Power-Line Communications: Smart Grid, Transmission, and Propagation

Journal of Electrical and Computer Engineering
Power-Line Communications:
Smart Grid, Transmission,
and Propagation
Guest Editors: Justinian Anatory, Moises V. Ribeiro, Andrea M. Tonello,
and Ahmed Zeddam
Power-Line Communications: Smart Grid,
Transmission, and Propagation

Guest Editors: Justinian Anatory, Moises V. Ribeiro,
Andrea M. Tonello, and Ahmed Zeddam
Copyright 2013 Hindawi Publishing Corporation. All rights reserved.
This is a special issue published in Journal of Electrical and Computer Engineering. All articles are open access articles distributed
under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, pro-
vided the original work is properly cited.
Editorial Board
The editorial board of the journal is organized into sections that correspond to
the subject areas covered by the journal.
Circuits and Systems

M. T. Abuelmaatti, Saudi Arabia H. Kuntman, Turkey Raj Senani, India
Ishfaq Ahmad, USA Parag K. Lala, USA Gianluca Setti, Italy
Dhamin Al-Khalili, Canada Shen-Iuan Liu, Taiwan Jose Silva-Martinez, USA
Wael M. Badawy, Canada Bin-Da Liu, Taiwan Nicolas Sklavos, Greece
Ivo Barbi, Brazil Joao Antonio Martino, Brazil Ahmed M. Soliman, Egypt
Martin A. Brooke, USA Pianki Mazumder, USA Dimitrios Soudris, Greece
Y. W. Chang, Taiwan Michel Nakhla, Canada Charles E. Stroud, USA
Tian-Sheuan Chang, Taiwan Sing Kiong Nguang, New Zealand Ephraim Suhir, USA
Chip Hong Chang, Singapore Shun-ichiro Ohmi, Japan Hannu Tenhunen, Sweden
Tzi-Dar Chiueh, Taiwan Mohamed A. Osman, USA George S. Tombras, Greece
M. Jamal Deen, Canada Ping Feng Pai, Taiwan Spyros Tragoudas, USA
A. El Wakil, UAE Marcelo Antonio Pavanello, Brazil Chi Kong Tse, Hong Kong
Denis Flandre, Belgium Marco Platzner, Germany Chi-Ying Tsui, Hong Kong
P. Franzon, USA Massimo Poncino, Italy Jan Van der Spiegel, USA
Andre Ivanov, Canada Dhiraj K. Pradhan, UK Chin-Long Wey, USA
Ebroul Izquierdo, UK F. Ren, USA Fei Yuan, Canada
Wen-Ben Jone, USA Gabriel Robins, USA
Yong-Bin Kim, USA Mohamad Sawan, Canada
Communications
Soene Aes, Canada Amoakoh Gyasi-Agyei, Ghana Samuel Pierre, Canada
Dharma Agrawal, USA Yaohui Jin, China Nikos C. Sagias, Greece
H. Arslan, USA Mandeep Jit Singh, Malaysia John N. Sahalos, Greece
Edward Au, China Peter Jung, Germany Christian Schlegel, Canada
Enzo Baccarelli, Italy Adnan Kavak, Turkey Vinod Sharma, India
Stefano Basagni, USA Rajesh Khanna, India Iickho Song, Korea
Jun Bi, China Kiseon Kim, Republic of Korea Ioannis Tomkos, Greece
Z. Chen, Singapore D. I. Laurenson, UK Chien Cheng Tseng, Taiwan
Rene Cumplido, Mexico Tho Le-Ngoc, Canada George Tsoulos, Greece
Luca De Nardis, Italy C. Leung, Canada Laura Vanzago, Italy
M.-Gabriella Di Benedetto, Italy Petri Mahonen, Germany Roberto Verdone, Italy
J. Fiorina, France Mohammad A. Matin, Bangladesh Guosen Yue, USA
Lijia Ge, China M. Najar, Spain Jian-Kang Zhang, Canada
Z. Ghassemlooy, UK M. S. Obaidat, USA
K. Giridhar, India Adam Panagos, USA
Signal Processing
S. S. Agaian, USA A. G. Constantinides, UK Karen Egiazarian, Finland
Panajotis Agathoklis, Canada Paul Dan Cristea, Romania W.-S. Gan, Singapore
Jaakko Astola, Finland Petar M. Djuric, USA Z. F. Ghassemlooy, UK
Tamal Bose, USA Igor Djurovic, Montenegro Ling Guan, Canada
Martin Haardt, Germany Antonio Napolitano, Italy Yannis Stylianou, Greece
Peter Handel, Sweden Sven Nordholm, Australia Ioan Tabus, Finland
Andreas Jakobsson, Sweden S. Panchanathan, USA Jarmo Henrik Takala, Finland
Jiri Jan, Czech Republic Periasamy K. Rajan, USA Clark N. Taylor, USA
S. Jensen, Denmark Cedric Richard, France A. H. Tewk, USA
Chi Chung Ko, Singapore W. Sandham, UK Jitendra Kumar Tugnait, USA
M. A. Lagunas, Spain Ravi Sankar, USA Vesa Valimaki, Finland
J. Lam, Hong Kong Dan Schonfeld, USA Luc Vandendorpe, Belgium
D. I. Laurenson, UK Ling Shao, UK Ari J. Visa, Finland
Riccardo Leonardi, Italy John J. Shynk, USA Jar Ferr Yang, Taiwan
S. Marshall, UK Andreas Spanias, USA
Contents
Power-Line Communications: Smart Grid, Transmission, and Propagation, Justinian Anatory,
Moises V. Ribeiro, Andrea M. Tonello, and Ahmed Zeddam
Volume 2013, Article ID 948598, 2 pages
Power Line Communications for Smart Grid Applications, Lars Torsten Berger, Andreas Schwager,
and J. Joaqun Escudero-Garzas
Radiation Mitigation for Power Line Communications Using Time Reversal, Amilcar Mescco,
Pascal Pagani, Michel Ney, and Ahmed Zeddam
An Overview of the HomePlug AV2 Technology, Larry Yonge, Jose Abad, Kaywan Afkhamie,
Lorenzo Guerrieri, Srinivas Katar, Hidayat Lioe, Pascal Pagani, Raaele Riva, Daniel M. Schneider,
and Andreas Schwager
A Study on the Optimal Receiver Impedance for SNR Maximization in Broadband PLC,
Massimo Antoniali, Andrea M. Tonello, and Fabio Versolatto
Enhancements of G3-PLC Technology for Smart-Home/Building Applications, Luca Di Bert,

Salvatore DAlessandro, and Andrea M. Tonello
Improved Maximum Likelihood S-FSK Receiver for PLC Modem in AMR,

Mohamed Chaker Bali and Chiheb Rebai
Hindawi Publishing Corporation
http://dx.doi.org/10.1155/2013/948598
Editorial
Justinian Anatory,1 Moises V. Ribeiro,2 Andrea M. Tonello,3 and Ahmed Zeddam4

1
School of Informatics, College of Informatics and Virtual Education, University of Dodoma, P.O. Box 490, Dodoma, Tanzania
2
Electrical Engineering Department, Faculdade de Engenharia, Federal University of Juiz de Fora, Campus Universitario,
36 033 330 Juiz de Fora, MG, Brazil
3
Wireless and Power Line Communications Lab, Universita di Udine, 33100 Udine, Italy
4
France Telecom, Orange Labs, 2, Avenue Pierre Marzin, 22 307 Lannion Cedex, France
Correspondence should be addressed to Justinian Anatory; janatory@udom.ac.tz
Received 28 January 2013; Accepted 28 January 2013
Copyright 2013 Justinian Anatory et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.
Power-line communication networks are gaining popularity peer-review process, six papers have been selected. These
in various service provisions such as in houses/offices, access papers make an inspiring ensemble with a topic spectrum
networks, in ships, aircrafts, trains, vehicles, in industry ranging from improved maximum likelihood detection of
systems control, and advanced metering infrastructure. This spread frequency shift keying (S-FSK) in PLC automatic
popularity is also striding towards smart grid implemen- meter reading (AMR), radiation mitigation for PLC, optimal
tations. However, the network structure affects the chan- receiver impedance design for SNR maximization in broad-
nel response which exhibits frequency selectivity and time band PLC, PLC for Smart Grid applications, improvements of
variant behavior. These effects are due to different terminal G3-PLC technology for Smart-Home/Building applications,
loads connected to such systems, number of branches, and and finally an overview of the HomePlug AV2 technology.
different branched line lengths. In addition, different types of Specifically, in the paper entitled Improved maximum
cables and signal injection methods used (i.e., with respect likelihood S-FSK receiver for PLC modem in AMR, Bali and
to adjacent lines/grounds and the grounding systems imple- Rebai present an optimized software implementation of a
mented in different countries) for such systems render the narrow band Power-line modem. The modem is a node in an
propagation difficult. Furthermore, Electromagnetic Com- automatic meter reading (AMR) system which is compliant
patibility (EMC) issues and more especially Electromagnetic to the IEC 61334-5-1 profile and operates in the CENELEC-
Interference (EMI) occurring at different frequencies of A band. Because of the hostile communication channel, a
operations still need more investigation. Power-line com- new design approach is carried out for an S-FSK demodulator
munications have demonstrated the acceptance to support capable of providing lower bit error rate (BER) than standard
various applications such as HD Television (HDTV), Internet specifications. The best compromise between efficiency and
Protocol Television (IPTV), interactive gaming, whole-home architecture complexity is investigated. Some implementa-
audio, security monitoring, and Smart Grid management. tion results are presented to show that a communication
These services have to be supported by the new home net- throughput of 9.6 kbps is reachable with the designed S-FSK
works/access and be deployed and standardized worldwide. modem.
This special issue was launched to gather state-of- In the paper entitled Radiation mitigation for Power-line
the-art research contributions in the field of Power-line communications using time reversal, Mescco et al. explain the
communications. It has attracted the submissions of many use of the Time Reversal (TR) technique to mitigate radi-
high-quality papers. After going through a conscientious ated emissions from PLC systems. The method was probed
2 Journal of Electrical and Computer Engineering
experimentally in real electrical networks with excellent Enhancements at the MAC layer are also discussed and they
results: in 40% of the observations, the electromagnetic include the implementation of power saving modes, the usage
interference generated by PLC transmission can be reduced of a short delimiter in the MAC frame, delayed acknowl-
by more than 3 dB, and this EMI mitigation factor may edgment, and the coexistence mechanisms with other PLC
increase to more than 10 dB in particular configurations. systems. The HomePlug AV2 performance is also assessed,
The paper proposes also future research directions in the through simulations reproducing real home scenarios.
field of wired TR techniques at higher frequencies and on We believe that the papers appearing in this special issue
other media, such as Digital Subscriber Line twisted pairs. In provide a good contribution and representation of significant
addition, optimal protocols are suggested for development of research topics in the field of PLC.
practical implementation of TR in future standards.
In the paper entitled A study on the optimal receiver Acknowledgments
impedance for SNR maximization in broadband PLC, Anto-
niali et al. present the design of the front-end receiver for We would like to thank the authors for their contributions
broadband Power-line communications. The paper focuses and the reviewers for their competent and timely work.
on the design of the input impedance that maximizes the
signal-to-noise ratio (SNR) at the receiver. The authors Justinian Anatory
show that the amplitude, rather than the power, of the Moises V. Ribeiro
received signal is important for communication purposes. Andrea M. Tonello
Furthermore, they analyse the receiver impedance impact on Ahmed Zeddam
the amplitude of the noise term for which a novel model
description is provided. Performance results are reported for
real in-home grids that have been assessed with experimental
measurements. The best attainable performance is studied
when the optimal receiver impedance is used and it is shown
that conventional power matching is suboptimal with respect
to the proposed impedance design approach in terms of
achievable information rate.
Two other papers deal with the topic of Smart Grid
in various environments. In the paper entitled Power-
line communications for Smart Grid applications, Berger
et al., surveys PLC technologies that are relevant in the
context of Smart Grid. The specifications of G3-PLC, PRIME,
HomePlug Green Phy, HomePlug AV2, as well as the
standards IEEE 1901/1901.2, and ITU-T G.hn/G.hnem are
discussed. Another paper entitled Enhancements of G3-PLC
technology for Smart-Home/Building applications, by Di Bert
et al., consider the in-home/building scenario, for which
a convergent network architecture is proposed to enhance
the performance of the narrow band G3-PLC technology
through its integration with an Ethernet-based network. The
paper defines the protocols implemented by the network
modules. Since Ethernet represents a convergent standard for
many communication devices, by adding this functionality to
G3-PLC, interconnectivity with other heterogeneous nodes
can be offered. Furthermore, since the G3-PLC medium
access control layer is based on a Carrier Sense Multiple
Access (CSMA) scheme, its performance decreases when
the number of network nodes contending for the channel
increases. Therefore, the paper evaluates the network per-
formance when an optimized time division multiple access
scheme is adopted showing that performance improvements
are reached.
In the paper entitled An overview of the HomePlug AV2
technology, Yonge et al. provide an overview of the HomePlug
AV2 system architecture and the key features at both the
PHY and MAC layers. The paper presents the new techniques
used at the PHY layer as multiple-input multiple-output
transmission, beamforming and precoding, adaptation of the
parameters, and power management for improved EMC.
http://dx.doi.org/10.1155/2013/712376
Review Article
Power Line Communications for Smart Grid Applications
Lars Torsten Berger,1 Andreas Schwager,2 and J. Joaqun Escudero-Garzs3

