structures, fabricated from
ultrathin layers of these semiconductor species, can act as optical superlattices
(Nelin, 2007).
2L z 2L0 U 2 z: (55)
Using the distribution (55) one can represent the generalized profile (z)
(47) as
2
2 L0 W z
2 2
z n0 , n20 1 : (56)
2 2T
Consider the model (Shvartsburg & Maradudin, 2013)
h z
z
i1
W z cos + M sin : (57)
L L
This profile is characterized by two free parameters (scale L and dimen-
sionless quantity M), connected with the thickness of gradient layer d and the
depth of modulation of the content of polarized molecules determined by
the value W 2min:
d 1
2 arctgM , Wmin
2
: (58)
L 1 + M2
Tunneling of Electromagnetic Waves 517
To solve the master equation (2) with the dielectric permittivity, given
by functions (9) and (57), it is worthwhile to introduce the new function f(z)
and variables and
z
f z
z p , W z1 dz1 , (59)
W z 0
2 z
3
6
1 + m tg p
7
1 +
p ln 4 2L
z 5, m 1 + M 2 M, (60)
L 1 + M2 1 m tg
2L
p
1 + M2
0 , 0 ln m + , jz0 0 , jzd 0 : (61)
L
The function W(z) can be expressed via the dimensionless coordinate
in an explicit form:
ch
W p , W 0 W 0 1: (62)
1 + M2
Bringing together the results (59)(62), we obtain from Equation (2)
equation, governing the function f(), in a form, familiar from quantum
mechanics (Schiff, 1968):
d2 f
q + 2 f 0,
2
(63)
d2 ch
" #
1 2 2L0 c
q 1+
2
, 0 , (64)
4 0
1 + M T
2 2 2 2n0 L
" 2 #
1 2
1 : (65)
4 0 2 2T
Substitution of values 1,2 to the expression for q2 (64) gives the value
q2 > 0; thus, the generating function is expressed via the evanescent
and anti-evanescent waves
Aexpq + Q expq 2 1 1 L0
2
p , q 1+ : (67)
W z 4 1 + M2
Note, that when > T these evanescent modes arise in the spectral
range > T , located beyond of the band stop of the unperturbed material
T < < L0 .
To illustrate the spectral properties of a spatially varying stop band, one
can consider the light wave, which is impinging normally on the gradient
subsurface layer in dielectric ZnSe, possessing a polariton bandgap between
L0 and T. Taking into account, that this material in a natural state is
characterized by the refractive index n0 1.98 and a polariton gap in the
near IR range between the frequencies T 0:47
1015 rad=s,
L0 0:88
1015 rad=s (Deych et al., 1998). Using the model (57), one
can find the variation of n; thus, in a case of modulation depth Umin 2
0.1
(58) and thickness of gradient layer d 150 nm the values of parameter M
and characteristic scale L, calculated from (58), are: M 3, L 60 nm;
herein the characteristic frequency, occurred due to heterogeneity (64),
0 1:26
1015 rad=s The frequency of evanescent mode , determined
from Equation (66) is 1.151015 rad/s; parameter q for this mode (67)
is q 0.51; thus this mode arises in the perturbed layer in the transparency
range of unperturbed dielectric > L0 0:88
1015 rad=s. Note, that
the refractive index of the perturbed dielectric mode for this frequency is real:
substitution of this value to (56) yields the value n 2.1. Side-by-side with
the analysis, performed in the Section 2.3, this example shows the drastic
changes of optical properties of material, providing the formation of evanes-
cent modes, existing surprisingly in the spectral range with the positive
values of n2. The similar effect can arise in another dielectrics too, e.g., in
LiF, characterized by a polariton gap between T 0:365
1015 rad=s and
L0 1:08
1015 rad=s (Serebryannikov, Ozbay, & Nojima, 2014).
Progress in crystal growth techniques has made feasible the realization of
multilayer periodic dielectric transparent structures, composed from
ultrathin layers with alternating high- and low-refractive indices; the periods
of these artificial one-dimensional structures, so-called superlattices, signif-
icantly exceed the natural periods of crystal lattices, constituting these
structures (Yeh, 1985). The prescribed distribution of refractive index in
these structures formed the basis for the design of systems for spectral
Tunneling of Electromagnetic Waves 519
(a)
1.0
2
0.9
|T |2 1
0.8
0.7
1.05 2.00 3.00
u
(b) 1
0.98
0.96
0.94
0.92
|Tm|2
0.9
0.88
0.86
m = 10
0.84
0.82
m = 20
0.8
1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3
u
Figure 6 Transmittance spectra of periodical gradient structures (20) providing the
wave energy transport in the tunneling regimes. The values of transmission coefficient
jTj2 are plotted versus the normalized frequency u (24). Structures containing m
nanolayers with thickness d 100 nm are supposed to be deposited on the thick sub-
strate with refractive index n 2.3. (a) n0 2.2187, y 0.75, spectra 1 and 2 relate to
nanostructures, containing 1 and 20 nanolayers, respectively. (b) n0 1.8928, y 0.577.
where
1 dJm n dJm 2a
am Jm nJm n , , m 0, 1, 2 . ..
d d
(69)
dJm n dH 1 Jm am Hm1
Hm1 nJm n m , cm :
d d Jm n
Here, n and a are the refractive index and radius of the cylinder, is
the wavelength of the incident wave, Jm and H(1)
m are the Bessel and Hankel
functions of the first kind, respectively. These forms satisfy continuity of
the tangential components of the total electric and magnetic fields at the
cylindrical surface.
The origin of magnetic activity in this configuration is connected
!
with the displacement currents j d (OBrien & Pendry, 2002). These
currents, induced inside the cylinder in the planes perpendicular to the
!
cylinder axis due to temporal variations of electric displacement D
( Jackson, 1998)
!
! 1 @D
jd ; (70)
4 @t
can be greatly enhanced, when the size parameter (69) is resonant with an
eigenmode of the field oscillations inside the cylinder. A characteristic equa-
tion is obtained by setting the denominator of the coefficient am (69) equal
to zero. In this case (m-th order resonance), the relevant cm coefficient
522 Alexander B. Shvartsburg et al.
dominates over the others and, the field inside the cylinder is described
mainly by the basis function Jm(nkr) (68). For the zeroth-order resonance,
only the azimuthally directed displacement current is enhanced and
results in the enhanced magnetic field along the cylinder axis. This effect
can provide the magnetic activity in the nonmagnetic dielectric inclusion;
herein the magnetic field is not confined inside the cylinder, but forms
some halo around it. In an array of identical cylinders, these localized mag-
netic modes could stimulate the macroscopic bulk magnetization of the
system, characterized by some frequency-dependent effective magnetic
permeability eff. This permeability can be presented in a dispersive form,
similar to the one, which was found at first for the arrays of split-ring metallic
resonators, embedded to the dielectric matrix (Pendry, Holden, Robbins, &
Stewart, 1999)
Y 2
eff 1 : (71)
2 20 + i
dHm1 dJm
2p Jm p1 Hm1 0: (72)
d d
Here, Jm and H(1) m are the spherical Bessel and Hankel functions, the
superscript p refers to either TM (p 1) or TE (p 2) wave polarization;
the size parameter and the relative refractive index are defined as
2anmed np
, ; (73)
nmed
np and nmed are the refractive indices of the particle and surrounding
medium, respectively, a is the particle radius. When these spherical Si inclu-
sions are arranged in square-lattice layers, the coupling between Mie modes
in the neighboring layers results in splitting and shifting of Mie resonances.
The refractive index nmed for Ar in the visible and near IR spectral ranges is
imaginary, so, the coupling among the nanospheres and between the
spheres and the film surfaces is mediated by evanescent modes. When
the imaginary part of dielectric function of Ar is small ImAr 1,
the transmittance of an Ar nanofilm with embedded Si spheres for some
wavelengths in the near IR range tends to 100% (Borisov, de Abajo, &
Shabanov, 2005).
1 @ 1 @
Ex , Hy : (74)
c @t @z
Owing to the representation (74), the set of Maxwell equations for a
monochromatic wave is reduced to the equation
d 2 2 1 d d
+ 2 ; (75)
dz 2 c dz dz
Equation (75) is valid for arbitrary combinations of and , including, in
particular, so-called right-handed ( > 0, > 0), left-handed ( < 0,
< 0), and single-negative ( > 0, < 0 or < 0, > 0) materials.
The propagation of EM waves through sharp boundaries between
homogeneous right-handed (RHM) and left-handed (LHM) materials
was examined by Ziolkowski and Heyman (2001). To visualize some pecu-
liarities of the gradual transition layer between these media, it is expedient
to present the spectral and spatial dependences of and as
, z U z and , z U z; here U(z) is an arbitrary real
odd function, the dependence () in the metamaterials is usually pres-
ented in a Lorentz-like resonant form (Pendry et al., 1999). Introducing
the new variable (23), we can eliminate the function U(z) from the
Equation (75):
d2 2 n20
+ 2 0, n20 : (76)
d2 c
i. Near the center of the transition layer jzj < L the distances between
zero-crossing points of Re are unequal, whereas deep in the media
jzj L these distances are equal, and expression (77) describes the
harmonic wave expiqz;
p
ii. The refractive index n0 and impedance Z = retain their
values for any given frequency in both half-spaces z 0 and z 0;
iii. The constancy of n0 and Z indicates that the wave travels from z 0
to z 0 without reflection from the plane z 0.
Note, that the generating function can be introduced, side-by-side with
(74), by an another representation
1 @ 1 @
Ex , Hy : (78)
@z c @t
attention. Within the ENZ material, the wave possesses small spatial var-
iations, an extremely large phase velocity, and its wavelength tends to
infinity; in other words, this wave exhibits a static-like character;
the source of radiation, embedded to the ENZ medium, can generate
an extraordinarily directed wave beam. These materials have been
realized in the microwave (Liu et al., 2008), mid-infrared (Mocella,
Dardano, Rendina, & Cabrini, 2010), and visible (Vesseur, Coenen,
Caglayan, Engheta, & Polman, 2013) spectral ranges. However, the
ENZ materials having a permittivity near zero and a permeability
p
about unity, possess a near-zero refractive index n and high opti-
p
are characterized by the phase shift /2 between electric and magnetic compo-
nents; these reactive components do not contribute to the radiation flux I (80).
However, the radiation of two dipoles, oscillating with the same fre-
quency , can create the interference of reactive components of dipole
fields, providing the nonradiating energy transfer between dipoles. Consider
two dipoles, p1 p10 expit and p2 p20 expit, spaced by distance l
at the x-axis and oscillating along the z-axis and let us suppose the dipoles to
be located symmetrically with respect to the plane x 0 at the points with
coordinates l=2,0,0 and l=2, 0,0. The total energy flow, irradiated
by dipoles through the plane x 0, contains three terms:
Ix I1x + I2x + Iix ; the first and second terms
4 jp10 j2 4 jp20 j2
I1x , I2x ; (82)
6c 3 6c 3
define the energy fluxes, irradiated by each dipole independently, mean-
while the term Iix corresponds to the interference flux, determined by the
EM fields of both sources. If the distance l between dipoles is much less than
the radiation wavelength 2 c=, the main contribution to the interfer-
ence flux is provided by the reactive components of the near zone:
ii. the the flux Iix is localized mainly in the area between dipoles:
l=2 < x < l=2;
! !
iii. The spatial distributions of the E H components of electric dipole,
! !
oscillating along the z-axis, and H E components of magnetic
dipole, located in a plane, orthogonal to z-axis, are known to coincide
(Kong, 1986); thus the nonradiating energy flux, stipulated by the inter-
ference of reactive field components, can arise in a near zone between
the magnetic dipoles too.
It is remarkable that the concept of formation of a wave pattern, providing
the energy transfer by the nonpropagating near-field components associated
with the source, seems to be valid for the near zones of sources of different
physical nature. Thus, in the optical range, the lifetime of a molecule excited
state (molecular fluorescence) was shown to depend upon the nonradia-
ting decay due to energy dissipation into the surrounding environment
(Liaw, Chen, Chen, & Kuo, 2009). At the other range of electromagnetic
radiation spectrum, at the GHz range, the physically similar situation was
presented in the experiments of Ranfagni, Fabeni, Pazzi, and Mugnai
(1993): in these experiments with closely located transmitter and receiver
horns the mouth of the microwave horn launcher was considered as a source
of radiation, and the field between horns was viewed as a reactive field. The
analysis of the pulse delay on the way between two tilted horns, analogous to
the tunneling of microwaves in the sub-cutoff waveguide, was treated in
favor of the concept of energy transfer by reactive wave fields.
parameters (for short, gradient TL) open one more new avenue for the con-
trol of microwave radiation flows. A theoretical analysis of these directional
systems illustrates the applicability of the concept of nonlocal HID to micro-
wave circuits. The propagation of EM radiation in these circuits can be
governed by the simultaneous action of geometrical dispersive effects, stip-
ulated by the shapes and sizes of perfectly conducting waveguides, and non-
local effects, arising due to heterogeneous distributions of capacity and/or
inductance in the circuit. Attention will be given below to the influence
of these phenomena on the tunneling of electromagnetic waves in
gradient TLs.