1
Kenus Informatica, Paterna, Valencia, Spain
2
European Technology Center (EuTEC), Sony Deutschland GmbH, Stuttgart, Germany
3
Department of Telecommunications and Systems Engineering, Universitat Autonoma of Barcelona (UAB),
Barcelona, Spain
Correspondence should be addressed to Lars Torsten Berger; lars.berger@kenus.es
Received 3 August 2012; Accepted 29 December 2012
Academic Editor: Ahmed Zeddam
Copyright 2013 Lars Torsten Berger et al. This is an open access article distributed under the Creative Commons Attribution
cited.
Power line communication, that is, using the electricity infrastructure for data transmission, is experiencing a renaissance in the
context of Smart Grid. Smart Grid objectives include the integration of intermittent renewable energy sources into the electricity
supply chain, securing reliable electricity delivery, and using the existing electrical infrastructure more efficiently. This paper surveys
power line communications (PLCs) in the context of Smart Grid. The specifications G3-PLC, PRIME, HomePlug Green PHY, and
HomePlug AV2, and the standards IEEE 1901/1901.2 and ITU-T G.hn/G.hnem are discussed.
1. Introduction deployments [3]. Industry adoption and large-scale customer

roll-outs are still in their infancies, and it is hard to make
Smart Grids, for many, the next big technological revolution an accurate prediction of the winners and losers. What
since the invention of the Internet, will play an important role seems clear is that power line communications (PLCs), that
in tomorrows societies. Governments around the world are is, communications over the existing electrical infrastructure,
pumping large sums of money into Smart Grid (SG) research, will have their part to play since they provide the natural
development, and deployments, their aims being manifold. upgrade from simple electricity conductors to hybrid and
Smart Grids have the potential to reduce carbon dioxide bidirectional electricity and data communication solutions.
emissions through the integration of distributed renewable The idea of using power lines also for communication
energy resources, energy storage, and plug-in hybrid electric purposes has already been around at the beginning of the
vehicles. Moreover, they can increase the reliability of the last century [4]. The obvious advantage is the wide spread
electricity supply (reduced blackout rate) by real-time mea- availability of electrical infrastructure, so that theoretically
surements, monitoring and control of the generation, and deployment costs are confined to connecting modems to the
transmission and distribution networks. Further, they can existing electrical grid.
render the utilization of base load power plants and electricity Power line technologies can be grouped into narrow-
transport infrastructure more efficient, deploying dynamic band PLC (NB-PLC), operating usually below 500 kHz, and
pricing and demand response strategies [1, 2]. broadband PLC (BB-PLC), operating usually at frequencies
Besides achievements in power electronics, sensing, above 1.8 MHz [5]. These are discussed in Sections 5 and 6,
monitoring, and control technology, key Smart Grid enablers respectively. Nevertheless, the following starts out with an
are the advances that in the last decade have been made in introduction to PLC scenarios, followed by channel, noise,
the area of telecommunications. There is a long list of com- and electromagnetic compatibility (EMC) aspects. Freely avail-
plementary and sometimes competing wireless and wireline able complementary reading on PLC state-of-the-art can also
specifications and standards that can be used in Smart Grid be found in [6]. Another valuable source of the PLC-related
Table 1: Domains and actors in the smart grid conceptual model, Medium-Voltage (MV) lines, with voltages in the range
based on [13, Table 3-1]. from 10 kV to 30 kV, are connected to the HV lines via
primary transformer substations. The MV lines are used
Domain Actors in the domain
for power distribution between cities, towns, and larger
The end users of electricity. May also generate, industrial customers. They can be realized as overhead or
store, and manage the use of energy.
underground lines. Further, they exhibit a low level of
Customers Traditionally, three customer types are
discussed, each with its own domain:
branches and directly connect to intelligent electronic devices
residential, commercial, and industrial. (IEDs) such as reclosers, sectionalizers, capacitor banks, and
The operators and participants in electricity
phasor measurement units. IED monitoring and control
Markets requires only relatively low data rates, and NB-PLC can
markets.
provide economically competitive communication solutions
The organizations providing services to
Service providers for these tasks. MV-related studies and trials can be found in
electrical customers and utilities.
[2123].
Operations The managers of the movement of electricity.
Low-voltage (LV) lines, with voltages in the range from
The generators of electricity in bulk quantities. 110 V to 400 V, are connected to the MV lines via secondary
Bulk generation
May also store energy for later distribution.
transformer substations. A communication signal on an MV
The carriers of bulk electricity over long line can pass through the secondary transformer onto the
Transmission distances. May also store and generate
LV line, however, with a heavy attenuation in the order of
electricity.
55 dB to 75 dB [24]. Hence, a special coupling device (induc-
The distributors of electricity to and from
tive, capacitive) or a PLC repeater is frequently required
Distribution customers. May also store and generate
electricity.
if one wants to establish a high data rate communications
path. As indicated in Figure 1, the LV lines lead directly or
over street cabinets to the end customers premises. Note
that a considerable regional topology difference exits. For
literature is the recently established IEEE Communication example in the USA a smaller secondary transformer on a
Societys web portal on Best Readings in Power Line Commu- utility pole might service a single house or a small number
nications [7]. of houses. In Europe, however, it is more common that
up to 100 households get served from a single secondary
2. Communication Scenarios transformer substation. Further, as pointed out in [25],
significant differences exist between building types. They
Many national and international organizations are currently may be categorized as multiflat buildings with riser, multiflat
drawing up roadmaps for SG standards [812]. For brevity, buildings with common meter room, single family houses, and
the following orients itself on the work by the US National high-rise buildings. Their different electrical wiring topologies
Institute of Standards and Technology (NIST). To structure influence signal attenuation as well as interference between
the various areas of the Smart Grid, NIST devised a domain- neighboring PLC networks [26].
based conceptual model [13]. Each domain contains actors In most cases the electrical grid enters the customers
that with the help of communications might act over domain premises over a house access point (HAP) followed by an
borders. The definitions of domains and actors are repro- electricity meter (M) and a distribution board (fuse box). One
duced in Table 1. The interconnections between domains frequently refers to PLC systems operating up to this point
are displayed in Figure 1. It has been common practice to as Access systems. Delivering broadband Internet Access over
distinguish power line communication scenarios according the electrical grid, also known as broadband over power
to operation voltages of the power lines [14]. Figure 1 links this line (BPL), amounted at the end of 2008 to less then 1%
voltage-based differentiation to the NIST conceptual model. of the worlds Access customers (65% used DSL, 21% used
High-voltage (HV) lines, with voltages in the range cable) [27]. BPL is, however, on the rise, especially in rural
from 110 kV to 380 kV, are used for nationwide or even areas and in developing countries with a poorly developed
international power transfer and consist of long overhead fixed telephone line and coaxial cable infrastructure [16].
lines with little or no branches. This makes them acceptable Apart from general Internet Access, automated meter reading
wave guides with less attenuation per line length as for (AMR) systems frequently used ultranarrowband power line
their medium- and low-voltage counterparts. However, their communication (UNB-PLC) technologies like Turtle [28]
potential for broadband SG communication services has up and TWACS [29, 30] to gain access and control over the
to the present day been limited. Time-varying high-voltage energy meters within private homes. UNB-PLC systems
arcing and corona noise with noise power fluctuations in the are usually designed to communicate over long distances
order of several tens of dBs and the practicalities and costs with their signals passing through the LV/MV transformers.
of coupling communication signals in and out of theses lines This helps to keep the amount of required modems and
have been an issue. Further, there is a fierce competition of repeaters to a minimum. Drawbacks are low data rates in
fiber optical links. In some cases, these links might even be the order of 0.001 bit/s and 60 bit/s for Turtle and TWACS,
spliced together with the ground conductor of the HV system respectively, and sometimes limitations to unidirectional
[15, 16]. Nevertheless, several successful trials using HV lines communications. These UNB-PLC technologies are men-
have been reported in [1720]. tioned here as they are among the pioneers in the AMR
Journal of Electrical and Computer Engineering 3
Markets
Bulk generation
Grid
Transmission, high voltage (HV)
operation
Primary
transformer
substations
Service Distribution, medium voltage (MV)

providers
M1 M2
Secondary
transformer
substations
Concentrator
Customer, low voltage (LV) M3
Room M6
Flat
. M5
.
. M4 Room
HAP M
ESI
EV
Flat
EVSE
M M M HAP
ESI ESIESI Single family house
with plug-in electric
Multiat building, common meter room vehicle (EV)
Communication link EVSE: electrical vehicle service equipment

Electricity and potential ESI: energy service interface
communication link M: meter
HAP: house access point
Figure 1: Smart Grid domains and their electrical and communication interconnections, based on Figure 7.1 in [75].
and distribution automation field. However, in the light of Load

many upcoming Smart Grid deployments, there are much
higher requirements on the communication infrastructure,
for example, to support demand response, distributed genera-
tion, and demand side management applications. It is believed Tx Rx
that these applications can, among others, be supported by
PLC-based Advanced Metering Infrastructure (AMI). A whole
wealth of material on AMI requirements and architectures Figure 2: Stub line example.
is available, for example, from the European OPEN meter
project [31]. To cope with increased AMI requirements, this
paper leaves UNB-PLC solutions at a sideline in favour of
Transmissions and reflections lead to a situation where a PLC
more recent narrowband PLC (NB-PLC) technologies, such
signal travels in form of a direct wave from over to .
as Power line-Related Intelligent Metering Evolution (PRIME)
Another signal travels from over to , bounces back
[32], and G3-PLC [33, 34]. NB-PLC bidirectional data rates
to , and reaches . All further signals travel from to
lie in the order of hundred kbit/s, while partly preserving
and undergo multiple bounces between and before they
the advantage to communicate over long ranges and through
finally reach . The result is a classical multipath situation,
transformers.
where frequency selective fading is caused by in-phase and
From the distribution board the LV lines run up to the
antiphase combinations of the arriving signal components.
different power sockets in every room. Lines may also run
The corresponding transfer function can readily be derived
to an electric vehicle service equipment (EVSE) as indicated in
in close form as an infinite impulse response filter [41].
Figure 1. For reliable home area network (HAN) high data rate
One important parameter capturing the frequency selectivity
applications, broadband power line communication (BB-PLC)
characteristics is the root mean square (rms) delay spread
technologies are becoming more and more attractive. Field-
(DS). For example, designing orthogonal frequency-division
proven BB-PLC technologies provide data rates of more than
multiplexing (OFDM) systems, the guard interval might be
200 Mbit/s [35], making it easy to fulfill the users home enter-
chosen as 2 to 3 times the rms DS to deliver good system
tainment needs including high definition television (HDTV).
performance [42]. To provide an orientation, the mean of the
Upcoming SG services in the home include granular control
observed rms DS for a band from 1 MHz up to 30 MHz in the
of smart appliances, the ability to remotely manage electrical
MV, LV-Access and LV-In-Home situations in [24, 42] was
devices, and the display of consumption data. Consumer
reported to be 1.9 s, 1.2 s, and 0.73 s, respectively.
awareness usually leads to a change in consumption habits
and in the sequel to energy savings between 10% and 20%
[36]. 3.2. Time Variation. Besides multipath fading, the PLC chan-
NB-PLC solutions have for a long time been used for nel exhibits time variation due to loads and/or line segments
home automation applications [14], and it is believed that the being connected or disconnected [43]. Further, through
well-established automation systems, like BACnet [37], KNX synchronizing channel measurements with the electrical grid
(ISO/IEC 14543-3-5, EN 50090) [38], and LON (ISO/IEC AC mains cycle Canete et al. were able to show that the In-
14908-3, ANSI 709.2) [39, 40], are being integrated into Home channel changes in a cyclostationary manner [4446].
upcoming Smart Home concepts [10]. Nevertheless the fol-
lowing also leaves these technologies at a sideline in favor of 3.3. Low-Pass Behavior. Untill now the low-pass behavior
more recent standards like IEEE 1901 and ITU-T G.hn for BB- of PLC channels has not been considered. It results from
PLC applications and IEEE 1901.2 and ITU-T G.hnem for NB- dielectric losses in the insulation between the conductors and
PLC applications. is more pronounced in long cable segments such as outdoor
underground cabling. Transfer function measurements on
different cable types and for different lengths can be found
3. Channel and Noise Aspects in [47, 48]. Using a large set of field trials, low-pass mean
gain models are derived in [24]. Over the range from 1
The power line channel and noise situations heavily depend to 30 MHz the mean gain in dB is approximated by linear
on the scenario and, hence, span a very large range. Generally, models. Consider again the PLC scenarios from Figure 1.
it can be said that the PLC channel exhibits frequency The mean gain from the secondary transformer to the HAP,
selective multipath fading and a low-pass behaviour. Further, labeled M3 and M4, writes [24]
alternating-current- (AC) related cyclic short-term variations
and abrupt long-term variations can be observed. LV-Access = (1 + 2 + 3 + 4 ) , (1)
where is frequency in MHz, is distance in meters, and the
3.1. Frequency Selectivity. To understand the effects that coefficients 1 to 4 are 0.0034 dB/(MHz m), 1.0893 dB/MHz,
lead to frequency selective fading consider, for example, 0.1295 dB/m, and 17.3481 dB, respectively.
the stub line schematic in Figure 2. An impedance-matched The mean gain model in dB for MV lines, as well as for
transmitter is placed at . marks the point of a branch, LV-In-Home situations is given by [24]
also called an electrical T-junction. An impedance-matched
receiver is placed at . A parallel load is connected at . MV or LV-In-Home = (1 + 2 ) . (2)
For the LV-In-Home situation the mean gain is given L T1

L
D1
from the mains distribution board to a socket in a room,
labelled M5 and M6 in Figure 1. The coefficients are 1 = D2 N
0.596 dB/MHz and 2 = 45.325 dB. The MV gain describes N
the channel between two primary transformers on the MV D3 PE T2
side, indicated by M1 and M2 in Figure 1. Its coefficients are PE
1 = 1.77 dB/MHz and 2 = 37.9 dB. In both situations the
model is not distant dependent. For the MV situation this Delta-style T-style
is due to the fact that not enough measurement results were (a) (b)
available to construct a distant-dependent model. Hence, in
L
this case the model is limited to situations where the distance
S1
between M1 and M2 is around 510 m. Nevertheless, correction N
S2
factors are proposed in [24] to determine the mean gain at PE
S3
other distances. For the LV-In-Home situation the model is
S4
not distance dependent either as distance in an In-Home
situation is a hard-to-define term. Power line networks in Star-style
such situation exhibit usually a large amount of branches, and (c)
a detailed floor plan to determine cable length cannot always Figure 3: MIMO PLC couplers.
be obtained. This leads to a situation where the low-pass
behavior is less pronounced in the In-Home case. Further,
in the MV- and the LV-Access situation the attenuation
drastically increases with frequency. This goes well in line consists of three baluns arranged in a triangle between L, N,
with the findings in [49] and is one of the reasons why BB- and PE. The T-style coupler feeds a differential mode signal
PLC Access networks are frequently operated in the lower between L and N plus a second signal between the middle
frequency range, for example, between 1 and 10 MHz, while point of L-N to PE. Receiving the PLC signals is also possible
BB-PLC In-Home networks might operate at frequencies using the star-style or longitudinal coupler. There three wires
above 10 MHz. are connected in a star topology to the center point. The
benefit of this coupler is the possibility to receive CM signals
which enables a forth reception path. On average CM signals
3.4. MIMO Channel. For a long time, the power line chan- are less attenuated than DM signals [64]. This is why it may
nel has been regarded as single-input single-output (SISO) be interesting to receive them, especially for highly attenuated
channel, based on two conductors. Nevertheless, in many In- channels.
Home installations three wires, namely, live (L) (also called
phase (P)), neutral (N), and protective earth (PE), are common
3.6. Noise Characterization and Modeling. Turning from the
[50]. Further, medium- and high-voltage installations often
channel to the noise situation, power line noises can be
make use of four or more conductors. In this respect, a
grouped based on temporal and spectral characteristics.
theoretical framework of multiconductor transmission line
Following, for example, [48, 66] one can distinguish colored
theory is extensively treated in [51]. Further, channel char-
background noise, narrowband noise, periodic impulsive noise
acterization and modeling work directly related to multicon-
asynchronous to the AC frequency, periodic impulsive noise
ductor PLC are available in [5263].
synchronous to the AC frequency, and aperiodic impulsive
noise. In [48] all these noises are modeled directly at the
3.5. MIMO Couplers. In general, the observed channel char- receiver using a superposition of spectrally filtered additive
acteristics are not independent from the coupling devices white Gaussian noise (AWGN), modulated sinusoidal signals,
used to inject and receive the power line signal. Figure 3 and Markov processes. Instead of modeling the noise directly
presents feeding and receiving possibilities for MIMO power at the receiver, Canete et al. proposed to model the noise at its
line communication, that is, (a) a Delta-style coupler, (b) a T- origin and to filter it by the channel transfer function [44, 67].
style coupler [55], and (c) a Star-style coupler [64]. Coupler Besides, specific results on MIMO system noise are presented
designs are tightly related to radiated emission treated in in [68, 69].
more detail in Section 4 [62]. According to the Biot-Savart A statistical approach to average colored background
law the main source of radiated emission is the common mode noise modeling is presented in [24] based on a large amount
(CM) current [65]. To avoid radiated emission, traditionally of noise measurements in MV, LV-Access and LV-In-Home
PLC modem manufacturers aim at injecting the signal as situations. Although a lot of the details get lost by averaging,
symmetrically as possible. In this way, two 180 out of the results can still deliver some interesting rule of thumb
phase electric fields are generated that neutralize each other when one wants to determine a likely average noise level.
resulting in little radiated emission. This desired symmetrical One general finding is that the mean noise power falls off
way of propagation is also known as differential mode (DM). exponentially with frequency. Derived from [24] the mean
Specifically, to avoid the injection of CM, feeding MIMO PLC noise power spectral density (PSD) in dBm/Hz is given by
signals can be done using the delta or T-style couplers from
Figures 3(a) or 3(b). The delta, also called transversal probe, = 1 (2 ) + 3 10 log10 (30000) , (3)
Table 2: Mean noise model coefficients [70] 2010 John Wiley & 55
Sons.
45
1 , 2 , 3 ,
Location
[dB] [1/MHz] [dBm/Hz] 35
M1 and 2, Secondary
Mean SNR (dB)

37 0,17 105
Transformer, MV 25
M3, Secondary Transformer, LV 24,613 0,105 116,72
M4, House access point, LV 29,282 0,12 114,94 15
M5, Main distribution board, LV 39,794 0,07 118,08
5
M6, Socket in private home, LV 17,327 0,074 115,172
5
where the last term normalizes out the 30 kHz bandwidth 15