The series of electromagnetic tunneling phenomena have both optical
and microwave counterparts. These analogies promote the elaboration of
some general physical insight on these phenomena despite the drastic differ-
ence in their temporal and spatial scales. TLs can be used for the modeling of
effects of gradient nanophotonics, especially as the making of microwave
circuitry proves to be easier, than the fabrication of gradient nanofilms
(Nelin, 2007).
To emphasize the dispersive effects, let us consider a lossless TL. The
spatial-temporal variations of current I and voltage V are known to be
governed by the pair of equations, describing the propagation of wave in
z-direction (Kong, 1986):
@I @V @V @I
+ C z 0, + M z 0: (85)
@z @t @z @t
Here, C(z) and M(z) are the coordinate-dependent capacitance and
inductance per unit length, characterized by dimensionless functions
W 2(z) and F 2(z), respectively:
C z C0 W 2 z, M z M0 F 2 z: (86)
The heterogeneous part of a TL, occupying the region 0 z d, is
assumed to be located between its homogeneous regions z 0 and z d,
where the quantities C and M possess constant values.
The equations of a heterogeneous TL with continuously distributed
parameters (85) are used for analysis of HID of a coaxial waveguide, stipu-
lated by all-dielectric 3D gradient porous diaphragm, installed inside this
waveguide (Section 4.1); this structure which is characterized by some dis-
tribution of capacitance C(z), mimics the gradient nanofilm with the similar
profile of effective dielectric permittivity eff (20). Calculation of GHz
Tunneling of Electromagnetic Waves 531
transmittance spectra for TEM mode for this structure, based on the exact
analytical solutions, describing the propagation of IR radiation through gra-
dient nanofilms, shows a good agreement with experimental measurement
of these spectra. The similar effect of HID is shown (Section 4.2) to provide
the formation of cutoff frequency and tunneling of waves in TL, character-
ized by different spatial distributions of capacitance and inductance. Tunnel-
ing of the fundamental TE10 mode through the gradient diaphragm with the
flexible distribution of eff, installed to the metallic waveguide, is considered
in Section 4.3; the reflectionless propagation of this evanescent mode,
analogous to the Brewster effect, is illustrated too.
The decrease of natural plasma frequencies of metals down to the GHz
range is considered in Section 4.4 as a perspective method for the creation
of a metamaterial with strong dispersion in the microwave spectral range.
The effect of nonradiating energy transfer by reactive waves in the near
zone of the source is illustrated in Section 4.4. The effective transmittance
of waves through a smoothly shaped subwavelength narrowing in the
waveguide is discussed in the Section 4.5. This analysis is continued by
the physically similar problem of tunneling of radiation through the
ultrathin slit, filled by a dielectric with constant permittivity, determined
by the geometrical parameters of narrowing. The last effect seems to be
rather intriguing, and the design of this configuration shows the richness
of possibilities offered by the transformation optics for control of the paths
of EM waves.
r
1 @ 1 @ 1 M0
V , I , v0 p , Z0 ; (87)
v0 @t Z0 @z C0 M0 C0
we will find that the second equation in (85) is reduced to identity, while the
function is governed by the first equation, which reads now as
532 Alexander B. Shvartsburg et al.
@ 2 U 2 z @ 2
+ 0; (88)
@z2 v02 @t2
Here, r0 and r1 are the external and internal radii of the waveguide,
0 and 0 are the dielectric susceptibility and magnetic permittivity of
vacuum. Substitution of (89) to (87) yields the well-known result:
p1
v0 0 0 c.
To verify the formation of HID in the microwave range, let us consider
the interferometer-like system, built from two similar coaxial waveguides
(Shvartsburg et al., 2008). We will examine the propagation of fundamental
TEM modes, launched to this coaxial waveguide from the generator. This
regime is chosen due to the following salient features, approaching to the
properties of EM waves in free space:
Tunneling of Electromagnetic Waves 533
i. The velocity of the TEM mode is known to coincide with the free-
space light velocity c.
ii. The TEM mode does not possess frequency dispersion.
iii. The polarization structure of the TEM mode contains only the trans-
versal electric and magnetic field components; unlike any other wave-
guide modes it is polarization does not contain any longitudinal
components.
One of these waveguides contains a plastic porous diaphragm, fabricated
from the dielectric with H with the air-filled pores (Figure 7a). The
ratio of plastic-filled and air-filled parts in each cross-section, varying along
the diaphragm, mimics the gradient profile of refractive index (20); unlike
534 Alexander B. Shvartsburg et al.
the sputtered nanofilm the thickness of this diaphragm is about several cm, as
is illustrated by the ruler on Figure 7b. The effective value of the dielectric
permittivity for this porous structure eff can be found by means of the
Maxwell-Garnett formula (Maxwell-Garnett, 1904):
H PD + 1 P H + G1 P D H
eff , (90)
PD + 1 P H GP D H
here G is the volume part of medium, filled by the admixture with the value
of dielectric permittivity D ; in the problem under discussion H 2:3
and D 1. The polarization factor P, dependent upon the shape of admix-
ture particles, is known to possess the values P 0:5 and P 0:33 for the
lengthy cylinder and sphere, respectively.
To compare the microwave transmittance of a heterogeneous diaphragm
with the transmittance of a plane photonic barrier (20), the spatial distribu-
tion of eff (z) in the diaphragm has to follow the profile U2(z) (20):
1 U 2 zP + 1 P H
Gz , (92)
H 1PU 2 z + 1 P
Our goal is to define the spectral range k < km , where the wave number
km is the lowest of the critical wave numbers kc, related to TE and TM
modes in the coaxial waveguide; herein the waveguide will support the
propagation of only fundamental TEM mode. The critical value kc for
TE modes is given by the first root of the dispersive equation
N0 kc r1 N0 kc r0
, (93)
J0 kc r1 J0 kc r0
J and N are the Bessel and Neuman functions, r0 and r1 are the internal and
external coaxial radii. The dispersive equation for the TM mode reads as:
N1 kc r1 N1 kc r0
: (94)
J1 kc r 1 J1 kc r0
Fixing the values r0 and r1, one can define the upper band of the spectral
range kc, related to the single-mode propagation of the TEM wave.
Tunneling of Electromagnetic Waves 535
0.98
1
0.96
0.94
0.92
2
|T|2
0.9
0.88
0.86
0.84
0.82
0.8
1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3
f (GHz)
Figure 8 Transmittance spectra of coaxial waveguide, shown on Figure 7a, with porous
gradient diaphragm, mimicking the photonic barrier (20); the values of transmission
coefficient jTj2 are plotted versus the frequency f. Curves 1 and 2 relate to the exper-
imental and theoretical spectra.
Since the incident and transmitted TEM modes are propagating along
the coaxial with free-space light velocity v0 c , we can calculate the trans-
mittance spectra using equation (28), obtained for the photonic barrier, put-
ting there the refractive index of the substrate n 1. The transmittance
spectra of the coaxial waveguide r1 1 cm, r0 2:3 cm with this gradient
diaphragm, calculated for the layer with effective permittivity (90) by means
of standard continuity conditions for voltage and current and measured in
experiment, are shown on Figure 8; the smallest values of kc for TE and
TM modes, defined from Equations (93) and (94), are 2.63 and
4.88 cm1, respectively. Thus, choosing the value km 0:628 cm1 ,
related to the frequency m km =c 1:88
1010 rad=s ( fm 3 GHz),
one can consider the frequencies of spectral range f < fm , shown on
Figure 8; herein the frequencies in the observed range 13 GHz exceed
the critical frequency of barrier fc =2, where the value is calculated
by means of (22); for the parameters, related to Figure 7, fc 0.27 GHz.
Comparison of calculated and measured transmittance spectra on this figure
indicates a good agreement between the theoretical and experimental data,
visualizing the effect of HID of this microwave TL.
536 Alexander B. Shvartsburg et al.
1 @ @
I , U Z0 : (95)
v0 @t @z
In this case, the first equation in (85) is reduced to an identity, while the
second one brings the equation, governing the function in a form, coin-
ciding with (88). However, to examine the combined influence of wave
propagation, produced by spatial distributions of both capacitance
C z C0 W 2 z and inductance M z M0 F 2 z of TL, special consid-
eration is needed.
To solve this problem, let us differentiate the first and second equations
in (85) with respect to the variables z and t, respectively, excluding the func-
tion V we will obtain one equation governing the unknown distribution of
current I:
@2I @2I 1 dC z @I
C zM z : (96)
@z2 @t 2 C z dz @z
Note, the similarity of Equations (96) and (75), describing the waves in a
metamaterial with coordinate-dependent parameters (z) and (z). The
replacement I ! V , C z ! M z in (96) brings the similar equation,
determining the distribution of voltage V. Due to using a new variable ,
distinguished from (23)
z
W 2 z1 dz1 , (97)
0
@ 2 I 2 F 2 z
+ I 0: (98)
@2 v2 W 2 z
z z 1
W z 1 + , F z 1 + , (99)
l1 l2
we can study the effects caused by the increase or decrease of capacity and
inductance of TL in a general form, considering both positive and negative
values of the lengths l1 and l2 independently. To distinguish these lengths,
related to models of different physical quantities, from the lengths L1 and
L2, describing the distribution of one quantity, e.g., the dielectric permittiv-
ity in model (99), lowercase letters are used in models (99).
Substitution of function W2(z) (99) into (97) yields the explicit expres-
sion of variable via z:
z 1
z 1+ : (100)
l1
Owing to (100) the product of the functions W(z) and F(z) reads as a
function of :
1
F zW z U 1 + , (101)
l
l1 l2
l : (102)
l1 l2
By substituting (101) into (98), we can rewrite this equation in -space in
a form similar to (88):
d2 u 2 2
+ U u 0: (103)
d2 v02
This equation is easy to solve by means of the algorithm, used above in
the Section 2.1. Introducing the new variable
l1 z + l2
U 1 d1 l ln , (104)
0 l2 z + l1
we can present the solution of Equation (103) in the form of forward and
backward waves, traveling along the -axis
Ar expiq + Q expiq
u p : (105)
U
The wave number q in (105) corresponds to the plasma-like dispersion of
the gradient layer
538 Alexander B. Shvartsburg et al.
s
2 v0
q 1 2, : (106)
v0 2jlj
1 W 2 z h z
z
i1
U 2 z 1 + , W z cos + M sin , (109)
g g L L
here the real positive quantities g, M, and L are the free parameters of dis-
tribution (109), dependent upon the diaphragm width d, minimum of
refractive index nmin, and half width of nmin. Considering the symmetrical
profile (W 0 W d 1), one can link these quantities with both physical
and geometrical parameters of the profile
d M2
2 arctgM, nmin n0 1
2 2
: (110)
L g1 + M 2
f z y
A p sin : (111)
W z b
Moreover, we will consider here the case q2 > 0, when the solution
of Equation (116) is presented by evanescent and anti-evanescent waves
in -space. Under these conditions the generating function (111),
corresponding to the tunneling regime, can be written as
y
r
z
z
2 4L 2q 1 + M 2 Mt :
2r 2 nev
H R , (123)
cR
where R is the distance from the wire axis in the plane, orthogonal to this
!
axis. Expressing the magnetic field in terms of the vector potential A
! !
H R rot A , (124)
and substituting from (124) into (123) one can find the vector potential
(Pendry et al., 1996):
2r 2 nev a
A R ln : (125)
c r
Electrons in a magnetic field are known to have an additional contribu-
tion to their momentum of eAc 1 , and therefore their momentum per unit
length of the wire is
2 2 e2 n2 r 4 a
r 2 neAc 1 ln : (126)
c2 r
Introducing the effective mass of the electron in the current-carrying
wire meff and equating (77) to the another expression for the momentum
per unit length of the wire meff r2nv, we find meff:
544 Alexander B. Shvartsburg et al.
2e2 nr 2 a
meff ln : (127)
c2 r
This effective mass, stipulated by the wire self-inductance, can exceed
drastically the value of the electron rest mass me: thus, in case of thin
aluminum wires (r 106 m, n 1.8
1029 m3, a 5
103 m) the effec-
tive mass is meff 2:72
104 me . The effective density of electrons in the
periodic structure as a whole is
r 2
neff n : (128)
a2
Substituting both the effective mass meff and effective density neff into the
definition of the plasma frequency 2p 4e2 neff =meff , we can calculate the
frequency, characterizing this plasma-like medium:
2c 2
2p a
: (129)
a2 ln
r
Thus, the large self-inductance of a thin wire is mimicking the increase of
electron mass. The periodic structure composed of infinite wires arranged in
a cubic lattice can be considered as an effective plasma-like medium. This
man-made medium, characterized by effective dielectric permittivity eff
and plasma frequency p
2p
eff 1 , (130)
2
possesses the following properties:
i. This value p for the effective medium, determined by the geometrical
factors and independent of the electron density of metallic wires, can be
several orders of magnitude less than the plasma frequency in metals;
thus, for the aforesaid values of geometric factors a and r we have from
(129): p 5:15
1010 rad= s ( f 8.2 GHz).
ii. Until the wavelength of external radiation, incident on this microstruc-
ture, exceeds essentially the period a, this regular array of thin wires
appears as a homogeneous dielectric medium, characterized by effective
dielectric permittivity (130); thus in the aforesaid case the wavelength ,
corresponding to p, is 35 mm, which is much greater than the
lattice spacing a 5 mm.
Tunneling of Electromagnetic Waves 545
0
z q : (132)
1 + z=L 2
Calculation of the radiation flow density along the radial direction for the
NT with a biconical shape (132) yields the following expression for the
parameter (z):
2 2 z 20 L
z exp : (134)
0 20
Thus, the radiation losses are distributed along the NT. This analysis illus-
trates the weakening of transparency of smoothly shaped subwavelength
narrowings in a dielectric waveguide, stipulated by the evanescent tails of
a guided mode, dependent upon the geometric parameters of narrowing,
L and 0. This weakening is an important factor, restricting the efficiency
of wave flow transfer through the subwavelength slit in a transparent material.
Another family of heterogeneous wave barriers, supporting the resonant
tunneling of light, may be found in directional systems, containing a
narrowing, confined by curvilinear interfaces. This peculiar type of
heterogeneity-induced tunneling proves to be controlled by the geometrical
parameters of the confining interfaces. To examine this effect, let us consider
again the propagation of the TE10 mode with EM field components
Ey, Hx, Hz along a single-mode waveguide with axis z and thickness
d d=2 x d=2, filled by a dielectric with dielectric permittivity ;
to avoid any confusion with the tunneling problem, considered above in
the Section 4.3, a slightly changed notation will be used below. The
waveguide contains a symmetric smoothly shaped narrowing in the area
b=2 z b=2, and the planes x d=2 and x d=2 are assumed to
Tunneling of Electromagnetic Waves 547
(a) Hx
Hz
Ey
X
b0 x = d/2
a<0 a>0
z
O
b 0
x = d/2
b
(b)
1 2 3
1
0.8 4
|T|2
0.6
0.4
0.2
0
700 750 800 850 900 950 1000 1050 1100 1150 1200
l
Figure 9 Tunneling of fundamental mode through the smooth narrowing in the
waveguide. (a) Geometry of the narrowed waveguide. The following geometrical
parameters are indicated: the narrowing length b, the distance between the wave-
guide walls d, and the distance between the tops of the coordinate lines 0 and
0 determines the minimal width s. The polarization structure of the guided wave
is shown on the inset. (b) Dependence of the transmittance upon the slit width s in
the range 225 nm < s < 300 nm, while b 1400 nm, d 500 nm are kept fixed; curves
1, 2, 3, and 4 relate to the values of width s 225, 250, 275, and 300 nm, respectively.
be perfectly conducting; the least width of narrowing is less than the wave-
length in the waveguide (Figure 9a). To avoid massive algebra, the analysis
below is restricted to the consideration of a single mode (Shvartsburg &
Kuzmiak, 2010).
548 Alexander B. Shvartsburg et al.
The wave equation governing the generating function inside the slit is
derived from the Maxwell equations. Using the (, ) coordinates and sep-
arating the variables F f , one can find the equations, determining
the unknown functions F() and f():
d2 F 2 2 a2
+ p 2
ch A F 0, p 2
, (137)
d2 c2
d2 f
2 + p A sin f 0:
2 2
(138)
d
Here, A is some dimensionless constant, which is determined from
the boundary conditions. Rigorously speaking, the regular solutions of
Eqs. (137) and (138) can be written in terms of Mathieu functions
(Abramowitz & Stegun, 1968). However, for our goals, it is convenient to
present the solutions of these equations directly by power series. Calculation
of the reflection/transmission coefficients R and T is based on the match-
ing of field components beyond of the narrowing z b=2, z b=2
with the field components inside the narrowing (b=2 z b=2). The
obtained expression for jT j2 illustrates the influence of each of the geo-
metric parameters of the slit (distance d, narrowing length b, and its width s)
on the tunneling of waves through the slit. Thus, the transmittance
Tunneling of Electromagnetic Waves 549
2 m
k2z + k2y 2
, ky , m 1,2,3 . .. (142)
c h
the values and are given in (141). When becomes large enough, the
value ky will change from imaginary to real, providing the tunneling of
the wave through the channel.
Moreover, the EM transmission in a hollow waveguide, periodically
loaded with SRR, may become feasible within a certain frequency band,
even if the transverse dimensions of the waveguide are much smaller than
the associated free-space wavelength (Marques, Martel, Mesa, & Medina,
2002). The systematic analysis of similar problems in the framework of trans-
formation electromagnetics was presented by Ochiai, Leonhardt & Nacher
(2008).
Tunneling of Electromagnetic Waves 551
0.6
0.5
1
0.4
2
V (x)
0.3
3
0.2
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
x
Figure 10 Spatial distributions of normalized velocity of energy transfer V x vE =c by
waves tunneling through the gradient barrier (y 0.75, n0 2.3) located on a homoge-
neous substrate with refractive index n 1.47, x z=dthe normalized coordinate
across the barrier. Curves 1, 2, and 3 relate to the normalized frequencies u 1, 1.5,
and 2.
vE 4 z
V x ; x ; 0 x 1; (145)
c U xx d
where is the profile of refractive index inside the barrier (20), the dimension-
less function is determined by the spatial distribution of field components.
A graph of V V x, presented on Figure 10 for different frequencies,
shows, that the velocity vE is spatially variable and highly dispersive. Note,
that vE depends on the global properties of the photonic barrier (20) via the
parameters y L2 =2L1 , with n being a local property. The difference
between the nonlocal velocity of energy transfer and the group velocity
Tunneling of Electromagnetic Waves 553
!
near-field zone is matter-free and the divergence of the total field E is zero,
!
the total field E is rotational-free, since it is curl in this zone is vanishing too;
! ! !
hence, E is not a transverse vector field. The components E T and E L are
linked by conditions (147), and the disappearance of longitudinal field indi-
cates that the transverse field is vanishing too; herein the system proves to be
in a stationary state, incapable to irradiate or absorb photons. The source for
the transverse field is shown to extend not only over the current-carrying
region, but also over the near-field zone of the source. Note, that a similar
analysis can be performed for the evanescent magnetic components of a near
field of magnetic dipole, presented by an aerial frame; the spatial depen-
dences of these components Hr and H are known to coincide with the spa-
tial dependences of the Er and E components of the electric dipole field
(Kong, 1986). Hence, the evanescent magnetic field in the near zone of a
!
magnetic dipole satisfies simultaneously the Maxwell equation div H 0
!
as well as the condition rot H 0. These results link the photon-tunneling
process to impossibility of localizing photons in space (Keller, 2001),
highlighting the nonlocality as one of the most intriguing aspects of the elec-
tromagnetic wave theory.
@2 1 @2 2 @
2 2 : (148)
@z 2 v @t iT0 v2 @t
@t @ ln jT j
p , T : (149)
@ @
The first equality in (149) defines so-called phase time p; note, that
under the barrier the phase is not accumulated, and the phase shift t comes
mainly from the boundaries. However, using this concept in the interpre-
tation of measurements of penetration times for GHz microwaves in a sub-
cutoff waveguide (Ranfagni, Ruggeri, Susini, Agresti, & Sandri, 2001) has
revealed a paradoxical result; the dependencies of both the real p and imag-
inary T tunneling times upon the length of barrier contained two ascending
branches, separated by one descending branch. To understand the origin of
the maximum and minimum on the graphs of p and T, restricting this
descending branch, further investigations are needed.
The interpretation of evanescent modes as the stochastic motions of pho-
tons admits the generalization to virtual photons. Some peculiar properties
of these modes, resembling the virtual particles in quantum electrodynamics,
can be stressed:
i. The attempts to quantize the evanescent modes in a medium with con-
stant frequency-independent refractive index has revealed an important
feature of these modes; unlike the propagating waves the commutator
of field operators between two space-like separated points in the near-
field zone does not vanish (Stahlhofen & Nimtz, 2006).
ii. To provide a valid recording of the tunneling signal, it is necessary that
its power be above the thermal noise level of the receiver. The power of
this noise, generated under the condition of thermal equilibrium, is
given by the Nyquist formula (Landau & Lifschitz, 1984)
556 Alexander B. Shvartsburg et al.
PN kT , (150)
ACKNOWLEDGMENTS
We are much obliged to Professor E. Wolf for his kind suggestion to present this paper to
Progress in Optics. The authors appreciate Prof. A. Alu, N. Engheta, N. Erokhin,
V. Fortov, J. Pendry, N. Silin, L. Vazquez and L. Zelenii for their immutable interest to
these topics and valuable discussions. The work of A.S. was supported by the Far Eastern
Federal University, Project No. 14-08-2/3-20.
REFERENCES
Abramowitz, M., & Stegun, I. (1968). Handbook of mathematical functions. NY: Dover
Publications.
Adachi, S. (1988). Model dielectric constants of Si and Ge. Physical Review B, 38, 12966.
Aharonov, Y., Erez, N., & Reznik, B. (2002). Superoscillations and tunneling times. Physical
Review A, 65, 052124.
Alu, A., & Engheta, N. (2007). Light squeezing through arbitrarily shaped plasmonic chan-
nels and sharp bends. Physical Review B, 78, 035440.
Alu, A., Silveirinha, M. G., Salandrino, A., & Engheta, N. (2007). Epsilon-near-zero
metamaterials and electromagnetic sources: Tailoring the radiation phase pattern. Physical
Review B, 75, 155410.
Azzam, R. M. A. (2006). Phase shifts in frustrated total internal reflection and optical
tunneling by an embedded low-index thin film. Journal of the Optical Society of America A,
23(4), 960.
Baena, F. D., Jelinek, L., Marques, R., & Medina, F. (2005). Near-perfect tunneling and
amplification of evanescent electromagnetic waves in a waveguide filled by a
metamaterial: Theory and experiments. Physical Review B, 72, 075116.
Balqou, Ph., & Dutriaux, L. (1997). Dual optical tunneling times in frustrated total internal
reflection. Physical Review Letters, 78(5), 851.
Tunneling of Electromagnetic Waves 559
Born, M., & Wolf, E. (1999). Principles of optics (7th ed.). Cambridge University Press.
Borisov, A. G., de Abajo, Garsia, & Shabanov, S. V. (2005). Role of electron trapped
modes in extraordinary transmission in nanostructured materials. Physical Review B,
71, 075408.
Bulgakov, A. A., Bulgakov, S. A., & Nieto Vesperinas, M. (1998). Inhomogeneous waves
and energy localization in dielectric superlattices. Physical Review B, 58(8), 4438.
Buttiker, M., & Washburn, S. (2003). Ado about nothing much? Nature, 422, 271272.
Chiao, R., & Steinberg, A. M. (1997). Tunneling times and superluminality. In E. Wolf
(Ed.), Progress in optics: Vol. 37 (p. 345).
Cohen Tannoudji, C., Dupont Roc, J., & Grynberg, G. (1989). Photons and atoms: Intro-
duction to quantum electrodynamics. John Viley & Sons, Canada, Ltd.
Collins, R. (1991). Field theory for guided waves. New York, NY: John Wiley.
Dalarsson, M., Jaksic, Z., & Tassin, P. (2009). Exact analytical solution for oblique
incidenceon a graded index interface between a right-handed and left-handed material.
Journal of Optoelectronics and Biomedical Materials, 1(4), 345.