used in the noise measurement process. The coefficients 0 5 10 15 20 25 30
1 to 3 are given in Table 2. The resulting noise models Frequency (MHz)
correspond to the measurement points M1 to M6 in Figure 1.
M1 to M2, subst. to MV subst.
M3 to M4, LV subst. to HAP
3.7. Mean SNR Considerations. Details on Tx limits will be M4 to M3, HPA to LV subst.
discussed in Section 4. For simplicity, assume for now that a M5 to M6, mains dist. to socket
M6 to M5, socket to main dist.
power line signal with = 50 dBm/Hz may be injected.
Using the gain and noise models from (1) to (3) the mean Figure 4: Mean SNR based on 50 dBm/Hz injected PSD, as well as
signal-to-noise ratio (SNR) can then be calculated as mean channel gain and mean noise power spectral density models.
SNR = + . (4)
The mean SNRs for the various connections between the one may distinguish between regulations for NB-PLC and
measurement points M1 to M6 in Figure 1 are plotted in BB-PLC.
Figure 4. One should note that, although the channel gain The NB-PLC regulations deal with the spectrum from
between two measurement points is symmetric, the noise at 3 kHz up to around 500 kHz. Important NB-PLC regulations
the measurement points differs. Hence, five different curves are listed in Table 3. Being a subset of all other bands,
are produced. the European Committee for Electrotechnical Standardization
It can be seen that especially the lower part of the (CENELEC) bands are the only ones available on a global
spectrum, up to 10 MHz, is very well suited for Access and basis. Four CENELEC bands are defined as (395 kHz),
Backhaul applications. Further, for In-Home applications the (95125 kHz), (125140 kHz), and (140148.5 kHz) [73].
entire spectrum from 1 to 30 MHz promises high mean SNRs Besides specifying transmission limits and their measure-
in the order of 40 dB, which goes also well in line with the ment procedures, the CENELEC standard also mandates that
findings in [71]. Further interesting results for the frequency the A band may only be used by Energy Suppliers and their
range up to 100 MHz are available in [6163, 69, 72]. licensees, while the other bands may be used by consumers.
In general, the results show that there is a high potential Further, devices operating in C band have to comply with a
for PLC if the estimated mean SNRs can be exploited in carrier sense medium access/collision avoidance (CSMA/CA)
PLC modems. However, the presented mean results have to protocol that allows a maximum channel holding period of
be handled with care. One should bear in mind that the 1 s, a minimum durations between channel uses by the same
mean SNR models from [24] exhibit a significant standard device of 125 ms, and a minimum time of 85 ms before declar-
deviation. Further, effects due to frequency selectivity, nar- ing the channel idle. For the USA, there are currently ongoing
rowband interference, impulsive noise, and time variation are efforts [74] to specify the band from 9 kHz to 534 kHz for
not reflected in Figure 4. Whether the estimated mean SNRs NB-PLC operations with a mandatory CSMA/CA protocol
translate into high PLC data rates depends not at last on the compliant with CENELEC EN 50065-1 [73]. The advantages
PLC modems signal processing algorithms, its component are that equipment manufacturers would be easily able to
quality, and permissible implementation complexity. adapt their NB-PLC products to the EU and USA market and
to many other markets that follow these standards.
4. Electromagnetic Compatibility Regulations Turning the view to BB-PLC one may again distinguish
two frequency ranges, that is, 1 MHz to 30 MHz, where
Power line cables were not designed to carry communication conducted emission is at the focus of regulation, and 30 MHz
signals and, hence, give rise to conducted emission, as well to 100 MHz, where the focus shifts to radiated emission.
as radiated emission that may interfere, for example, with The Comite International Special des Perturbations
Amateur Radio or radio broadcast receivers. When looking at Radioelectriques (CISPR), founded in 1934 and now part of
power line electromagnetic compatibility (EMC) regulations, the International Electrotechnical Commission (IEC), was
Table 3: Important regulations related to NB-PLC [75] 2012 John to the telecommunication port or to the mains port by
Wiley & Sons. introducing a Power Line Telecommunication (PLT) port.
In this document the electro magnetic interference (EMI)
Frequ.
States Institution References mitigation techniques are specified as normative. However,
[kHz]
CISPR never proposed a committee draft for voting (CDV).
Comite Europeen de
This is why CENELEC became active after the lifetime of the
EU 3148.5 Normalisation [73]
Electrotechnique (CENELEC)
CISPR committee to find at least in Europe an acceptable
solution. The result, prEN 50561-1:2012 [85] additionally
Federal Communications
US 10490 [76] excludes the aeronautical frequencies from power line
Commission (FCC)
communication and specifies the following.
Association of Radio Industries
Japan 10450 [77]
and Businesses (ARIB) (i) An emission measurement procedure at the PLT port
while no communication takes place.
(ii) A second emission measurement procedure at the
making efforts to regulate BB-PLC generated interference. PLT port when normal PLC takes place.
At the beginning CISPR 22 [78] defined two sets of limits (iii) A general cap on the injected PSD of 55 dBm/Hz.
and measurement methods for conducted emissions of
telecommunication equipment. One set was defined for the (iv) Permanent notching of certain parts of the radio spec-
telecommunication port, and the other was defined for the trum, that is related to amateur radio and aeronautical
mains port. For PLC modems, it is not defined whether the bands.
PLC signal port, which at the same time is used for power (v) A procedure for adaptive notching, meaning that the
supply, is considered as a mains or as telecommunication PLC equipment senses the presence of radio services
port. The method for telecommunication ports respects the and notches the affected frequencies for its own
symmetry properties of the attached cable. Asymmetries like operation.
an open light switch or asymmetrical parasitic capacitances (vi) A procedure of adaptive transmit power manage-
convert the symmetrically fed signals into common ment, meaning that the transmitting equipment limits
mode signals [65]. However, the method used for devices its transmit power as a function of channel attenua-
connected to the mains, specified in CISPR 16 [79], is based tion and noise to a level below the allowed maximum,
on measuring the asymmetric voltage level of either the that is just sufficient to achieve the required data rate.
phase or neutral wire to the ground. From the perspective
of a PLC modem, this is the worst case, because this voltage EN50561-1:2012 was approved in November 2012, which
consists of asymmetric and symmetric voltage; that is, not finally gives certainty to PLC stakeholders on interference
only the emission, and therewith the interference relevant limits.
part, but also the desired signal is measured and compared In the USA the Federal Communications Commission
against emission limits. IEC CISPR/I/89/CD [80] tried to (FCC) is in charge of regulating electromagnetic emissions.
clarify the situation by interpreting PLC as an application In general, all digital equipments have to comply with the
following the telecommunication limits of CIPSR 22. FCC part 15 standard (47 CFR 15) [76]. Specifically, Access
Therefore, the longitudinal conversion loss (LCL) parameter PLC systems over medium- and low-voltage power lines
was used in an identical way as, for example, in the testing and for a frequency range from 1.705 to 80 MHz are treated
of digital subscriber line (DSL) equipment. The benefit of in the standards Section G. Conducted emission limits are
using the LCL parameter is the simplicity of measuring it. explicitly not applicable, but radiated emission limits are
It is a reflection parameter whose measurement requires imposed through a transmit power spectral density mask.
an equipment to be connected to only one location of the Additionally, PLC systems have to be able to notch certain
grid. Other possibilities to verify the potential of interference frequencies that might be used by other services. Further,
would have been to place antennas and measure the field the FCC defines excluded bands where no PLC signal shall
generated by the fed PLC signal. However, this would make be injected, as well as geographical exclusion zones close to
a measurement setup significantly more cumbersome and which no Access PLC systems may be deployed. Further, pro-
error prone. Nevertheless, with respect to highly attenuated cedures in which service providers inform about prospective
wires like the power lines the much simpler measurement PLC Access deployments and complaint handling procedures
of a reflection parameter as in [80] only describes the are a requirement.
local situation instead of giving detailed insight of what is More details on PLC EMC regulations as well as con-
happening if a PLC signal travels deeply into the electrical ducted and radiated interference measurement results can be
network. In 2008, CISPR/I/257/CD [81] was published found in [6, 61, 86]. Besides, the IEEE Standard for Power
with an LCL parameter reduced by 6 dB. Simultaneously, Line Communication EquipmentElectromagnetic Compati-
CISPR/I/258/CD [82] indicated that mitigation techniques bility (EMC) RequirementsTesting and Measurement Meth-
like cognitive notching for PLC modems [64, 83] and ods [87] was recently released, intending to provide an
dynamic transmit power management are the compromise internationally recognized EMC measurement and testing
to solve the never-ending EMC discussions. CIS/I/301/CD methodology. It endorses among others CISPER22 and FCC
[84] answered the question of whether PLC is connected part 15 as normative references, but does not establish
any emission limits itself. Looking at the developments in responsible for PLC channel access arbitration. A contention
CISPR/CENELEC and at FCC part 15, it becomes clear that free and a contention-based access mechanism exists, whose
next generation PLC equipment has to be highly configurable usage time and duration are decided by the base node. Within
to apply power spectral density-shaping masks, as well as the contention-free time division multiplex (TDM) channel
adaptive notching. access period, the base node assigns the channel to only one
node at a time. The contention-based access uses CSMA/CA
5. Narrowband PLC [88, 92]. To assure privacy, authentication, and data integrity a
Security Profile 1 is defined, that uses 128-bit AES encryption
Narrowband power line communication systems usually [93]. The specification also defines a Security Profile 0 equiva-
operate in the frequency range from 3 kHz to 500 kHz, that is, lent to no encryption and leaves room for the definition of two
the CENELEC/ARIB/FCC bands. Following, for example, the further security profiles in future releases. To interface MAC
nomenclature in [5], they can be subdivided into low data rate and application layer, PRIME defines a convergence layer
(LDR) and high data rate (HDR) systems. LDR systems have (CL) between the two. The CL can be split into a Common
throughputs of a few kbit/s and usually are based on single Part Convergence Sublayer (CPCS) and a Service Specific
carrier technology. Example standards, listed in the NIST Convergence Sublayer (SSCS). The CPCS performs tasks of
Smart Grid Interoperability Panel (SGIP) Catalog of Standards data segmentation and reassembling and is adjusted to the
[3, 13], are ISO/IEC 14908-3 (LON, ANSI/EIA 709.2) [39, specific application. Three SSCSs are currently defined: the
40] and ISO/IEC 14543-3-5 (KNX, EN 50090) [38]. These NULL Convergence Sublayer provides the MAC layer with a
standards span all layers of the open systems interconnection transparent way to the application, being as simple as possible
(OSI) model and can, besides over power line, also be used and minimizing the overhead. It is intended for applications
over other media such as twisted pair and in some cases even that do not need any special convergence capability. The
wirelessly. Their main area of application has been industrial IPv4 Convergence Layer provides an efficient method for
and building automation. In this respect a further popular transferring IPv4 packets over the PRIME network. Finally,
protocol is BACnet (ISO 16484-5 [37]). The following, how- the IEC 61334-4-32 Convergence Layer supports the same
ever, will focus on the physical layer (PHY) of high data rate primitives as the IEC 61334-4-32 standard [94], making it
(up to 1 Mbit/s) NB-PLC. As in many other communication easy, for example, to support advanced metering applications
systems OFDM has emerged as the modulation scheme of that make use of the standardised data models of IEC 62056-
choice for HDR NB-PLC. Example HDR NB-PLC systems 62 [95]. PRIME could therefore also be used to replace the
are G3-PLC [33] and PRIME [88] that have made it into ITU aging PHY and MAC layer of the single carrier power line
G.hnem standardization as ITU G.9903 [89] and ITU G.9904 standard IEC 61334-5-1 [96], also known as S-FSK, for spread
[90], respectively. Both are also supported via interoperability frequency shift keying.
modes in IEEE 1901.2. With respect to coexistence, the
NIST priority action plan 15 (PAP15) working group recently 5.2. G3-PLC. The other OFDM-based HDR NB-PLC speci-
approved that newly developed NB-PLC standards shall all fication, G3-PLC [9799] was published in August 2009. It
implement a single coexistence protocol as to have minimal can be configured to operate in the internationally accepted
performance impact on the existing deployed devices, that is, bands from 10 kHz to 490 kHz (FCC, CENELEC, and ARIB).
devices using ISO/IEC 14543-3-5, IEC 61334-3-1, IEC 61334- Using differential BPSK, QPSK, and 8PSK for constellation
5-1, IEC 61334-5-2, and IEC 61334-4-32 [91]. mapping and concatenated convolutional Reed-Solomon for-
ward error correction coding, it reaches PHY peak data rates
5.1. PRIME. PRIME, for Power line-Related Intelligent- close to 300 kbit/s. Peak and typical data rates for various
Metering Evolution, was developed within the PRIME Alli- frequency bands have been reported in [100] and are, for
ance, with its steering committee chaired by the Spanish convenience, also listed in Table 4. The MAC layer is based
utility heavyweight Iberdrola [32]. The PRIME system uses on IEEE 802.15.4-2006 [101]. 6LoWPAN [102] is used to adapt
a total of 96 ODFM subcarriers over the frequencies from the IEEE 802.15.4-2006 MAC to IPv6 [103]. This allows the
42 kHz to 89 kHz, that is, within the CENELEC A-band. application layer to comply with ANSI C12.19/C12.22 [104] or
Further, it deploys differential binary, quaternary, and eight- IEC 62056-61/62 (DLMS/COSEM) [105, 106] to run standard
phase shift keying (BPSK, QPSK, and 8PSK) and an optional Internet services.
1/2-rate convolutional code. Therewith it is able to achieve a A comparison between PRIME and G3-PLC, mainly
PHY peak data rate of 128.6 kbit/s [92]. The OFDM symbol focusing on physical layer aspects is presented in [107].
interval is of 2240 s including a 192 s cyclic prefix which There it is found that G3-PLC is slightly more robust when
suffices to deal with most common power line delay spreads. disturbed by AWGN and narrowband interference, while
Further, to deal with unpredictable impulsive noise PRIME PRIME is the less complex system, which could allow cheaper
offers the option to implement automatic retransmission implementations.
request (ARQ), based on the selective repeat mechanism.
Turning to the system architecture, PRIME is forming sub- 5.3. IEEE 1901.2 and ITU-T G.hnem. The specifications
networks, where each subnetwork has one Base Node and PRIME and G3-PLC form the baseline in the ongoing NB-
several Service Nodes. The base node is the master that PLC standardization processes within IEEE 1901.2 [108].
manages the subnetworks resources and connections using Compliant devices are supposed to support interoperability
a periodically sent beacon signal. The base node is further modes with PRIME and G3-PLC in the CENELEC A band
Table 4: G3-PLC data rates, based on [100]. Forum was founded to promote the ITU-T G.hn standard
and to address certification and interoperability issues [123].
Peak rate, Typical rate,
Frequency band Simultaneously, IEEE P1901 [124] was working on the Draft
[kbit/s] [kbit/s]
Standard for Broadband over Power Line Networks: Medium
CENELEC (36 kHz to 91 kHz) 46 44 Access Control and Physical Layer Specifications [125]. It
FCC (150 kHz to 487.5 kHz) 234 187 covers the aspects Access, In-Home, and coexistence of
FCC (10 kHz to 487.5 kHz) 298 225 Access-In-Home and In-Home-In-Home networks, and the
official IEEE Std 19012010 was published on December, 30,
2010, with the HomePlug Powerline Alliance [117] being a
certifying body for IEEE 1901 compliant products. In analogy
and have as minimum a requirement to implement one of
to the introduction of MIMO to ITU G.hn, the HomePlug
the PHY/MACs referred to as Main 1901.2, G3 CENELEC A,
Alliance introduced the HomePlug AV2 specification in Jan-
and PRIME CENELEC A [109]. G3-PLC and PRIME have also
uary 2012. The HomePlug AV2 specification includes features
been approved as ITU recommendations G.9903 [89] and
like MIMO with beamforming, an extended frequency range
G.9904 [90], respectively in October 2012. Together with ITU
of up to 86 MHz, efficient notching, several transmit power
G.9901 [110] and ITU G.9902 [111] they now supersede what
optimization techniques, 4096-QAM, power save modes,
was formally ITU G.9955/G9956. ITU G.hnem development
short delimiter, and delayed acknowledgement, boosting
is oriented on Smart Grid use cases, like support of pricing
the maximum PHY rate to around 2 Gbit/s. Further, to
awareness, load control, and demand response, and it has been
cover multiple home networking media under one umbrella,
agreed to meet the requirements set forth by NIST PAP15.
IEEE P1905.1 is working on a standard for a convergent
There are still some open issues, for example, with respect
digital home network for heterogeneous technologies [126].
to ITU G.hnem IEEE 1901.2 coexistence. Implementers that
It defines an abstraction layer for multiple home networking
cannot wait might consider that due to relatively low signal
technologies like IEEE 1901, IEEE 802.11 (Wi-Fi), IEEE 802.3
processing complexity, NB-PLC modems can be imple-
(Ethernet), and MoCA 1.1 and is extendable to work with
mented on a digital signal processor (DSP). This allows for
other home networking technologies.
upgradeability via software updates and can especially be
an advantage for early stage customer premises deployments
6.1. IEEE 1901 and ITU-T G.hn. IEEE 1901 uses the band
where a hardware update would be prohibitively expensive.
from 2 MHz up to 50 MHz with services above 30 MHz being
An upgradeable DSP solution that supports PRIME as well as
optional. ITU-T G.hn (G.9960/G.9961) operates from 2 MHz
G3-PLC is, for example, offered by Texas Instruments [112].
up to 100 MHz using bandwidth scalability, with three dis-
tinct and interoperable bands defined as 225, 250, and 2
6. Broadband PLC 100 MHz. The architectures defined by IEEE 1901 and ITU-T
G.hn (G.9960/G.9961) are similar in several aspects. In G.hn
In the last decade, BB-PLC chips by semiconductor vendors, one refers to a subnetwork as Domain. Operation, as well
such as Intellon (in 2009 acquired by Atheros, Atheros in as communication is organized by the Domain Master who
2011 acquired by Qualcomm) [113], DS2 (in 2010 acquired communicates with various Nodes. Similarly, the subnetwork
by Marvell) [114], Gigle (in 2010 acquired by Broadcom) in 1901 is referred to as Basic Service Set (BSS). The equivalent
[115], and Panasonic [116], came to market that operate to the domain master is the BSS Manager, which connects
in the band from around 1 MHz to 300 MHz. The chips to so-called Stations. These basic network components with
are mainly based on three consortia backed specifications their system specific terminology are summarized in Table 5.
developed within the frameworks of the HomePlug Pow- While the general concepts are similar, one should note
erline Alliance (HomePlug) [117], the Universal Powerline that G.hn defines a PHY/DLL used for operation over
Association (UPA) [118], and the High Definition Power Line any wireline medium. Primarily, the OFDM parameters are
Communication (HD-PLC) Alliance [119]. Related products adjusted to account for different medium-dependent channel
allow data rates around 200 Mbit/s and are not interoper- and noise characteristics. On the contrary, IEEE 1901 defines
able. However, to make PLC systems a broad success, an two disparate PHY/MAC technologies based on HomePlug
internationally adopted BP-PLC standard became essential. AV and HD-PLC. One of the key differences is their fre-
The International Telecommunications Union (ITU) as well quency division-multiplexing scheme. The HomePlug AV-
as the Institute of Electrical and Electronics Engineers (IEEE) based version uses the Fast Fourier Transform (FFT), while
commenced work on such next generation standards, namely, the HD-PLC based version uses Wavelets. Hence, they are
ITU-T G.hn and IEEE 1901. ITU-T G.hn is not only applicable sometimes also referred to as FFT-PHY and Wavelet-PHY,
to power lines, but also to phone lines and coaxial cables, respectively. A special coexistence mechanism has to be used
therewith for the first time defining a single standard for all when operating IEEE 1901 devices from both PHYs on the
major wireline communications media. At the end of 2008, same power line which is standardized within IEEE 1901 as
the PHY layer and the overall architecture were consented Inter-System Protocol (ISP) (see also [127] or [128]). A nearly
in ITU-T Recommendation G.9960 [120]. The Data Link identical mechanism was standardized by ITU-T in G.9972
Layer (DLL) Recommendation G.9961 [121] was approved in [129], also known as G.cx. Technical contributions to ITU,
June 2010, and a MIMO transceiver extension G.9963 was and IEEE from members of the NIST PAP15 assured the
consented in September 2011 [122]. Alongside, the HomeGrid alignment of both standards. As a result, it is likely that the
Table 5: Synopsis of terms used in the BB-PLC standards ITU-T operating in the frequency range 225 MHz. This allows for
G.hn and IEEE 1901 [75] 2012 John Wiley & Sons. reduced component cost and power consumption. Example
SGH nodes could be heating and air conditioning appliances,
Network item ITU-T G.hn IEEE P1901
Plug-in Electric Vehicles (PEV), and Electric Vehicle Service
Basic Service Set Equipments (EVSEs) as indicated in Figures 1 and 5. Together
Sub-network Domain
(BSS) they form a multi-domain HAN.
Transceiver Node Station (STA) The SGHs interact with the utilitys access network (UAN)
Sub-network controller Domain Master BSS Manager and its advanced metering infrastructure (AMI) through an
Media Access Plan energy service interface (ESI). The AMI domain comprises
Access control schedule Beacon
(MAP) AMI Meters (AM), AMI Submeters (ASM), and an AMI Head
Timeframe MAC cycle Beacon Interval End (HE). The HE is a local hub (concentrator), that controls
CSMA/CA, TDMA, all meters downstream from it and interfaces to the utilitys
STXOP CSMA/CA, wide area/backhaul network upstream from it. Each AMI
Access methods HE supports up to 250 AM and/or ASM nodes forming
(shared transmission TDMA
opportunities) an AMI domain (in dense urban areas 150 to 200 meters
Relaying Relay (L2) Repeater (L2) are a frequently encountered maximum). Further, a network
Network controller Relay (assigned Proxy BSS supports up to 16 AMI domains, delivering support for up
to 16 250 = 4000 AMI devices. The ability to support
proxy as a proxy) Manager
16 domains with 250 nodes each is a general property of
Transceiver Hidden Hidden G.hn not limited to Smart Grid/AMI applications. Domains
(not directly reached) Node Station may be formed over any kind of wiring. The nodes within
a domain are grouped into SGH and non-SG nodes. For
security reasons, non-SG nodes are logically separated from
SGH nodes using a secure upper-layer protocol.
In every domain there is a domain master that coor-
NIST SGIP will mandate that all BB-PLC technologies imple- dinates operation of all nodes. G.hn nodes of different
ment Recommendation ITU-T G.9972 or ISP [130]. ISP/G.cx domains communicate with each other via Interdomain
provides coexistence by splitting time equally among systems Bridges (IDBs). IDBs are simple data communications bridges
present on the line. IEEE 1901 additionally specifies the on OSI Layer 3 and above, enabling a node in one domain to
optional Coexistence Protocol (CXP), that provides coexis- pass data to a node in another domain. In a multi-domain
tence among multiple technologies based on a first-come- situation, a Global Master (GM) provides coordination of
first-serve basis. resources, priorities, and operational characteristics between
The coexistence mechanism is thought for the case where G.hn domains. Besides, G.hn domains can be bridged to
disparate networks would otherwise be interfering with alien (non-G.hn) domains, for example, to IEEE 1901/1901.2,
each other. Another likely scenario is that same technology wireless technologies, and so forth. For example, besides the
networks exist in close proximity, with the risk of so-called UAN/AMI connection through the ESI, the HAN might be
neighboring network interference. To deal with neighboring connected to the outside world via a DSL or cable modem
network interference G.hn uses different preamble-symbol gateway communicating with the G.hn HAN via an alien
seeds in each network. Therewith, G.hn networks are able to domain bridge.
coexist and communicate simultaneously, that is, not using
time division. Instead, link adaptation procedures adjust the
throughput to cope with degraded signal to interference plus 6.3. HomePlug Green PHY. In analogy to the ITU G.hn low
noise ratios (SINR). In many cases the throughput will be complexity profile (LCP), the HomePlug Powerline Alliance
throttled only slightly allowing G.hn networks to coexist has released the HomePlug Green PHY specification. Home-
nearly unimpeded. On the other hand, IEEE 1901 relies on Plug GreenPhy is a subset of HomePlug AV that is intended
a CSMA/CA medium access strategy, which may lead to an for use within Smart Grid applications. It is compatible to
increased number of collisions. As countermeasure, IEEE HomePlug AV, AV2, and IEEE 1901 and is optimized for
1901 introduces a coordinated mode that allows neighboring low power applications and costs. GreenPhy uses the most
networks to allocate times over the shared medium for robust communication mode (called ROBO) of HomePlug
specific communications. This coordinated time division AV technology. OFDM carrier spacing, signal preamble,
multiple access (TDMA) mode enables traffic to get through frame control, and FEC are identical to HomePlug AV to
unimpeded albeit at the price of time division (orthogonal ensure interoperability. This also results in identical coverage
throughput sharing). and reliability to Smart Grid functions. CSMA/CA is used
as channel access scheme. Data rates are up to 10 Mbit/s.
GreenPhy nodes may use long power save periods if a higher
6.2. ITU-T G.hn Low Complexity Profile. It is envisioned latency is acceptable. In the sleep state modems have only a
that G.hn nodes are in the future embedded into Smart 3% power consumption compared to the awake time resulting
Grid home (SGH) area network devices. SGH nodes will in an average power reduction of more than 90% with respect
typically make use of the G.hn low complexity profile (LCP), to standard HomePlug AV products.
Domain A
Bridge to alien
domain
Domain B
Non-SG Non-SG
G.9960 G.9960 Non-SG Non-SG Bridge between Non-SG
node node G.9960 G.9960 domains A and B G.9960
PC DMA node node node
Non-SG G. SGH SGH SGH Non-SG