Darmanyan, S. A., Neviere, M., & Zayats, A. V. (2004). Light tunneling via resonant surface
plasmon polariton state. Physical Review B, 70, 075103.
de Dios-Leyva, M., & Drake-Perez, J. C. (2012). Group velocity and non-local energy trans-
port velocity in finite photonic structures. Journal of the Optical Society of America B, 9,
2275.
de Fornel, F. (2001). Evanescent waves: From Newtonian optics to atomic optics.
Springer ser. in optical sciences 73. Berlin: Springer.
De Witt Morette, & Foong, S. K. (1989). Path integral solutions of wave equations with
dissipation. Physical Review Letters, 62, 2201.
Deych, I., Livdan, L., & Lisyansaky, A. A. (1998). Resonant tunneling of electromagnetic
waves through polariton gaps. Physical Review E, 57, 7254.
Dresselhaus, G. (1955). Spin-orbit coupling in III-IV semiconductor heterostructures. Phys-
ical Review, 100(2), 580.
Dunster, T. M. (1990). Bessel functions of purely imaginary order. SIAM Journal on Mathe-
matical Analysis, 21, 995.
Edwards, B., Alu, A., Young, M. E., Silveirinha, M., & Engheta, N. (2008). Experimental
verification of epsilon-near-zero metamaterial coupling and energy squeezing using a
microwave waveguide. Physical Review Letters, 100, 033903.
Engheta, N. (2007). Circuits with light at nanoscales: Optical nanocircuits inspired by
metamaterials. Science, 317, 1698.
Efimov, V. V., Ivanov, O. V., & Sementsov, D. I. (2001). Phase relation for electromagnetic
waves reflected and transmitted through a dielectric slab. Journal of Optics A: Pure and
Applied Optics, 3, 514.
Eikhenwald, A. A. (2009). Energy flow with total internal reflection. Journal of Russian
Physical Chemical Society, 41, 131.
Erokhin, N. S., & Zakharov, V. E. (2011). Non-reflective propagationof electromagnetic
wave through heterogeneous plasma layer. Plasma Physics Reports, 37(9), 818.
Erokhin, N. S., Zueva, Yu. M., & Shvartsburg, A. B. (2013). Polarization phenomena in
gradient nanooptics. Quantum Electronics, 43(9), 785.
Fratalocchi, A., & Assanto, G. (2006). Dispersion spectroscopy of photonic barrier. Optics
Letters, 31(22), 3351.
Frias, W., Smolyakov, A., Hirose, A., Frias, W., Smolyakov, A., Hirose, A., et al. (2011).
Non-local energy transport in tunneling and plasmonic structures. Optics Express,
19(16), 15281.
Fu, C., Zhang, Z. M., & Tanner, D. B. (2005). Planar heterogeneous structures for coherent
emission of radiation. Optics Letters, 30(14), 1873.
560 Alexander B. Shvartsburg et al.
Futterman, J. (1995). Magnetic Brewster angle. American Journal of Physics, 63, 471.
Gamow, G. (1928). Zur quantentheorie des atomkernes. (On quantum theory of atomic
nuclei). Zeitschrift f
ur Physik, 51, 204.
Garcia de Abajo, F. J., Gomez-Santos, G., Blanco, L. A., Borisov, A. G., & Shabanov, S. V.
(2005). Tunneling mechanism of light transmission through metallic films. Physical
Review Letters, 95, 067403.
Ghulinyan, M., Oton, C. J., Gaburro, Z., Pavesi, L., Toninelli, C., & Wiersma, D. S. (2005).
Zener tunneling of light waves in the optical superlattice. Physical Review Letters, 94,
127401.
Ginzburg, V. L. (1970). Propagation of electromagnetic waves in plasma. Oxford: Pergamon Press.
Grbic, A., & Eleftheriades, G. V. (2002). Experimental verification of backward-wave radiation
from a negative refractive index metaqmaterial. Journal of Applied Physics, 92(10), 5930.
Haibel, A., Nimtz, G., & Stahlhofen, A. A. (2001). Frustrated total reflection: The double -
prisms revisited. Physical Review E, 63, 047601.
Hartman, T. E. (1962). Tunneling of a wave packet. Applied Physics, 33, 3427.
Hayata, K., & Koshiba, M. (1995). Algebraically decaying modes of dielectric planar wave-
guides. Optics Letters, 20(10), 1131.
Huang, J. P., & Yu, K. W. (2005). Effective nonlinear optical properties of graded metal-
dielectric composite films of anisotropic particles. Journal of the Optical Society of America
B, 22(8), 1640.
Jackson, J. D. (1998). Classical electrodynamics. New York, NY: Wiley.
Kaplan, A. E. (2012). Hysteresis in cyclotron resonance based on weak-relativistic mass-effect
of the electron. Physical Review Letters, 109, 153901.
Keller, O. (2001). Optical tunneling: A fingerprint of the lack of photon localizability. Journal
of the Optical Society of America B, 18(2), 206214.
Kim, K. (2008). Excitation of S-polarized surface electromagnetic waves in homogeneous
dielectric media. Optics Express, 16(17), 13354.
Kolokolov, A. A., & Skrotsky, G. V. (1992). Interference of reactive components of an elec-
tromagnetic field. Soviet PhysicsUspekhi, 162, 165.
Kong, J. A. (1986). Electromagnetic wave theory. New York, NY: Wiley-Interscience.
Kuzelev, M. V., & Rukhadze, A. A. (2009). Methods of waves theory in dispersive media.
Singapore: World Scientific Publishing Co.
Lakhtakia, A., & Messier, R. (2005). Sculptured thin films: Nanoengineered morphology and optics.
Bellinghem, Washington: SPIE Press.
Landau, L. D., & Lifschitz, E. M. (1984). Electrodynamics of continuous media. Amsterdam:
Elsevier.
Lee, S. J., Moon, S. E., Ryu, Y. C., Kwak, M. H., Kim, E. T., & Han, S. K. (2003). Micro-
wave properties of compositionally graded (Ba, SR)TiO3 thin films according to the
direction of the composition gradient for tunable microwave applications. Applied Physics
Letters, 82(13), 2133.
Lekner, J. (2000). Omnidirectional reflection by multilayer dielectric mirrors. Journal of Optics
A: Pure and Applied Optics, 2, 349.
Leonhardt, U. (2006). Optical conforming mapping. Science, 312, 1777.
Li, M., Liu, X., Zhang, J., Ma, X., Sun, X., Li, Y., et al. (2007). Rigorous analysis of the
spatial dispersion effect in multilayer thin-film coatings. Journal of the Optical Society of
America A, 24(8), 2328.
Li, C.-F., & Wang, Q. (2001). Duration of tunneling photons in a frustrated-total-internal-
reflection structure. Journal of Optical Society of America B, 18(8), 1174.
Liaw, J. W., Chen, J.-H., Chen, C.-S., & Kuo, M.-K. (2009). Purcell effect on nanoshell
dimmer on single molecules fluorescence. Optics Express, 17, 13532.
Litchinitser, N. M., Maimistov, A. I., Gabitov, I. R., Sagdeev, R. Z., & Shalaev, V. M.
(2010). Metamaterials: Electromagnetic enhancementat zero-index transition. Optics
Letters, 33(20), 2350.
Tunneling of Electromagnetic Waves 561
Liu, R., Cheng, Q., Hand, T., Mock, J. J., Cui, T. J., Cummer, S. A., et al. (2008). Exper-
imental demonstration of electromagnetic tunneling through an epsilon-near-zero
metamaterial at microwave frequencies. Physical Review Letters, 100, 023903.
Liu, L., Hu, C., Zhang, Z., & Luo, X. (2009). Multi-passband tunneling effect in multilay-
ered epsilon-near-zero metamaterials. Optics Express, 17(14), 12183.
Longhi, S. (2005). Resonant tunneling in frustrated total internal reflection. Optics Letters, 30,
2781.
Lui, L., Hu, C., Zhao, Z., & Luo, X. (2009). Multipassband effect in multilayered epsilon
near zero metamaterial. Optics Express, 17(14), 12183.
Lukyanchuk, B. S., & Qiu, C.-W. (2008). Peculiarities of light scattering by spherical par-
ticles with radial anisotropy. Journal of the Optical Society of America A, 25, 1623.
Lukyanchuk, B. S., Miroshnichenko, A. E., & Kivshar, Yu. S. (2013). Fano resonances and
topological optics: An interplay of far and near field interference phenomena. Journal
of Optics, 15, 073001.
Maier, S. A. (2007). Plasmonics: Fundamentals and applications. New York, NY: Springer.
Marques, R., Martel, J., Mesa, F., & Medina, F. (2002). Left-handed media simulation and
transmission of EM waves in subwavelength split-ring-resonator-loaded metallic wave-
guides. Physical Review Letters, 89(18), 183901.
Maxwell-Garnett, J. C. (1904). Colours in metal glasses and in metallic films. Philosophical
Transactions of the Royal Society A, 203, 385.
Miroshnichenko, A. E., Kuznetsov, A. I., Wei, L., Fu, Y. H., Neshev, D., Lukyanchuk, B.,
et al. (2012). Magnetic light: Optical magnetism of dielectric nanoparticles. Optics &
Photonics News, 23(12), 35.
Mocella, V., Dardano, P., Rendina, I., & Cabrini, S. (2010). An extraordinary directive radi-
ation based on optical antimatter at near infrared. Optics Express, 18(24), 25068.
Nelin, E. A. (2007). Impedance model for quantum mechanical barrier problems. Physics
Uspekhi, 50, 293.
OBrien, S., & Pendry, J. P. (2002). Photonic band-gap effectsand magnetic activityin dielec-
tric composites. Journal of Physics: Condensed Matter, 14, 4035.
Ochiai, T., Leonhardt, U., & Nacher, J. C. (2008). A novel design of dielectric perfect invis-
ibility devices. Journal of Mathematical Physics, 49, 032903.
Olkhovsky, V. S., Recami, E., & Jakiel, J. (2004). Unified time analysis of photon and particle
tunneling. Physics Reports, 398, 133.
Pendry, J. B., Holden, A. J., Robbins, D. J., & Stewart, W. J. (1999). Magnetism from
conductors and enhanced nonlinear phenomena. Microwave Theory and Techniques,
47, 2075.
Pendry, J. B., Holden, A. J., Stewart, W. J., & Youngs, I. (1996). Extremely low frequency
plasmons in metallic mesostructures. Physical Review Letters, 76, 4773.
Pendry, J. B., Schurig, D., & Smith, D. (2006). Controlling electromagnetic fields. Science,
312, 1780.
Rand, S. C., Fisher, W. M., & Oliveira, S. L. (2008). Optically induced magnetization in
homogeneous, undoped dielectric media. Journal of the Optical Society of America B,
25(7), 1106.
Ranfagni, A., Mugnai, D., Fabeni, P., & Pazzi, G. P. (1991). Time-delay measurements in
narrowed waveguides as a test of tunneling. Applied Physics Letters, 58(7), 774.
Ranfagni, A., Fabeni, F., Pazzi, G. P., & Mugnai, D. (1993). Anomalous pulse delay in
microwave propagation: A plausible connection to the tunneling time. Physical Review
E, 48, 1473.
Ranfagni, A., Ruggeri, R., Susini, C., Agresti, A., & Sandri, P. (2001). Experimental evi-
dence of tunneling as a stochastic process. Physical Review E, 63, 025102.
Rayleigh, J. W. S. (1937). The theory of sound. London: McMillan & Co.
Recami, E. (2004). Superluminal tunneling through successive barriers: Does quantum
mechanics predict infinite group velocities? Journal of Modern Optics, 51, 913.
562 Alexander B. Shvartsburg et al.
Sarid, D., & Challener, W. (2010). Modern introduction to surface plasmons: Theory metamaterial
modelling and applications. NY: Cambridge University Press.
Schiff, L. I. (1968). Quantum mechanics. Tokyo: Mc Grow Hill.
Semouchkina, E., Werner, D. H., Semouchkin, G. B., & Pantano, C. (2010). An infrared
invisibility cloak composed of glass. Applied Physics Letters, 96, 233503.
Serebryannikov, A. E., Ozbay, E., & Nojima, S. (2014). Asymmetric transition of terahertz
waves using polar dielectrics. Optics Express, 22(3), 3075.
Shelby, R. A., Smith, D. R., & Schultz, S. (2001). Experimental verification of a negative
index of refraction. Science, 292, 77.
Shvartsburg, A. B., & Kuzmiak, V. (2010). Reflectionless tunneling of microwave radiation
through subwavelength slit in a waveguide. Journal of Optical Society of America B, 27(12),
2766.