9960 node G.9960 EVSE G.9960 G.9960 G.9960
multimedia node node node node
EMS DMA
EV
G.9960
node (ex.)
EV
G.9960
node (ex.) EVCF Multidomain HAN
SGH AMI domain

G.9960
node
Bridge Residence
G.9960 boundary
AM
node
G.9960 G.9960
AM AM
node node
Meter
Utility access network
(AMI and Smart Grid)
Figure 5: Smart Grid HAN implementation based on G.9960, based on Figure 22 in [138].
7. Application Layer Interoperability 8. Conclusions

Although there is still an ongoing struggle between stake- Although there are strong wireless and wireline communi-
holders promoting the various NB-PLC and BB-PLC tech- cation competitors, it is believed that power line communi-
nologies, it seems clear that IP support will become cations (PLCs) will fulfill various communication tasks in
paramount to assure Smart Grid interoperability. Along these upcoming Smart Grid deployments as PLC provides the natu-
lines, the HomePlug Alliance (promoting IEEE P1901) and ral upgrade from simple electricity conductors to hybrid and
the HomeGrid Forum (promoting ITU G.hn/G.hnem) have bidirectional electricity and data communication solutions.
joint forces with wireless stakeholders, namely, the Wi-Fi Seen from a utility point of view, one of the main advan-
Alliance and the ZigBee Alliance. Together they formed the tages of PLC is the full control over the physical medium,
Consortium for Smart Energy Profile 2 (SEP2) Interoperability without the need to depend on third party providers like
with the intention to develop common testing documents telecommunication companies or cellular operators. Espe-
and processes for certifying SEP 2 interoperability [131]. cially, PLC standardization and harmonization as, for exam-
Smart Energy Profile 2 (SEP2) originates from smart ple, fostered by NIST PAP15, are important for the PLC indus-
energy-related upper-layer developments (OSI layer 5 to 7) try as a whole when defending territory against competing
within the ZigBee Alliance [132]. It is compatible to the wireless and wireline options.
International Electrotechnical Commissions Common Infor- In terms of broadband PLC (BB-PLC), the coexisting
mation Model (CIM) [133, 134] and is kept link layer agnostic. standards ITU-T G.hn and IEEE P1901 with the expanded
Further, it is described using extensible markup language HomePlug AV2 specification are currently the most promis-
(XML) [135] and follows a Representational State Transfer ing, while narrowband PLC (NB-PLC) standardization, that
(REST) [136] architecture. Moreover, data transport takes is, IEEE P1901.2 and ITU-T G.hnem, is still ongoing.
place using Hypertext Transfer Protocol (HTTP) [137]. Cases where NB-PLC and BB-PLC systems operate on
By opting for these widely adopted building blocks, SEP 2 the same physical medium will become more frequent with
can benefit from a large knowledge and developer base and increasing PLC system penetration. Nevertheless, coexis-
is regarded by NIST as an import specification to allow tence of the two should be straight forward as they are
interoperability between various home area network devices using different frequency ranges. In general, interconnections
[13]. between different coexisting NB-PLC and BB-PLC standards
can then be established by OSI Layer 3 bridges as suggested, [9] ITU Telecommunication Standardization Bureau Policy &
for example, in [108, 138]. Technology Watch Division, Activities in smart grid standard-
When looking at higher layer interoperability, the recent ization repository, version 1.0., April 2010, http://www.itu.int/
foundation of the Consortium for Smart Energy Profile dmspub/itu-t/oth/48/01/T48010000020002PDFE.pdf.
2 Interoperability, is a promising step forward to allow [10] DKE German Commission for Electrical, Electronic &
users and operators to deploy various wireless and wireline Information Technologies of DIN and VDE, The German
technologies with seamless interoperability at the application roadmap e-energy/smart grid, April 2010, http://www.e-energy
.de/documents/DKE%20Roadmap%20SmartGrid%20230410
layer which is regarded as essential for widespread Smart Grid
%20Engllish.pdf.
acceptance.
[11] State Grid Corporation of China, SGCC framework and
roadmap for strong and smart grid standards, August 2010,
Acknowledgments http://collaborate.nist.gov/twiki-sggrid/pub/SmartGrid/SGIP
DocumentsAndReferencesSGAC/China State Grid Frame
Although this article is distributed under the Creative Com- work and Roadmap for SG Standards.pdf.
mons Attribution License, it should be noted that it is in [12] International Energy Agency, Technology roadmap smart
parts based on [70, 75]. The material from [70] and [75] is grids, April 2011, http://www.iea.org/publications/freepub
reproduced with explicit permission of John Wiley & Sons, lications/publication/smartgrids roadmap.pdf.
Ltd. and John Wiley & Sons, Inc., respectively. However, [13] National Institute of Standards and Technology (NIST) and U.S.
Department of Commerce, NIST framework and roadmap for
related copyrights and the right to grant reprint permissions
smart grid interoperability standards, NIST Draft Publication,
of the affected parts remain with John Wiley & Sons. This Release 1. 0, January 2010.
work is cofinanced by the Spanish Ministry of Science
[14] K. Dostert, Powerline Communications, Prentice-Hall, 2001.
and Innovation (MICINN) Program INNCORPORA-Torres
[15] G. Held, Understanding Broadband Over Power Line, CRC Press,
Quevedo 2011, and Grant SICCNALS, TEC2011-28219. Fur-
2006.
ther it is cofunded by the European Social Fund and the
[16] P. Sobotka, R. Taylor, and K. Iniewski, Broadband over power
European Regional Development Fund within the ERDF line communications: Home networking, broadband access,
Operational Programme of the Valencian Community 2007 and smart power grids, in Internet Networks: Wired, Wireless,
2013, grant no. IMIDTA/2011/905. Special thanks also to the and Optical Technologies, K. Iniewski, Ed., Devices, Circuits,
BreezeSolve Team for editorial support. and Systems, chapter 8, CRC Press, New York, NY, USA, 2009.
[17] R. Pighi and R. Raheli, On multicarrier signal transmission for
high-voltage power lines, in Proceedings of the 9th International
References Symposium on Power Line Communications and Its Applications
(ISPLC 05), pp. 3236, Vancouver, Canada, April 2005.
[1] B. A. Hamilton, J. Miller, and B. Renz, Understanding the
benefits of smart grid, Tech. Rep. DOE/NETL-2010/1413, U.S. [18] H. Duckhwa and L. Younghun, A study on the compound
Department of Energy, 2010. communication network over the high voltage power line for
distribution automation system, in Proceedings of the 2nd
[2] L. T. Berger and K. Iniewski, Smart GridApplicacions, Com- International Conference on Information Security and Assurance
munications and Security, John Wiley & Sons, 2012. (ISA 08), pp. 410414, Busan, Korea, April 2008.
[3] M. Burns, NIST SGIP catalog of standards, October 2010, [19] R. Aquilu, I. G. J. Pijoan, and G. Sanchez, Highvoltage multi-
http://collaborate.nist.gov/twiki-sggrid/bin/view/SmartGrid carrier spread-spectrum system field test, IEEE Transactions on
SGIPCatalogOfStandards. Power Delelivery, vol. 24, no. 3, pp. 11121121, 2009.
[20] N. Strandberg and N. Sadan, HV-BPL phase 2 field test report,
[4] P. A. Brown, Power line communicationspast present and
Tech. Rep. DOE/NETL-2009/1388, U.S. Department of Energy,
future, in Proceedings of the International Symposium on Power
2009.
Line Communicatons and Its Applications (ISPLC 99), pp. 18,
September 1999. [21] P. A. A. F. Wouters, P. C. J. M. van der Wielen, J. Veen, P. Wage-
naars, and E. F. Wagenaars, Effect of cable load impedance on
[5] S. Galli, A. Scaglione, and Z. Wang, For the grid and through coupling schemes for MV power line communication, IEEE
the grid: the role of power line communications in the smart Transactions on Power Delivery, vol. 20, no. 2, pp. 638645, 2005.
grid, Proceedings of the IEEE, vol. 99, no. 6, pp. 9981027, 2011. [22] R. Benato and R. Caldon, Application of PLC for the control
[6] The OPEN meter Consortium, Description of current and the protection of future distribution networks, in Pro-
state-of-the-art of technology and protocols description ceedings of the IEEE International Symposium on Power Line
of state-of-the-art of PLC-based access technology, Communications and Its Applications (ISPLC 07), pp. 499504,
European Union Project Deliverable FP7-ICT-2226369, d 2.1 Pisa, Italy, March 2007.
Part 2, Version 2.3, March 2009, http://www.openmeter.com/ [23] A. Cataliotti, A. Daidone, and G. Tine, Power line communica-
files/deliverables/OPEN-Meter%20WP2%20D2.1%20part2% tion in medium voltage systems: characterization of MV cables,
20v2.3.pdf. IEEE Transactions on Power Delivery, vol. 23, no. 4, pp. 1896
1902, 2008.
[7] IEEE Communication Society, Best readings in power line
[24] P. Meier, M. Bittner, H. Widmer et al., Pathloss as a function of
communications, http://www.comsoc.org/best-readings.
frequency, distance and network topology for various LV and
[8] SMB Smart Grid Strategic Group (SG3), IEC smart grid stan- MV European powerline networks, The OPERA Consortium,
dardization roadmap, June 2010, http://www.iec.ch/smartgrid/ Project Deliverable, EC/IST FP6 Project no. 507667 D5v0.9,
downloads/sg3 roadmap.pdf. April 2005.
[25] A. Rubinstein, F. Rachidi, M. Rubinstein et al., EMC guide- [42] S. Galli, A simplified model for the indoor power line channel,
lines, The OPERA Consortium, IST Integrated Project Deliver- in Proceedings of the IEEE International Symposium on Power
able D9v1. 1, IST Integrated Project No 026920, October 2008. Line Communications and its Applications (ISPLC 09), pp. 13
[26] A. Vukicevic, Electromagnetic compatibility of power line com- 19, Dresden, Germany, April 2009.