Shvartsburg, A. B., Kuzmiak, V., & Petite, G. (2007). Optics of subwavelength gradient
nanofilms. Physics Reports, 452, 33.
Shvartsburg, A. B., & Maradudin, A. A. (2013). Waves in gradient metamaterials. Singapore:
WSPC.
Shvartsburg, A. B., Marklund, M., Brodin, G., & Stenflo, L. (2008). Superluminal tunneling
of microwaves in smoothly varying transmission lines. Physical Review E, 78, 016601.
Shvartsburg, A. B., & Petite, G. (2005). Concave and convex photonic barriers in gradient
optics. European Physical Journal D, 36, 111.
Siddique, O. F., Ramahi, O. M., & Kashanianfard, M. (2010). Frequency-selective energy
tunneling in wire-roaded narrow waveguide channels. Progress in Electromagnetic Research
Letters, 15, 153.
Silveirinha, M. G., & Engheta, N. (2006). Tunneling of electromagnetic energy through sub-
wavelength channels and bends using 0-near-zero materials. Physical Review Letters, 97,
157403.
Silveirinha, M. G., & Engheta, N. (2009). Transporting of image through the subwavelength
hole. Physical Review Letters, 102, 103902.
Smith, D. D., & Fuller, K. A. (2002). Photonic bandgaps in Mie scattering by con-
centrationally stratified spheres. The Journal of the Optical Society of America B, 19(10),
2449.
Smith, D. R., Vier, D. C., Koschny, T., & Soukolis, M. C. (2005). Electromagnetic param-
eter retrieval from inhomogeneous metamaterials. Physical Review E, 71, 036617.
Stahlhofen, A. A., & Nimtz, G. (2006). Evanescent modes and virtual photons. Europhysics
Letters, 76(2), 189.
Sumetsky, M. (2007). Radiation loss of a nanotaper: Singular Gaussian beam model. Optics
Express, 15(4), 1480.
Tretyakov, S. A. (2003). Analytical modelling in applied electromagnetics. Norwood, MA, USA:
Artech House, INC.
Vahala, K. J. (2003). Optical microcavities. Nature, 424, 839.
Veselago, V. G. (1968). The electrodynamics of substanceswith simultaneously negative
and . Soviet PhysicsUspekhi, 10, 509.
Vesseur, E., Coenen, T., Caglayan, H., Engheta, N., & Polman, A. (2013). Experimental
verifications of n 0 structures for visible light. Physical Review Letters, 110, 013902.
Wang, J., Li, L., & Jia, S. (2008). Nonlinear tunneling of optical similaritons in nonlinear
waveguides. Journal of the Optical Society of America B, 25(8), 1254.
Wannier, G. H. (1955). Elements of solid state theory. London: Cambridge University Press.
Wilson, D. W., Glytsis, E. N., & Gaylord, T. X. (1993). Electron waveguiding characteristics
and ballistic current capacity of semiconductor quantum slabs. IEEE Journal of Quantum
Electron, 29, 1364.
Winful, H. G. (2003). Nature of superluminal barrier tunneling. Physical Review Letters,
90(2), 023901.
Tunneling of Electromagnetic Waves 563
Yang, F., & Sambles, J. R. (2002). Resonant transition of microwaves through a narrow
metallic slit. Physical Review Letters, 89, 063901.
Yeh, P. (1985). Resonant tunneling of electromagnetic radiation in superlattice structures.
Journal of the Optical Society of America A, 2(4), 568.
Zayatz, A. V., Smolyaninov, I. I., & Maradudin, A. A. (2005). Nano-optics of surface
plasmon polaritons. Physics Reports, 408, 131.
Zhang, J., Luo, Y., & Mortensen, N. A. (2010). Transmission of electromagnetic waves
through sub-wavelength channels. Optics Express, 18(4), 3864.
Zhao, Q., Zhou, J., Zhang, J., & Lippens, F. (2009). Mie resonance- based dielectric
metamaterial. Materials Today, 12, 60.
Ziolkowski, R. W., & Heyman, E. (2001). Negative effective permeability and left-handed
materials at optical frequencies. Physical Review E, 64, 056625.
This page intentionally left blank
AUTHOR INDEX
565
566 Author Index
O Passow, T., 59
OBrien, J.L., 78, 390, 424425 Patel, R.B., 51
OBrien, S., 521522 Paternostro, M., 82
Oates, C.W., 369, 371373, 426427 Paturzo, M., 173174, 239
Oberthaler, M.K., 373 Pavesi, L., 520
Obod, Y.A., 490558 Paz, S., 144
Ochoa, E., 175 Pazzi, G.P., 538539
Oesterhelt, D., 176177 Peaudecerf, B., 358
Ohnishi, K., 127128 Pedrosa, H.N., 59
Okada, M., 61 Pegg, D.T., 348350, 392
Okamoto, N., 171172 Pelc, J.S., 96
Okamoto, R., 390, 424425 Pellizzari, T., 350, 407408
Okamoto, T., 175 Pelton, M., 45, 47, 49
Olivares, S., 353, 367, 404 Pendry, J.B., 543, 550
Oliveira, S.L., 520521 Pendry, J.P., 521522
Olivos-Perez, L.I., 179 Penfield, P., 441, 459460
Olkhovsky, V.S., 491 Pennebaker, W.B., 134135, 137138
Olver, F.W.J., 290292 Pepper, M., 33
Ong, J.R., 96 Perez, J., 144
Ono, T., 396397 Perez-Delgado, C.A., 396
Oppenheim, A.V., 136137 Perina, J., 911, 1718, 349350, 390392
Oravec, M., 119120, 125126, 141, 153, Pernicka, J.C., 209210
169170, 239, 244245 Peters, N.A., 77
Orwa, J., 68 Peterson, C.G., 94
Oton, C.J., 520 Peterson, J., 59
Ou, Z.Y., 50 Petite, G., 492, 501502, 539
Ouerhani, Y., 136138, 154 Petroff, P.M., 4344, 79, 89
Oughstun, K.E., 264343 Petz, D., 385386
Ozaktas, H.M., 131132, 241242 Pezze, L., 359360, 367, 389390, 396397,
Ozbay, E., 13, 518 399, 424425
Pfeifer, R.N.C., 451
Pfitzmann, A., 139
P Phlypo, R., 209
Palombini, C.L., 341343 Phong, B., 193194
Pan, J.-W., 358 Piazza, F., 373
Pandalai, S.G., 209210 Piederrie`re, Y., 210211, 219221
Panotopoulos, G., 239 Piestun, R., 166167
Pant, M., 96 Pinel, O., 353
Pantano, C., 522 Pinel, P., 241
Papoulis, A., 205 Priz, G., 171172
Paris, M.G.A., 353, 367, 396397, 404 Pitaevskii, L.P., 467468, 471472
Park, G., 128 Pitkin, M., 346347, 366367, 388
Park, H., 8283 Pittman, T.B., 78, 88
Park, H.M., 197200 Plant, J., 47
Parka, J., 179 Plenio, M.B., 350, 407408, 426427
Parkins, A.S., 18 Plick, W.N., 349350, 396
Parks, T., 136137 Podolky, B., 75
Pasquini, T.A., 373 Poggi, C., 210211, 219221
Author Index 577
A B
Abraham formulation, 450451 Barbaras spectrum, 155, 155f
AbrahamMinkowski debate, 440 Bayesian approach
Abraham momentum density decoherence-free interferometry,
field momentum, 458459 390392
transmitted, 457 definition, 349350
Adapted all-numerical correlator estimation theory
AUC, 157159, 158f classical parameter, 377379
comparison, FFT and NO_FT algorithm, quantum parameter, 386387
161162, 163f realistic interferometry
correlation results, 159161, 160f, 162f phase diffusion, 420421
drawbacks and stringent requirements, photonic losses, 419420
156 Bayesian inference, 348
face database, 157, 158f BEC interferometry, 373
numerical algorithm, 156157, 157f Brewster effect
reconstructed image, 159, 159f generating function, 541
reconstruction, input plane, 160161, polarization structure, 542
161f problems, 538
ROC curve, 157159, 158f Brillouin precursor
Adaptive block scanning, 242 application, 301302
Antibunching asymptotic analysis, 294
example, 9f characteristic, 302303
Hanbury Brown and Twiss dynamic behavior, 302303, 303f
experiment, 8f first-order saddle points, 304305,
optical microcavity, 25 307308
resonant fluorescence, 24 limited validity, 300301
second-order correlation function, 78 second-order saddle point, 306307
Area under curve (AUC), 157159, 158f
Asymmetric cryptography, 235 C
Asymptotic analysis Cauchys residue theorem, 289
Brillouin precursor field, 294 Chu formulation
Cauchys residue theorem, 289 electrodynamics, moving media, 443444
dynamical behavior, 294 field and material contributions, 468470
FourierLaplace integral representation, Cipher, digital encryption, 140
289 Circular polarimeter, 220221, 220221f
Olver-type path, 290, 291f Classical electrodynamics
single-resonance Lorentz model Abraham formulation, 450451
dielectric, 292 EinsteinLaub formulation, 450451
Sommerfeld precursor, 292 energy and momentum continuity, 441
space-time intervals, 295 momentum density and stress tensors,
Autonomous underwater vehicles 451, 451t
(AUVs), 209 moving media
583
584 Subject Index
L classical electrodynamics
LempelZivWelch technique, 135 Abraham formulation, 450451
Lenas spectrum, 155, 155f Amperian formulation, 444445
Light interference, 346 Chu formulation, 443444
Linear canonical transforms (LCTs), Doppler shifted photon, 448449, 449f
162164 EinsteinLaub formulation, 450451
Linearly crossed polarized (LCP) images, energy and momentum continuity, 441
230231, 230f Minkowski formulation, 442443
Linearly parallel polarized (LPP) images, plane wave, 446448
230231, 230f electromagnetic fields, 439
Linear polarimeter, 219220, 219220f electromagnetic force
Linear polarizer, 212213, 214t equivalence of total force, 453454,
Livens momentum density, 457 453f
Lorentz model dielectric, 264266 equivalence of total time-average force,
Lossless compression 453f, 454455
image transmission, OIP, 234 linear media, 439, 439t
LempelZivWelch technique, 135 electromagnetic momentum
Lossy compression field momentum, 458460
image transmission, OIP, 234 momentum densities, comparison of,
JPEG, 135 455458
lossy encoding, 136 electromagnetic stress
Lossy interferometer. See Photonic losses dielectric fluids, 462464
momentum flow into media, 460462
radiation pressure, 464467, 465f, 467f
M electromagnetic wave theory, 437438
MachZehnder (MZ) interferometer fundamental aim, 440
BEC, 373 Minkwoski stress tensor, 474475
classical theory, 346 momentum and stress, dispersive media
coherent + squeezed-vacuum field and material contributions,
interferometry, 366367 468470
coherent state interferometry, 364365 time-averaged energy and momentum
definite photon-number state propagation, 470472
interferometry optical manipulation
Heisenberg scaling, 367369 classical optical tweezers, 476477,
simple error propagation formula, 369 476f
twin-Fock state, 367369 description, 438439
FabryPerrot interferometer, 369370, Gaussian beam, 475476, 476f
370f interference, 477, 478f
Fock state interferometry, 365366 modeling process, 475
Michelson interferometer, 369, 370f radiation pressure, 477478
phase-sensing uncertainty optical tweezers, 438
JordanSchwinger map, 363364 plane wave scattering
standard configuration, 362363, 362f cylinder, 478481
Ramsey interferometry, 371373, 372f sphere, 481485
shot noise, 347 radiation pressure, 438
Macroscopic theory, optical momentum stationary media, 473
AbrahamMinkowski debate, 440 total average electromagnetic force, 474
causality requirements, 474 Magnetometry interferometer, 369
590 Subject Index
VOLUME 1 (1961)
1 The modern development of Hamiltonian optics, R.J. Pegis 1 29
2 Wave optics and geometrical optics in optical design, K. Miyamoto 31 66
3 The intensity distribution and total illumination of aberration-free diffraction
images, R. Barakat 67 108
4 Light and information, D. Gabor 109 153
5 On basic analogies and principal differences between optical and electronic
information, H. Wolter 155 210
6 Interference color, H. Kubota 211 251
7 Dynamic characteristics of visual processes, A. Fiorentini 253 288
8 Modern alignment devices, A.C.S. Van Heel 289 329
VOLUME 2 (1963)
1 Ruling, testing and use of optical gratings for high-resolution spectroscopy,
G.W. Stroke 1 72
2 The metrological applications of diffraction gratings, J.M. Burch 73 108
3 Diffusion through non-uniform media, R.G. Giovanelli 109 129
4 Correction of optical images by compensation of aberrations and by spatial
frequency filtering, J. Tsujiuchi 131 180
5 Fluctuations of light beams, L. Mandel 181 248
6 Methods for determining optical parameters of thin films, F. Abele`s 249 288
VOLUME 3 (1964)
1 The elements of radiative transfer, F. Kottler 1 28
2 Apodisation, P. Jacquinot, B. Roizen-Dossier 29 186
3 Matrix treatment of partial coherence, H. Gamo 187 332
VOLUME 4 (1965)
1 Higher order aberration theory, J. Focke 1 36
2 Applications of shearing interferometry, O. Bryngdahl 37 83
3 Surface deterioration of optical glasses, K. Kinosita 85 143
4 Optical constants of thin films, P. Rouard, P. Bousquet 145 197
5 The Miyamoto-Wolf diffraction wave, A. Rubinowicz 199 240
6 Aberration theory of gratings and grating mountings, W.T. Welford 241 280
7 Diffraction at a black screen, Part I: Kirchhoffs theory, F. Kottler 281 314
VOLUME 5 (1966)
1 Optical pumping, C. Cohen-Tannoudji, A. Kastler 1 81
2 Non-linear optics, P.S. Pershan 83 144
Volumes IXL were previously distinguished by roman rather than by arabic numerals.
601
602 Contents of Previous Volumes
VOLUME 6 (1967)
1 Recent advances in holography, E.N. Leith, J. Upatnieks 1 52
2 Scattering of light by rough surfaces, P. Beckmann 53 69
3 Measurement of the second order degree of coherence,
M. Frantcon, S. Mallick 71 104
4 Design of zoom lenses, K.Yamaji 105 170
5 Some applications of lasers to interferometry, D.R. Herriot 171 209
6 Experimental studies of intensity fluctuations in lasers,
J.A. Armstrong, A.W. Smith 211 257
7 Fourier spectroscopy, G.A. Vanasse, H. Sakai 259 330
8 Diffraction at a black screen, Part II: electromagnetic theory, F. Kottler 331 377
VOLUME 7 (1969)
1 Multiple-beam interference and natural modes in open resonators,
G. Koppelman 1 66
2 Methods of synthesis for dielectric multilayer filters, E. Delano, R.J. Pegis 67 137
3 Echoes at optical frequencies, I.D. Abella 139 168
4 Image formation with partially coherent light, B.J. Thompson 169 230
5 Quasi-classical theory of laser radiation, A.L. Mikaelian, M.L. Ter-Mikaelian 231 297
6 The photographic image, S. Ooue 299 358
7 Interaction of very intense light with free electrons, J.H. Eberly 359 415
VOLUME 8 (1970)
1 Synthetic-aperture optics, J.W. Goodman 1 50
2 The optical performance of the human eye, G.A. Fry 51 131
3 Light beating spectroscopy, H.Z. Cummins, H.L. Swinney 133 200
4 Multilayer antireflection coatings, A. Musset, A. Thelen 201 237
5 Statistical properties of laser light, H. Risken 239 294
6 Coherence theory of source-size compensation in
interference microscopy, T. Yamamoto 295 341
7 Vision in communication, L. Levi 343 372
8 Theory of photoelectron counting, C.L. Mehta 373 440
VOLUME 9 (1971)
1 Gas lasers and their application to precise length measurements, A.L. Bloom 1 30
2 Picosecond laser pulses, A.J. Demaria 31 71
3 Optical propagation through the turbulent atmosphere, J.W. Strohbehn 73 122
Contents of Previous Volumes 603
VOLUME 10 (1972)
1 Bandwidth compression of optical images, T.S. Huang 1 44
2 The use of image tubes as shutters, R.W. Smith 45 87
3 Tools of theoretical quantum optics, M.O. Scully, K.G. Whitney 89 135
4 Field correctors for astronomical telescopes, C.G. Wynne 137 164
5 Optical absorption strength of defects in insulators, D.Y. Smith, D.L. Dexter 165 228
6 Elastooptic light modulation and deflection, E.K. Sittig 229 288
7 Quantum detection theory, C.W. Helstrom 289 369
VOLUME 11 (1973)
1 Master equation methods in quantum optics, G.S. Agarwal 1 76
2 Recent developments in far infrared spectroscopic techniques, H. Yoshinaga 77 122
3 Interaction of light and acoustic surface waves, E.G. Lean 123 166
4 Evanescent waves in optical imaging, O. Bryngdahl 167 221
5 Production of electron probes using a field emission source, A.V. Crewe 223 246
6 Hamiltonian theory of beam mode propagation, J.A. Arnaud 247 304
7 Gradient index lenses, E.W. Marchand 305 337
VOLUME 12 (1974)
1 Self-focusing, self-trapping, and self-phase modulation of laser beams,
O. Svelto 1 51
2 Self-induced transparency, R.E. Slusher 53 100
3 Modulation techniques in spectrometry, M. Harwit, J.A. Decker Jr 101 162
4 Interaction of light with monomolecular dye layers, K.H. Drexhage 163 232
5 The phase transition concept and coherence in atomic emission, R. Graham 233 286
6 Beam-foil spectroscopy, S. Bashkin 287 344
VOLUME 13 (1976)
1 On the validity of Kirchhoffs law of heat radiation for a body in a
nonequilibrium environment, H.P. Baltes 1 25
2 The case for and against semiclassical radiation theory, L. Mandel 27 68
3 Objective and subjective spherical aberration measurements of the human
eye, W.M. Rosenblum, J.L. Christensen 69 91
4 Interferometric testing of smooth surfaces, G. Schulz, J. Schwider 93 167
5 Self focusing of laser beams in plasmas and semiconductors,
M.S. Sodha, A.K. Ghatak, V.K.Tripathi 169 265
6 Aplanatism and isoplanatism, W.T. Welford 267 292
604 Contents of Previous Volumes
VOLUME 14 (1976)
1 The statistics of speckle patterns, J.C. Dainty 1 46
2 High-resolution techniques in optical astronomy, A. Labeyrie 47 87
3 Relaxation phenomena in rare-earth luminescence, L.A. Riseberg, M.J. Weber 89 159
4 The ultrafast optical Kerr shutter, M.A. Duguay 161 193
5 Holographic diffraction gratings, G. Schmahl, D. Rudolph 195 244
6 Photoemission, P.J. Vernier 245 325
7 Optical fibre waveguides a review, P.J.B. Clarricoats 327 402
VOLUME 15 (1977)
1 Theory of optical parametric amplification and oscillation, W. Brunner,
H. Paul 1 75
2 Optical properties of thin metal films, P. Rouard, A. Meessen 77 137
3 Projection-type holography, T. Okoshi 139 185
4 Quasi-optical techniques of radio astronomy, T.W. Cole 187 244
5 Foundations of the macroscopic electromagnetic theory of dielectric media,
J. Van Kranendonk, J.E. Sipe 245 350
VOLUME 16 (1978)
1 Laser selective photophysics and photochemistry, V.S. Letokhov 1 69
2 Recent advances in phase profiles generation, J.J. Clair, C.I. Abitbol 71 117
3 Computer-generated holograms: techniques and applications, W.-H. Lee 119 232
4 Speckle interferometry, A.E. Ennos 233 288
5 Deformation invariant, space-variant optical pattern recognition,
D. Casasent, D. Psaltis 289 356
6 Light emission from high-current surface-spark discharges, R.E. Beverly III 357 411
7 Semiclassical radiation theory within a quantum-mechanical framework,
I.R. Senitzky 413 448
VOLUME 17 (1980)
1 Heterodyne holographic interferometry, R. Dandliker 1 84
2 Doppler-free multiphoton spectroscopy, E. Giacobino, B. Cagnac 85 161
3 The mutual dependence between coherence properties of light and
nonlinear optical processes, M. Schubert, B. Wilhelmi 163 238
4 Michelson stellar interferometry, W.J. Tango, R.Q. Twiss 239 277
5 Self-focusing media with variable index of refraction, A.L. Mikaelian 279 345
VOLUME 18 (1980)
1 Graded index optical waveguides: a review, A. Ghatak, K. Thyagarajan 1 126
2 Photocount statistics of radiation propagating through random and
nonlinear media, J. Perina 127 203
3 Strong fluctuations in light propagation in a randomly inhomogeneous
medium, V.I. Tatarskii, V.U. Zavorotnyi 204 256
4 Catastrophe optics: morphologies of caustics and their diffraction patterns,
M.V. Berry, C. Upstill 257 346
Contents of Previous Volumes 605
VOLUME 19 (1981)
1 Theory of intensity dependent resonance light scattering and resonance
fluorescence, B.R. Mollow 1 43
2 Surface and size effects on the light scattering spectra of solids, D.L. Mills,
K.R. Subbaswamy 45 137
3 Light scattering spectroscopy of surface electromagnetic waves in solids,
S. Ushioda 139 210
4 Principles of optical data-processing, H.J. Butterweck 211 280
5 The effects of atmospheric turbulence in optical astronomy, F. Roddier 281 376
VOLUME 20 (1983)
1 Some new optical designs for ultra-violet bidimensional detection of
astronomical objects, G. Courte`s, P. Cruvellier, M. Detaille 1 61
2 Shaping and analysis of picosecond light pulses, C. Froehly, B. Colombeau,
M. Vampouille 63 153
3 Multi-photon scattering molecular spectroscopy, S. Kielich 155 261
4 Colour holography, P. Hariharan 263 324
5 Generation of tunable coherent vacuum-ultraviolet radiation, W. Jamroz,
B.P. Stoicheff 325 380
VOLUME 21 (1984)
1 Rigorous vector theories of diffraction gratings, D. Maystre 1 67
2 Theory of optical bistability, L.A. Lugiato 69 216
3 The Radon transform and its applications, H.H. Barrett 217 286
4 Zone plate coded imaging: theory and applications, N.M. Ceglio,
D.W. Sweeney 287 354
5 Fluctuations, instabilities and chaos in the laser-driven nonlinear ring cavity,
J.C. Englund, R.R. Snapp, W.C. Schieve 355 428
VOLUME 22 (1985)
1 Optical and electronic processing of medical images, D. Malacara 1 76
2 Quantum fluctuations in vision, M.A. Bouman, W.A. Van De Grind,
P. Zuidema 77 144
3 Spectral and temporal fluctuations of broad band laser radiation,
A.V. Masalov 145 196
4 Holographic methods of plasma diagnostics, G.V. Ostrovskaya,
Yu.I. Ostrovsky 197 270
5 Fringe formations in deformation and vibration measurements using laser
light, I. Yamaguchi 271 340
6 Wave propagation in random media: a systems approach, R.L. Fante 341 398
VOLUME 23 (1986)
1 Analytical techniques for multiple scattering from rough surfaces,
J.A. DeSanto, G.S. Brown 1 62
2 Paraxial theory in optical design in terms of Gaussian brackets, K. Tanaka 63 111
606 Contents of Previous Volumes
3 Optical films produced by ion-based techniques, P.J. Martin, R.P. Netterfield 113 182
4 Electron holography, A. Tonomura 183 220
5 Principles of optical processing with partially coherent light, F.T.S. Yu 221 275
VOLUME 24 (1987)
1 Micro Fresnel lenses, H. Nishihara, T. Suhara 1 37
2 Dephasing-induced coherent phenomena, L. Rothberg 39 101
3 Interferometry with lasers, P. Hariharan 103 164
4 Unstable resonator modes, K.E. Oughstun 165 387
5 Information processing with spatially incoherent light, I. Glaser 389 509
VOLUME 25 (1988)
1 Dynamical instabilities and pulsations in lasers, N.B. Abraham, P. Mandel,
L.M. Narducci 1 190
2 Coherence in semiconductor lasers, M. Ohtsu, T. Tako 191 278
3 Principles and design of optical arrays, Wang Shaomin, L. Ronchi 279 348
4 Aspheric surfaces, G. Schulz 349 415
VOLUME 26 (1988)
1 Photon bunching and antibunching, M.C. Teich, B.E.A. Saleh 1 104
2 Nonlinear optics of liquid crystals, I.C. Khoo 105 161
3 Single-longitudinal-mode semiconductor lasers, G.P. Agrawal 163 225
4 Rays and caustics as physical objects, Yu.A. Kravtsov 227 348
5 Phase-measurement interferometry techniques, K. Creath 349 393
VOLUME 27 (1989)
1 The self-imaging phenomenon and its applications, K. Patorski 1 108
2 Axicons and meso-optical imaging devices, L.M. Soroko 109 160
3 Nonimaging optics for flux concentration, I.M. Bassett, W.T. Welford,
R. Winston 161 226
4 Nonlinear wave propagation in planar structures, D. Mihalache,
M. Bertolotti, C. Sibilia 227 313
5 Generalized holography with application to inverse scattering and
inverse source problems, R.P. Porter 315 397
VOLUME 28 (1990)
1 Digital holography computer-generated holograms, O. Bryngdahl,