munication systems [Ph.D. thesis], cole Polytechnique Federale [43] F. J. Canete Corripio, L. Dez del Ro, and J. T. Entrambasaguas
de Lausanne, Lausanne, Switzerland, 2008. Munoz, A time variant model for indoor power-line chan-
[27] Broadband Forum, Next generation broadband access, nels, in Proceedings of the International Symposium on Power
White Paper MR-185, Issue: 1, August 2009, http://www Line Communications (ISPLC 01), pp. 8590, Malmo, Sweden,
.broadband-forum.org/marketing/download. March 2001.
[28] D. E. Nordell, Communication systems for distribution [44] F. J. Canete, L. Dez, J. A. Cortns, and J. T. Entrambasaguas,
automation, in Proceedings of the IEEE Transmission and Dis- Broadband modelling of indoor powerline channels, IEEE
tribution Conference and Exposition, Bogota, Colombia, April Transactions on Consumer Electronics, vol. 48, no. 1, pp. 175183,
2008. 2002.
[29] S. Mak and D. Reed, TWACS, a new viable twoway automatic [45] J. A. Cortns, F. J. Canete, L. Dez, and J. T. Entrambasaguas,
communication system for distribution networks. Part I: out- Characterization of the cyclic short-time variation of indoor
bound communication, IEEE Transactions on Power Apparatus power-line channels response, in Proceedings of the Inter-
and Systems, vol. 101, no. 8, pp. 29412949, 1982. national Symposium on Power Line Communications and Its
Applications (ISPLC 05), pp. 326330, Vancouver, Canada,
[30] S. Mak and T. Moore, TWACS, a new viable twoway auto-
April 2005.
matic communication system for distribution networks. Part
II: inbound communication, IEEE Transactions on Power [46] F. J. Canete Corripio, J. A. Cortns Arrabal, L. Dez del Ro,
Apparatus and Systems, vol. 103, no. 8, pp. 21412147, 1984. and J. T. Entrambasaguas Munoz, Analysis of the cyclic short-
term variation of indoor power line channels, IEEE Journal on
[31] The OPEN meter Consortium, Open Public Extended Net- Selected Areas in Communications, vol. 24, no. 7, pp. 13271338,
work Metering, Home page, http://www.openmeter.com/. 2006.
[32] PRIME Alliance, Powerline related intelligent metering evolu- [47] M. Zimmermann and K. Dostert, A multipath model for the
tion (PRIME), http://www.prime-alliance.org,. powerline channel, IEEE Transactions on Communications, vol.
[33] Electricite Reseau Distribution France, G3-plc: Open standard 50, no. 4, pp. 553559, 2002.
for smart grid implementation, http://www.maximintegrated [48] M. Babic, M. Hagenau, K. Dostert, and J. Bausch, Theoretical
.com/products/powerline/g3-plc/. postulation of PLC channel model, The OPERA Consortium,
[34] K. Razazian, M. Umari, A. Kamalizad, V. Loginov, and M. IST Integrated Project Deliverable D4v2.0, March 2005.
Navid, G3-PLC specification for powerline communication: [49] H. Liu, J. Song, B. Zhao, and X. Li, Channel study for medium-
Overview, system simulation and field trial results, in Proceed- voltage power networks, in Proceedings of the IEEE Interna-
ings of the 14th Annual International Symposium on Power Line tional Symposium on Power Line Communications (ISPLC 06),
Communications and its Applications (ISPLC 10), pp. 313318, pp. 245250, Orlando, Fla, USA, March 2006.
Rio de Janeiro, Brazil, March 2010. [50] European Telecommunication Standards Institute (ETSI),
[35] P. Siohan, A. Zeddam, G. Avril et al., State of the art, Powerline Telecommunications (PLT); MIMO PLT; Part 1:
application scenario and specific requirements for PLC, Measurement Methods of MIMO PLT, February 2012.
OMEGA, European Union Project Deliverable D3.1 v1.0, IST [51] C. R. Paul, Analysis Of Multiconductor Transmission Lines, John
Integrated Project No ICT-213311, 2008, http://www.ict-omega Wiley & Sons, 1994.
.eu/publications/deliverables.html.
[52] T. Magesacher, P. Odling, P. O. Borjesson et al., On the capacity
[36] G. Wood and M. Newborough, Dynamic energy-consumption of the copper cable channel using the common mode, in
indicators for domestic appliances: environment, behaviour and Proceedings of the IEEE Global Telecommunications Conference,
design, Energy and Buildings, vol. 35, no. 8, pp. 821841, 2003. pp. 12691273, November 2002.
[37] International Organization for Standartization, Building [53] T. Sartenaer, Multiuser communications over frequency selective
automation and control systemspart 5: data communication wired channels and applications to the powerline access network
protocol, international standard ISO 16484-5, 2010. [Ph.D. thesis], Faculty of Applied Sciences of the Universite
[38] International Organization for Standartization, Information Catholique de Louvain, Louvain-la-Neuve, Belgium, 2004.
technologyHome electronic system (HES) architecturepart [54] D. Sabolic, A. Bazant, and R. Malaric, Signal propagation
35: Media and media dependent layers Powerline for network modeling in power-line communication networks, IEEE Trans-
based control of HES class 1, international standard ISO/IEC actions on Power Delivery, vol. 20, no. 4, pp. 24292436, 2005.
14543-3-5, First edition, May 2007. [55] T. Banwell and S. Galli, A novel approach to the modeling
[39] International Organization for Standartization, Interconnec- of the indoor power line channel part I: circuit analysis and
tion of information technology equipmentControl network companion model, IEEE Transactions on Power Delivery, vol.
protocolPart 3: Power line channel specification, interna- 20, no. 2, pp. 655663, 2005.
tional standard ISO/IEC 14908-3, Rvision 11, January 2011. [56] S. Galli and T. Banwell, A novel approach to the modeling of
[40] American National Standards Institute/Electronic Industries the indoor power line channelpart II: transfer function and
Association (ANSI/EIA), Control network power line (PL) its properties, IEEE Transactions on Power Delivery, vol. 20, no.
channel specification, ANSI/CEA-709.2-A, September 2006. 3, pp. 18691878, 2005.
[41] L. T. Berger and G. Moreno-Rodrguez, Power line com- [57] T. Magesacher, P. Odling, P. O. Borjesson, and S. Shamai,
munication channel modelling through concatenated IIR-filter Information rate bounds in common-mode aided wireline
elements, Academy Publisher Journal of Communications, vol. communications, European Transactions on Telecommunica-
4, no. 1, pp. 4151, 2009. tions, vol. 17, pp. 533545, 2006.
[58] R. Hashmat, P. Pagani, A. Zeddam, and T. Chonave, MIMO [72] M. Tlich, P. Pagani, G. Avril et al., PLC channel
communications for inhome PLC Networks: Measurements characterization and modelling, OMEGA, European Union
and results up to 100 MHz, in Proceedings of the 14th Annual Project Deliverable D3.3 v1.0, IST Integrated Project No ICT-
International Symposium on Power Line Communications and 213311, December 2008, http://www.ict-omega.eu/publications/
its Applications (ISPLC 10), pp. 120124, Rio de Janeiro, Brasil, deliverables.html.
March 2010. [73] European Committee for Electrotechnical Standardization
[59] R. Hashmat, P. Pagani, A. Zeddam, and T. Chonave, A channel (CENELEC), Signalling on low-voltage electrical installations
model for multiple input multiple output in-home power line in the frequency range 3 kHz to 148,5 kHz - Part 1: General
networks, in Proceedings of the IEEE International Symposium requirements, frequency bands and electromagnetic distur-
on Power Line Communications and Its Applications (ISPLC 11), bances, Standard EN 50065-1, September 2010.
pp. 3541, April 2011. [74] National Institute of Standards and Technology (NIST), Pri-
[60] F. Versolatto and A. M. Tonello, A MIMO PLC random channel ority Action Plan 15 (PAP15): Harmonize Power Line Car-
generator and capacity analysis, in Proceedings of the IEEE rier Standards for Appliance Communications in the Home,
International Symposium on Power Line Communications and Coexistence of narrow band power line communication tech-
Its Applications (ISPLC 11), pp. 6671, Udine, Italy, April 2011. nologies in the unlicensed FCC band, April 2010, http://www
[61] L. T. Berger, A. Schwager, P. Pagani, and D. Schneider, MIMO .fcc.gov/encyclopedia/rules-regulations-title-47.
Power Line Communications: Narrow and Broadband Standards, [75] L. T. Berger, Wireline communications in smart grids, in
EMC, and Advanced Processing, CRC Press, 2013. Smart GridApplicacions, Communications and Security, L. T.
[62] A. Schwager, W. Baschlin, J. Moreno et al., European MIMO Berger and K. Iniewski, Eds., chapter 7, John Wiley & Sons,
PLC field measurements: overview of the ETSI STF410 cam- Hoboken, NJ, USA, 2012.
paign & EMI analysis, in Proceedings of the International [76] FCC, Title 47 of the code of federal regulations (CFR), Tech.
Symposium on Power Line Communications and Its Applications Rep. 47 CFR 15, Federal Communications Commission, 2008,
(ISPLC 12), Beijing, China, 2012. http://www.fcc.gov/encyclopedia/rules-regulations-title-47,.
[63] D. Schneider, A. Schwager, W. Baschlin, and P. Pagani, Euro- [77] Association of Radio Industries and Businesses (ARIB), Power
pean MIMO PLC field measurements: channel analysis, in line communication equipment (10kHz-450kHz), sTD-T84,
Proceedings of the 16th IEEE International Symposium on Power Ver. 1.0, (in Japanese), November 2002, http://www.arib.or.jp/
Line Communications and Its Applications (ISPLC 12), pp. 304 english/html/overview/doc/1-STD-T84v1 0.pdf.
309, March 2012.
[78] Comite International Special des Perturbations Radioelectri-
[64] A. Schwager, Powerline communications: significant technolo-
ques, Information technology equipment; Radio disturbance
gies to become ready for integration [Ph.D. thesis], Univer-
characteristics; Limits and methods of measurement, ICS
sitat Duisburg-Essen, Fakultat fur Ingenieurwissenschaften,
CISPR, International Standard Norme CISPR 22:1997, 1997.
Duisburg-Essen, Germany, 2010.
[79] Comite International Special des Perturbations Radio-
[65] A. Schwager, D. Schneider, W. Baschlin, A. Dilly, and J. Speidel,
electriques, Specification for radio disturbance and immunity
MIMO PLC: theory, measurements and system setup, in IEEE
measuring apparatus and methods part 1-1: Radio disturbance
International Symposium on Power Line Communications and Its
and immunity measuring apparatusMeasuring apparatus,
Applications (ISPLC 11), pp. 4853, April 2011.
2003.
[66] M. Zimmermann and K. Dostert, An analysis of the broadband
noise scenario in power-line networks, in Proceedings of the [80] Comite International Special des Perturbations Radio-
International Symposium on Power Line Communications and electriques, Amendment to CISPR 22: Clarification of its
Its Applications (ISPLC 00), pp. 131138, Limerick, Ireland, April application to telecommunicationsystem on the method of
2000. disturbance measurement at ports used for PLC, 2003.
[67] F. J. Canete, J. A. Cortns, L. Dez, and J. T. Entrambasaguas, [81] Comite International Special des Perturbations Radioelectri-
Modeling and evaluation of the indoor power line transmis- ques, CISPR 22 am3 f1 ed. 5.0, Limits and method of measure-
sion medium, IEEE Communications Magazine, vol. 41, no. 4, ment of broadband telecommunication equipment over power
pp. 4147, 2003. lines, February 2008.
[68] R. Hashmat, P. Pagani, A. Zeddam, and T. Chonavel, Measure- [82] Comite International Special des Perturbations Radioelectri-
ment and analysis of inhome MIMO PLC channel noise, in ques, Report on mitigation factors and methods for power line
Proceedings of the 4th Workshop on Power Line Communications, telecommunications, February 2008.
Boppard, Germany, September 2010. [83] European Telecommunication Standards Institute (ETSI),
[69] P. Pagani, R. Hashmat, A. Schwager, D. Schneider, and W. PowerLine Telecommunications (PLT); Coexistence between
Baschlin, European mimo plc field measurements: noise anal- PLT Modems and Short Wave Radio broadcasting services,
ysis, in Proceedings of the 16th IEEE International Symposium August 2008.
on Power Line Communications and Its Applications (ISPLC 12), [84] Comite International Special des Perturbations Radio-
pp. 310315, March 2012. electriques, Amendment 1 to CISPR 22 ed. 6.0: Addition of
[70] L. T. Berger, Broadband powerline communications, in Con- limits and methods of measurement for conformance testing
vergence of Mobile and Stationary Next Generation Networks, of power line telecommunication ports intended for the
K. Iniewski, Ed., chapter 10, pp. 289316, John Wiley & Sons, connection to the mains, July 2009.
Hoboken, NJ, USA, 2010. [85] C. European Committee for Electrotechnical Standardization,
[71] A. Schwager, L. Stadelmeier, and M. Zumkeller, Potential Power line communication apparatus used in low-voltage
of broadband power line home networking, in Proceedings installationsRadio disturbance characteristicsLimits and
of the 2nd IEEE Consumer Communications and Networking methods of measurementPart 1: Apparatus for in-home use,
Conference (CCNC 05), pp. 359363, January 2005. November 2012.
[86] R. Razafferson, P. Pagani, A. Zeddam et al., Report on [101] Institute of Electrical and Electronics Engineers, Local and
electro magnetic compatibility of power line communica- metropolitan area networksSpecific requirements part 15.4:
tions, OMEGA, European Union Project Deliverable D3.3 Wireless medium access control (MAC) and physical layer
v3.0, IST Integrated Project No ICT-213311, April 2010, (PHY) specifications for low-rate wireless personal area net-
http://www.ict-omega.eu/publications/deliverables.html. works (WPANs), standard for Information Technology
[87] Institute of Electrical and Electronics Engineers, IEEE stand- Telecommunications and Information Exchange Between Sys-
ard for power line communication equipmentElectromagnetic tems, September 2006.
compatibility (EMC) requirementsTesting and measurement [102] Z. Shelby and C. Bormann, 6LoWPAN: TheWireless Embedded
methods, January 2011. Internet, John Wiley & Sons, 2009.
[88] PRIME Alliance, Draft standard for PoweRline Intelligent [103] S. Deering and R. Hinden, Internet protocol, version 6 (IPv6)
Metering Evolution, 2010, www.prime-alliance.org/Docs/Ref/ specification, RFC 2460, December 1998, http://tools.ietf.org/
PRIME-Spec v1.3.6.pdf. html/rfc2460.
[89] ITU - Telecommunication Standardization Sector STUDY [104] American National Standards Institute (ANSI), Utility indus-
GROUP 15, Narrowband orthogonal frequency division mul- try end device data tables, ANSI Standard C12.19, 2008.
tiplexing power line communication transceivers for G3-PLC [105] International Electrotechnical Commission (IEC), Electricity
networks, Recommendation ITU-T G.9903, October 2012, meteringData exchange for meter reading, tariff and load
approved. controlPart 61: Object identification system (OBIS), interna-
[90] ITU - Telecommunication Standardization Sector STUDY tional standard IEC 62056-61, second edition, November 2006.
GROUP 15, Narrowband orthogonal frequency division mul- [106] International Electrotechnical Commission (IEC), Electricity
tiplexing power line communication transceivers for PRIME meteringData exchange for meter reading, tariff and load
networks, Recommendation ITU-T G.9904, October 2012, controlPart 62: Interface classes, international standard IEC
approved. 62056-62, second edition, November 2006.
[91] NIST Priority Action Plan 15, Narrowband PLC coexistence [107] M. Hoch, Comparison of PLC G3 and PRIME, in Proceedings
requirement, October 2011, http://collaborate.nist.gov/twiki- of the IEEE International Symposium on Power Line Communi-
sggrid/pub/SmartGrid/PAP15PLCForLowBitRates/Require cations and Its Applications (ISPLC 11), pp. 165169, Udine, Italy,
ments on NB PLC coexistence Final Oct 11r1.xls. April 2011.
[92] I. Berganza, A. Sendin, and J. Arriola, Prime: powerline [108] V. Oksman and J. Zhang, G.HNEM: the new ITU-T standard
intelligent metering evolution, in Proceedings of the CIRED on narrowband PLC technology, IEEE Communications Maga-
Seminar 2008: SmartGrids for Distribution, pp. 13, Frankfurt, zine, vol. 49, no. 12, pp. 3644, 2011.
Germany, June 2008. [109] J. LeClare, IEEE 1901.2 PAP15 update, Presentation to NIST
[93] National Institute of Standards and Technology (NIST), U.S. PAP-15 SGIP Plenary Meeting, Chicago, Ill, USA, at Grid
Department of Commerce, Specification for the advanced Interop, December 2010, http://collaborate.nist.gov/twiki-
encryption standard (AES), Federal Information Processing sggrid/pub/SmartGrid/PAP15PLCForLowBitRates/PAP15
Standards Publication 197, November 2001. IEEE 1901 2 Update.ppt.
[94] International Electrotechnical Commission (IEC), Distribu- [110] ITU - Telecommunication Standardization Sector STUDY
tion automation using distribution line carrier systemsPart 4: GROUP 15, Narrowband orthogonal frequency division mul-
Data communication protocolsSection 32: Data link layer tiplexing power line communication transceiversPower spec-
Logical link control (LLC), November 1997. tral density specification, Recommendation ITU-T G.9901,
[95] International Electrotechnical Commission (IEC), Electricity November 2012, approved.
meteringData exchange for meter reading, tariff and load [111] ITU - Telecommunication Standardization Sector STUDY
controlPart 62: Interface classes, Standard IEC 62056-62, 2nd GROUP 15, Narrowband orthogonal frequency division mul-
ed., November 2006. tiplexing power line communication transceivers for G.hnem
[96] International Electrotechnical Commission (IEC), Distribu- networks, Recommendation ITU-T G.9902, October 2012,
tion automation using distribution line carrier systemsPart approved.
5-1: Lower layer profilesThe spread frequency shift keying (S- [112] O. Monnier, TI delivers flexible power line communications
FSK) profile, Standard IEC 61334-5-1, 2nd ed., 2001. solutions, White Paper, September 2010, http://www.ti.com/lit/
[97] Electricite Reseau Distribution France (ERDF), G3-PLC wp/slyy026/slyy026.pdf.
physical layer specification, August 2009, http://www. [113] Intellon, http://www.intellon.com/.
maximintegrated.com/products/powerline/pdfs/G3-PLC- [114] DS2, Design of Systems on Silicon, Your world. Connected,
Physical-Layer-Specification.pdf. http://www.marvell.com/.
[98] Electricite Reseau Distribution France (ERDF), G3-PLC MAC [115] Gigle Networks Inc., http://www.giglenetworks.com/.
layer specification, August 2009, http://www.maximintegrated [116] Panasonic Communications Co., Ltd., Ideas for life, http://
.com/products/powerline/pdfs/G3-PLC-MAC-Layer-Specifi panasonic.net/corporate/segments/psn/.
cation.pdf. [117] HomePlug Powerline Alliance, https://www.homeplug.org/
[99] Electricite Reseau Distribution France (ERDF), G3-PLC pro- home.
file specification, August 2009, http://www.maximintegrated [118] Universal Powerline Association (UPA), http://en.wikipedia
.com/products/powerline/pdfs/G3-PLC-Profile-Specification .org/wiki/Universal Powerline Association.
.pdf. [119] High Definition Power Line Communication Alliance (HD-
[100] K. Razazian, G3-PLC provides an ideal communication plat- PLC), http://www.hd-plc.org/.
form for the smart gird, in Proceedings of the IEEE Internatinal [120] International Telecommunications Union (ITU), ITU-T Rec-
Symposium on Power Line Communications and Its Applications ommendation G.9960, Unified highspeed wire-line based
(ISPLC 10), Rio de Janeiro, Brazil, March 2010. home networking transceiversFoundation, August 2009.
[121] International Telecommunications Union (ITU), ITU-T Rec- [137] R. T. Fielding, J. Gettys, J. C. Mogul et al., RFC 2616:
ommendation G.9961, Data link layer (DLL) for unified high- Hypertext Transfer Protocol HTTP/1.1, June 1999, http://tools
speed wire-line based home networking transceivers, June .ietf.org/html/rfc2616.
2010. [138] V. Oksman and J. Egan, Applications of ITUT G.9960,
[122] International Telecommunications Union (ITU), ITU-T Rec- ITU-T G.9961 transceivers for Smart Grid applications:
ommendation G.9963, Unified high-speed wire-line based Advanced metering infrastructure, energy management
home networking transceiversMultiple Input/Multiple Out- in the home and electric vehicles, Telecommunication
put (MIMO), g.9963 (ex G.hn-MIMO), September 2011. Standartization Sector of ITU; Series G: Transmission
[123] HomeGrid Forum, For any wire, anywhere in your home, Systems and Media, Digital Systems and Networks, ITU-T
http://www.homegridforum.org/. Technical Paper, June 2010, http://www.itu.int/dms pub/itu-
[124] Institute of Electrical and Electronic Engineers (IEEE), Stan- t/opb/tut/T-TUT-HOME-2010-PDF-E.pdf.
dards Association, Working group P1901, IEEE standard for
broadband over power line networks: Medium access control
and physical layer specifications, http://grouper.ieee.org/
groups/1901/.
[125] Institute of Electrical and Electronics Engineers (IEEE) Stan-
dards Association and P1901 Working Group, IEEE P1901 Draft
standard for broadband over power line networks: Medium
access control and physical layer specifications, July 2009.
[126] Institute of Electrical and Electronics Engineers Standards
Association and Working Group P1905.1, Draft standard
for a convergent digital home network for heterogeneous
technologies, 2012, http://standards.ieee.org/develop/project/
1905.1.html.
[127] S. Galli, A. Kurobe, and M. Ohura, The inter-PHY protocol
(IPP): A simple coexistence protocol for shared media, in
Proceedings of the IEEE International Symposium on Power Line
Communications and its Applications (ISPLC 09), pp. 194200,
Dresden, Germany, April 2009.
[128] H. C. Ferreira, L. Lampe, J. Newbury, and T. G. Swart, Eds.,
Power Line Communications: Theory and Applications For Nar-
rowband and Broadband Communications Over Power Lines,
John Wiley & Sons, 2010.
[129] International Telecommunications Union (ITU), ITU-T Rec-
ommendation G.9972, Coexistence mechanism for wireline
home networking transceivers, June 2010.
[130] D. Su and S. Galli, PAP 15 recommendations to SGIP on
broadband PLC coexistence, December 2010, http://collaborate
.nist.gov/twiki-sggrid/pub/SmartGrid/
PAP15PLCForLowBitRates/PAP15 - Recommendation to
SGIP BB CX - Final - APPROVED 2010-12-02.pdf.
[131] Consortium for Smart Energy Profile 2 Interoperability, CSEP
consortium for SEP2 interoperability, 2011, http://www.csep
.org/.
[132] ZigBee Alliance, Smart Energy Profile 2.0 Application Protocol
Specification, v 0.7, April 2010, http://zigbee.org/Standards/
ZigBeeSmartEnergy/Version20Documents.aspx.
[133] International Electrotechnical Commission (IEC), Application
integration at electric utilitiesSystem interfaces for distribu-
tion management, international standard IEC 61968, under
development.
[134] International Electrotechnical Commission (IEC), Energy
management system application program interface (EMS-
API), international standard IEC 61970, under development.
[135] World Wide Web Consortium (W3C), Extensible markup lan-
guage (XML) 1.0, W3C Recommendation, 5th ed., November
2008, http://www.w3.org/TR/REC-xml/.
[136] R. T. Fielding, Architectural styles and the design of network-
based software architectures [Ph.D. thesis], University of
California, Irvine, Calif, USA, 2000, http://www.ics.uci.edu/
fielding/pubs/dissertation/top.htm.
http://dx.doi.org/10.1155/2013/402514
Research Article
Radiation Mitigation for Power Line Communications Using
Time Reversal
Amilcar Mescco,1 Pascal Pagani,2 Michel Ney,2 and Ahmed Zeddam1