F. Wyrowski 1 86
2 Quantum mechanical limit in optical precision measurement and
communication, Y. Yamamoto, S. Machida, S. Saito, N. Imoto,
T. Yanagawa, M. Kitagawa 87 179
3 The quantum coherence properties of stimulated Raman scattering,
M.G. Raymer, I.A. Walmsley 181 270
4 Advanced evaluation techniques in interferometry, J. Schwider 271 359
5 Quantum jumps, R.J. Cook 361 416
Contents of Previous Volumes 607
VOLUME 29 (1991)
1 Optical waveguide diffraction gratings: coupling between guided modes,
D.G. Hall 1 63
2 Enhanced backscattering in optics, Yu.N. Barabanenkov, Yu.A. Kravtsov,
V.D. Ozrin, A.I. Saichev 65 197
3 Generation and propagation of ultrashort optical pulses, I.P. Christov 199 291
4 Triple-correlation imaging in optical astronomy, G. Weigelt 293 319
5 Nonlinear optics in composite materials. 1. Semiconductor and metal
crystallites in dielectrics, C. Flytzanis, F. Hache, M.C. Klein, D. Ricard,
Ph. Roussignol 321 411
VOLUME 30 (1992)
1 Quantum fluctuations in optical systems, S. Reynaud, A. Heidmann,
E. Giacobino, C. Fabre 1 85
2 Correlation holographic and speckle interferometry, Yu.I. Ostrovsky,
V.P. Shchepinov 87 135
3 Localization of waves in media with one-dimensional disorder,
V.D. Freilikher, S.A. Gredeskul 137 203
4 Theoretical foundation of optical soliton concept in fibers, Y. Kodama,
A. Hasegawa 205 259
5 Cavity quantum optics and the quantum measurement process, P. Meystre 261 355
VOLUME 31 (1993)
1 Atoms in strong fields: photoionization and chaos, P.W. Milonni,
B. Sundaram 1 137
2 Light diffraction by relief gratings: a macroscopic and microscopic view,
E. Popov 139 187
3 Optical amplifiers, N.K. Dutta, J.R. Simpson 189 226
4 Adaptive multilayer optical networks, D. Psaltis, Y. Qiao 227 261
5 Optical atoms, R.J.C. Spreeuw, J.P. Woerdman 263 319
6 Theory of Compton free electron lasers, G. Dattoli, L. Giannessi,
A. Renieri, A. Torre 321 412
VOLUME 32 (1993)
1 Guided-wave optics on silicon: physics, technology and status, B.P. Pal 1 59
2 Optical neural networks: architecture, design and models, F.T.S. Yu 61 144
3 The theory of optimal methods for localization of objects in pictures,
L.P. Yaroslavsky 145 201
4 Wave propagation theories in random media based on the path-integral
approach, M.I. Charnotskii, J. Gozani, V.I. Tatarskii, V.U. Zavorotny 203 266
5 Radiation by uniformly moving sources. Vavilov-Cherenkov effect,
Doppler effect in a medium, transition radiation and associated
phenomena, V.L. Ginzburg 267 312
6 Nonlinear processes in atoms and in weakly relativistic plasmas,
G. Mainfray, C. Manus 313 361
608 Contents of Previous Volumes
VOLUME 33 (1994)
1 The imbedding method in statistical boundary-value wave problems,
V.I. Klyatskin 1 127
2 Quantum statistics of dissipative nonlinear oscillators, V. Perinova,
A. Luks 129 202
3 Gap solitons, C.M. De Sterke, J.E. Sipe 203 260
4 Direct spatial reconstruction of optical phase from phase-modulated
images, V.I. Vlad, D. Malacara 261 317
5 Imaging through turbulence in the atmosphere, M.J. Beran,
I. Oz-Vogt 319 388
6 Digital halftoning: synthesis of binary images, O. Bryngdahl,
T. Scheermesser, F. Wyrowski 389 463
VOLUME 34 (1995)
1 Quantum interference, superposition states of light, and nonclassical effects,
V. Buzrek, P.L. Knight 1 158
2 Wave propagation in inhomogeneous media: phase-shift approach,
L.P. Presnyakov 159 181
3 The statistics of dynamic speckles, T. Okamoto, T. Asakura 183 248
4 Scattering of light from multilayer systems with rough boundaries,
I. Ohldal, K. Navratil, M. Ohldal 249 331
5 Random walk and diffusion-like models of photon migration in turbid
media, A.H. Gandjbakhche, G.H. Weiss 333 402
VOLUME 35 (1996)
1 Transverse patterns in wide-aperture nonlinear optical systems,
N.N. Rosanov 1 60
2 Optical spectroscopy of single molecules in solids, M. Orrit, J. Bernard,
R. Brown, B. Lounis 61 144
3 Interferometric multispectral imaging, K. Itoh 145 196
4 Interferometric methods for artwork diagnostics, D. Paoletti, G. Schirripa
Spagnolo 197 255
5 Coherent population trapping in laser spectroscopy, E. Arimondo 257 354
6 Quantum phase properties of nonlinear optical phenomena, R. Tanas,
A. Miranowicz, Ts. Gantsog 355 446
VOLUME 36 (1996)
1 Nonlinear propagation of strong laser pulses in chalcogenide glass films,
V. Chumash, I. Cojocaru, E. Fazio, F. Michelotti, M. Bertolotti 1 47
2 Quantum phenomena in optical interferometry, P. Hariharan,
B.C. Sanders 49 128
3 Super-resolution by data inversion, M. Bertero, C. De Mol 129 178
4 Radiative transfer: new aspects of the old theory, Yu.A. Kravtsov,
L.A. Apresyan 179 244
5 Photon wave function, I. Bialynicki-Birula 245 294
Contents of Previous Volumes 609
VOLUME 37 (1997)
1 The Wigner distribution function in optics and optoelectronics,
D. Dragoman 1 56
2 Dispersion relations and phase retrieval in optical spectroscopy,
K.-E. Peiponen, E.M. Vartiainen, T. Asakura 57 94
3 Spectra of molecular scattering of light, I.L. Fabelinskii 95 184
4 Soliton communication systems, R.-J. Essiambre, G.P. Agrawal 185 256
5 Local fields in linear and nonlinear optics of mesoscopic systems,
O. Keller 257 343
6 Tunneling times and superluminality, R.Y. Chiao, A.M. Steinberg 345 405
VOLUME 38 (1998)
1 Nonlinear optics of stratified media, S. Dutta Gupta 1 84
2 Optical aspects of interferometric gravitational-wave detectors,
P. Hello 85 164
3 Thermal properties of vertical-cavity surface-emitting semiconductor
lasers, W. Nakwaski, M. Osi nski 165 262
4 Fractional transformations in optics, A.W. Lohmann, D. Mendlovic,
Z. Zalevsky 263 342
5 Pattern recognition with nonlinear techniques in the Fourier domain,
B. Javidi, J.L. Horner 343 418
6 Free-space optical digital computing and interconnection, J. Jahns 419 513
VOLUME 39 (1999)
1 Theory and applications of complex rays, Yu.A. Kravtsov, G.W. Forbes,
A.A. Asatryan 1 62
2 Homodyne detection and quantum-state reconstruction, D.-G. Welsch,
W. Vogel, T. Opatrny 63 211
3 Scattering of light in the eikonal approximation, S.K. Sharma,
D.J. Somerford 213 290
4 The orbital angular momentum of light, L. Allen, M.J. Padgett,
M. Babiker 291 372
5 The optical Kerr effect and quantum optics in fibers, A. Sizmann,
G. Leuchs 373 469
VOLUME 40 (2000)
1 Polarimetric optical fibers and sensors, T.R. Woli
nski 1 75
2 Digital optical computing, J. Tanida, Y. Ichioka 77 114
3 Continuous measurements in quantum optics, V. Perinova,
A. Luks 115 269
4 Optical systems with improved resolving power, Z. Zalevsky,
D. Mendlovic, A.W. Lohmann 271 341
5 Diffractive optics: electromagnetic approach, J. Turunen, M. Kuittinen,
F. Wyrowski 343 388
6 Spectroscopy in polychromatic fields, Z. Ficek, H.S. Freedhoff 389 441
610 Contents of Previous Volumes
VOLUME 41 (2000)
1 Nonlinear optics in microspheres, M.H. Fields, J. Popp, R.K. Chang 1 95
2 Principles of optical disk data storage, J. Carriere, R. Narayan, W.-H. Yeh,
C. Peng, P. Khulbe, L. Li, R. Anderson, J. Choi, M. Mansuripur 97 179
3 Ellipsometry of thin film systems, I. Ohldal, D. Franta 181 282
4 Optical true-time delay control systems for wideband phased array
antennas, R.T. Chen, Z. Fu 283 358
5 Quantum statistics of nonlinear optical couplers, J. Perina Jr, J. Perina 359 417
6 Quantum phase difference, phase measurements and Stokes operators,
A. Luis, L.L. Sanchez-Soto 419 479
7 Optical solitons in media with a quadratic nonlinearity, C. Etrich,
F. Lederer, B.A. Malomed, T. Peschel, U. Peschel 483 567
VOLUME 42 (2001)
1 Quanta and information, S. Ya. Kilin 1 91
2 Optical solitons in periodic media with resonant and off-resonant
nonlinearities, G. Kurizki, A.E. Kozhekin, T. Opatrny, B.A. Malomed 93 146
3 Quantum Zeno and inverse quantum Zeno effects, P. Facchi, S. Pascazio 147 217
4 Singular optics, M.S. Soskin, M.V. Vasnetsov 219 276
5 Multi-photon quantum interferometry, G. Jaeger, A.V Sergienko 277 324
6 Transverse mode shaping and selection in laser resonators, R. Oron,
N. Davidson, A.A. Friesem, E. Hasman 325 386
VOLUME 43 (2002)
1 Active optics in modern large optical telescopes, L. Noethe 1 69
2 Variational methods in nonlinear fiber optics and related fields,
B.A. Malomed 71 193
3 Optical works of L.V. Lorenz, O. Keller 195 294
4 Canonical quantum description of light propagation in dielectric media,
A. Luks, V. Perinova 295 431
5 Phase space correspondence between classical optics and quantum
mechanics, D. Dragoman 433 496
6 Slow and fast light, R.W. Boyd, D.J. Gauthier 497 530
7 The fractional Fourier transform and some of its applications to optics,
A. Torre 531 596
VOLUME 44 (2002)
1 Chaotic dynamics in semiconductor lasers with optical feedback, J. Ohtsubo 1 84
2 Femtosecond pulses in optical fibers, F.G. Omenetto 85 141
3 Instantaneous optics of ultrashort broadband pulses and rapidly varying
media, A.B. Shvartsburg, G. Petite 143 214
4 Optical coherence tomography, A.F. Fercher, C.K. Hitzenberger 215 301
5 Modulational instability of electromagnetic waves in inhomogeneous
and in discrete media, F.Kh. Abdullaev, S.A. Darmanyan, J. Garnier 303 366
Contents of Previous Volumes 611
VOLUME 45 (2003)
1 Anamorphic beam shaping for laser and diffuse light, N. Davidson,
N. Bokor 1 51
2 Ultra-fast all-optical switching in optical networks, I. Glesk, B.C. Wang,
L. Xu, V. Baby, P.R. Prucnal 53 117
3 Generation of dark hollow beams and their applications, J. Yin,
W. Gao, Y. Zhu 119 204
4 Two-photon lasers, D.J. Gauthier 205 272
5 Nonradiating sources and other invisible objects, G. Gbur 273 315
6 Lasing in disordered media, H. Cao 317 370
VOLUME 46 (2004)
1 Ultrafast solid-state lasers, U. Keller 1 115
2 Multiple scattering of light from randomly rough surfaces, A.V. Shchegrov,
A.A. Maradudin, E.R. Mendez 117 241
3 Laser-diode interferometry, Y. Ishii 243 309
4 Optical realizations of quantum teleportation, J. Gea-Banacloche 311 353
5 Intensity-field correlations of non-classical light, H.J. Carmichael,
G.T. Foster, L.A. Orozco, J.E. Reiner, P.R. Rice 355 404
VOLUME 47 (2005)
1 Multistep parametric processes in nonlinear optics, S.M. Saltiel,
A.A. Sukhorukov, Y.S. Kivshar 1 73
2 Modes of wave-chaotic dielectric resonators, H.E. T ureci, H.G.L. Schwefel,
Ph. Jacquod, A.D. Stone 75 137
3 Nonlinear and quantum optics of atomic and molecular fields, C.P. Search,
P. Meystre 139 214
4 Space-variant polarization manipulation, E. Hasman, G. Biener, A. Niv,
V. Kleiner 215 289
5 Optical vortices and vortex solitons, A.S. Desyatnikov, Y.S. Kivshar,
L.L. Torner 291 391
6 Phase imaging and refractive index tomography for X-rays and
visible rays, K. Iwata 393 432
VOLUME 48 (2005)
1 Laboratory post-engineering of microstructured optical fibers,
B.J. Eggleton, P. Domachuk, C. Grillet, E.C. Magi, H.C. Nguyen,
P. Steinvurzel, M.J. Steel 1 34
2 Optical solitons in random media, F. Abdullaev, J. Garnier 35 106
3 Curved diffractive optical elements: Design and applications, N. Bokor,
N. Davidson 107 148
4 The geometric phase, P. Hariharan 149 201
5 Synchronization and communication with chaotic laser systems,
A. Uchida, F. Rogister, J. Garca-Ojalvo, R. Roy 203 341
612 Contents of Previous Volumes
VOLUME 49 (2006)
1 Gaussian apodization and beam propagation, V.N. Mahajan 1 96
2 Controlling nonlinear optical processes in multi-level atomic systems,
A. Joshi, M. Xiao 97 175
3 Photonic crystals, H. Benisty, C. Weisbuch 177 313
4 Symmetry properties and polarization descriptors for an arbitrary
electromagnetic wavefield, C. Brosseau, A. Dogariu 315 380
5 Quantum cryptography, M. Dusek, N. L utkenhaus, M. Hendrych 381 454
6 Optical quantum cloning, N.J. Cerf , J. Fiurasek 455 545
VOLUME 50 (2007)
1 From millisecond to attosecond laser pulses, N. Bloembergen 1 12
2 Conical diffraction: Hamiltons diabolical point at the heart of crystal optics,
M.V. Berry, M.R. Jeffrey 13 50
3 Historical papers on the particle concept of light, O. Keller 51 95
4 Field quantization in optics, P.W. Milonni 97 135
5 The history of near-field optics, L. Novotny 137 184
6 Light tunneling, H.M. Nussenzveig 185 250
7 The influence of Youngs interference experiment on the development
of statistical optics, E. Wolf 251 273
8 Planck, photon statistics, and Bose-Einstein condensation,
D.M. Greenberger, N. Erez, M.O. Scully, A.A. Svidzinsky, M.S. Zubairy 275 330
VOLUME 51 (2008)
1 Negative refractive index metamaterials in optics, N.M. Litchinitser,
I.R. Gabitov, A.I. Maimistov, V.M. Shalaev 1 67
2 Polarization techniques for surface nonlinear optics, M. Kauranen,
S. Cattaneo 69 120
3 Electromagnetic fields in linear bianisotropic mediums, T.G. Mackay
A. Lakhtakia 121 209
4 Ultrafast optical pulses, C.R. Pollock 211 249
5 Quantum imaging, A. Gatti, E. Brambilla, L. Lugiato 251 348
6 Assessment of optical systems by means of point-spread functions,
J.J.M. Braat, S. van Haver, A.J.E.M. Janssen, P. Dirksen 349 468
7 The discrete Wigner function, G. Bj ork, A.B. Klimov, L.L. Sanchez-Soto 469 516
VOLUME 52 (2009)
1 Coherent backscattering and Anderson localization of light, C.M. Aegerter,
G. Maret 1 62
2 Soliton shape and mobility control in optical lattices, Y.V. Kartashov,
V.A. Vysloukh, L. Torner 63 148
3 Signal and quantum noise in optical communications and cryptography,
P. Gallion, F. Mendieta, S. Jiang 149 259
4 Invisibility cloaking by coordinate transformation, M. Yan, W. Yan, M. Qiu 261 304
Contents of Previous Volumes 613
VOLUME 53 (2009)
1 The resolution challenge in 3D optical microscopy, M. Martnez-Corral,
G. Saavedra 1 67
2 Transformation optics and the geometry of light, U. Leonhardt,
T.G. Philbin 69 152
3 Photorefractive solitons and their underlying nonlocal physics, E. DelRe,
B. Crosignani, P. Di Porto 153 200
4 Stimulated scattering effects of intense coherent light, G.S. He 201 292
5 Singular optics: Optical vortices and polarization singularities,
M.R. Dennis, K. OHolleran, M.J. Padgett 293 363
6 Quantum feed-forward control of light, U.L. Andersen, R. Filip 365 414
VOLUME 54 (2009)
1 Propagation-Invariant Optical Fields, J.Turunen, A.T. Friberg 1 88
2 Theoretical Tools for Quantum Optics in Structured Media,
D. Mogilevtsev, S. Kilin 89 148
3 Polarization and Coherence Optics: Historical Perspective, Status,
and Future Directions, C. Brosseau 149 208
4 Optical Quantum Computation, T.C. Ralph, G.J. Pryde 209 269
5 Science and Engineering of Photonic Crystals, K. Sakoda, J.W. Haus 271 317
6 The Mathematical Theory of Laser Beam-Splitting Gratings,
L.A. Romero, F.M. Dickey 319 386
VOLUME 55 (2010)
1 Optical Pulse Propagation in Biological Media: Theory and Numerical
Methods, M. Premaratne 1 40
2 Manipulating Light States by Single-Photon Addition and Subtraction,
M. Bellini, A. Zavatta 41 84
3 Vacuum-Induced Processes in Multilevel Atoms, M. Kiffner, M. Macovei,
J. Evers, C.H. Keitel 85 198
4 Adaptive Lens, Guoqiang Li 199 284
5 The Structure of Partially Coherent Fields, G. Gbur, T.D. Visser 285 342
VOLUME 56 (2011)
1 Space-Time Analogies in Optics V. Torres-Company, J. Lancis, P. Andres 1 80
2 Unconventional Polarization States: Beam Propagation, Focusing, and
Imaging, T.G. Brown 81 129
3 Quantum Scissors Finite-Dimensional States Engineering,
W. Leonski, A. Kowalewska-Kudaszyk 131 186
4 Polarization and Spectral Properties of Biphotons, M. Chekhova 187 226
5 Engineering Nonlinear Optic Sources of Photonic Entanglement,
J.P. Torres, K. Banaszek, I.A. Walmsley 227 331
614 Contents of Previous Volumes
VOLUME 57 (2012)
1 The Microscope in a Computer: Image Synthesis from Three-Dimensional
Full-Vector Solutions of Maxwells Equations at the Nanometer Scale,
Ilker R. C
apoglu, Jeremy D. Rogers, Allen Taflove, Vadim Backman 1 91
2 Microstructures and Nanostructures in Nature, Dora Kroisova 93 132
3 Quantitative Phase Imaging, Mustafa Mir, Basanta Bhaduri, Ru Wang,
Ruoyu Zhu, Gabriel Popescu 133 217
4 Tight Focusing of Light Beams: Effect of Polarization, Phase, and
Coherence, Ziyang Chen, Limin Hua, Jixiong Pu 219 260
5 Direct and Inverse Problems in the Theory of Light Scattering,
Daomu Zhao, Tao Wang 261 308
VOLUME 58 (2013)
1 Dynamic Photonic Materials Based on Liquid Crystals,
Luciano De Sio, Nelson Tabiryan, Timothy Bunning, Brian R. Kimball,
and Cesare Umeton 1 64
2 Invisibility Physics: Past, Present, and Future, Greg Gbur 65 114
3 Phase Anomalies in Micro-Optics,
Myun-Sik Kim, Toralf Schaf, Carsten Rockstuhl, and Hans Peter Herzig 115 197
4 Subwavelength Atom Localization, Kishore T. Kapale 199 250
5 Low-Dimensional Silicon Structures for Use in Photonic Circuits,
Tyler Roschuk, Iain F. Crowe, Andrew P. Knights, and Matthew P. Halsall 251 315
VOLUME 59 (2014)
1 Active Optical Metamaterials,
Sebastian Wuestner and Ortwin Hess 1 88
2 Spontaneous Parametric Down-Conversion in Nonlinear Layered
Structures, Jan Perina Jr. 89 158
3 Spatial Heterodyne Fourier-Transform Waveguide Spectrometers,
Aitor V. Velasco, Pavel Cheben, Mirosaw Florjanczyk, and Maria L. Calvo 159 208
4 Precursors and Dispersive Pulse Dynamics, A Century after the Sommerfeld-
Brillouin Theory: Part I. The Original Theory, Natalie A. Cartwright and
Kurt E. Oughstun 209 266
5 The Role of Coherence in Image Formation in Holographic Microscopy,
Radim Chmelik, Michala Slaba, Vera Kollarova, Tomas Slaby, Martin Lostak,
Jana Collakova, and Zbynek Dostal 267 336
CUMULATIVE INDEX VOLUMES 160
Volumes IXL were previously distinguished by roman rather than by arabic numerals.
615
616 Cumulative Index Volumes 160
Eberly, J.H.: Interaction of very intense light with free electrons 7, 359
Eggleton, B.J., P. Domachuk, C. Grillet, E.C. Magi, H.C. Nguyen, P.
Steinvurzel and M.J. Steel: Laboratory post-engineering of microstructured
optical fibers 48, 1
Englund, J.C., R.R. Snapp and W.C. Schieve: Fluctuations, instabilities and
chaos in the laser-driven nonlinear ring cavity 21, 355
Ennos, A.E.: Speckle interferometry 16, 233
Erez, N., see Greenberger, D.M. 50, 275
Essiambre, R.-J. and G.P. Agrawal: Soliton communication systems 37, 185
Etrich, C., F. Lederer, B.A. Malomed, T. Peschel and U. Peschel: Optical
solitons in media with a quadratic nonlinearity 41, 483
Evers, J., see Kiffner, M. 55, 85
Greenberger, D.M., N. Erez, M.O. Scully, A.A. Svidzinsky and M.S. Zubairy:
Planck, photon statistics, and Bose-Einstein condensation 50, 275
Grillet, C., see Eggleton, B.J. 48, 1
Ralph, T.C. and Pryde, G.J.: Optical quantum computation 54, 209
Raymer, M.G. and I.A. Walmsley: The quantum coherence properties of
stimulated Raman scattering 28, 181
Reiner, J.E., see Carmichael, H.J. 46, 355
Renieri, A., see Dattoli, G. 31, 321
Reynaud, S., A. Heidmann, E. Giacobino and C. Fabre: Quantum fluctuations
in optical systems 30, 1
Ricard, D., see Flytzanis, C. 29, 321
Rice, P.R., see Carmichael, H.J. 46, 355
Riseberg, L.A. and M.J. Weber: Relaxation phenomena in rare-earth
luminescence 14, 89
Risken, H.: Statistical properties of laser light 8, 239
Rockstuhl, C., see Kim, M.S. 58, 115
Roddier, F.: The effects of atmospheric turbulence in optical astronomy 19, 281
Rogers, J.D., see C apoglu, I.R. 57, 1
Rogister, F., see Uchida, A. 48, 203
Roizen-Dossier, B., see Jacquinot, P. 3, 29
Romero, L.A. and Dickey, F.M.: The mathematical theory of laser beam-
splitting gratings 54, 319
Ronchi, L., see Wang Shaomin 25, 279
Rosanov, N.N.: Transverse patterns in wide-aperture nonlinear optical systems 35, 1
Roschuk, T., I.F. Crowe, A.P. Knights, M.P. Halsall: Low-Dimensional
Silicon Structures for Use in Photonic Circuits 58, 251
Rosenblum, W.M. and J.L. Christensen: Objective and subjective spherical
aberration measurements of the human eye 13, 69
Rothberg, L.: Dephasing-induced coherent phenomena 24, 39
Rouard, P. and P. Bousquet: Optical constants of thin films 4, 145
Rouard, P. and A. Meessen: Optical properties of thin metal films 15, 77
Roussignol, Ph., see Flytzanis, C. 29, 321
Roy, R., see Uchida, A. 48, 203
Rubinowicz, A.: The Miyamoto-Wolf diffraction wave 4, 199
Rudolph, D., see Schmahl, G. 14, 195
628 Cumulative Index Volumes 160