1
Orange Labs Networks, 2 Avenue Pierre Marzin, 22200 Lannion, France
2
Telecom Bretagne, Microwave Department, Technopole Brest-Iroise, 29238 Brest, France
Correspondence should be addressed to Amilcar Mescco; amescco@gmail.com
Received 4 August 2012; Revised 16 January 2013; Accepted 21 January 2013
Academic Editor: Moises V. Ribeiro
Copyright 2013 Amilcar Mescco et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Power Line Communication (PLC) is the response for nowadays high demand of multimedia services in domestic environment,
not only for its fast and reliable transfer characteristics but also for its flexible low cost implementation, since the PLC technology
uses the existing electrical network infrastructure and the omnipresent outlets throughout the home. The transfer of such a high bit
rate through the mains network generates acceptable radiated emission regulated by international standards, but the increment in
speed for new generation PLC may cause higher levels of emissions. This paper explains the use of the Time Reversal (TR) technique
to mitigate radiated emissions from PLC systems. This method was probed experimentally in real electrical networks with excellent
results: in 40% of the observations, the Electromagnetic Interference (EMI) generated by PLC transmission could be reduced by
more than 3 dB, and this EMI mitigation factor could increase to more than 10 dB in particular configurations.
1. Introduction at frequencies below 500 kHz, in both indoor and out-

door configurations, using LV or Medium Voltage (MV)
With the increasing demand for both high data rate appli- infrastructures [4]. These systems allow the transmission of
cations and reliable links for command and control systems, command and control information over longer distances for
Power Line Communication (PLC) has emerged in recent Smart Grid applications. The ITU-T G.9955 standard is an
years as an attractive communication technique [1]. The main example of such NB PLC systems [5].
advantage of this technology is its ability to benefit from the LV or MV electrical wires were not initially designed to
existing electrical network infrastructure for the transmission propagate communication signals at frequencies above 1 kHz.
of electromagnetic signals. Hence, it becomes possible to As a consequence, the communication channel between the
build large communication networks without the need for transmitter and the receiver is a difficult channel, generating
installing new wires. attenuation and multiple propagation paths. The channel
In the home or office environment, indoor PLC uses capacity is, hence, limited, and signal processing needs to be
the Low Voltage (LV) infrastructure. The presence of several optimized so as to maximize the offered data rate and Quality
electrical outlets in each room of the house allows ubiquitous of Service (QoS).
coverage of the communication network. In addition, the This paper focuses on one of the main limitations related
relatively short distance between different outlets allows the to the PLC technology, namely, the generation of uninten-
system to operate within a limited attenuation. Current Broad tional radiated signal. This phenomenon is mainly due to
Band (BB) in-home systems primarily operate in the fre- the unbalanced nature of the electrical network [6]. The
quency range from 2 MHz to 30 MHz. However, recent spec- variation of the impedances of the loads connected to the
ifications, such as IEEE 1901 [2] or ITU-T G.9960 [3] allow network as well as the unequal length of the phase and
signal transmission at higher frequencies up to 100 MHz. On neutral wires (due to single-phase switches) converts the
the other hand, Narrow Band (NB) PLC systems are deployed differential PLC signal into common mode current flowing
50 at the transmitter, hence generating less Electro Magnetic

55 Interference (EMI). Second, we target to reduce the level of
PSD limit (average) (dBm/Hz)
energy dissipated at any location except the intended receiver.

60
In particular, it is desirable to minimize the level of radiated
65 power from the electrical wires. These two benefits already
70 appear as features of a known technique in the field of
75 wireless transmission: Time Reversal (TR) [13]. Experimental
investigations conducted using Ultra Wide Band (UWB)
80
radio waves demonstrated both the focusing and interference
85 mitigation properties of this technique [14, 15].
90 In this paper, we present for the first time an experimental
95
analysis of TR as a mean to mitigate radiation effects for
0 5 10 15 20 25 30 35 40 wired signal transmission. Our investigation focuses on High
Frequency (MHz) Frequency (HF) BB PLC signals but could be extended to NB
PLC and other wired transmission systems, such as Digital
Figure 1: Example of PSD limit (average) for North America. Subscriber Line (DSL) access. The paper is organized as
follows. Section 2 presents the concepts of the TR technique
and its application to wired systems. Section 3 details the
through the network. Consequently, the copper wires used experiment conducted to assess the merits of TR for BB PLC,
for transmitting the useful signal act as an antenna, and part and the results are statistically analyzed in Section 4. Finally,
of the transmitted power is radiated. This not only results in conclusions are drawn in Section 5.
stronger signal attenuation at the receiver but also leads to
Electromagnetic Compatibility (EMC) issues, as the radiated 2. Time Reversal for Power
signal may interfere with other existing services, such as
amateur radio (HAM) or Short Wave (SW) broadcasting. Line Communications
The impact of PLC transmission on EMC has been studied, 2.1. Time Reversal for Wireless Transmission. The TR tech-
for example, within the ICT FP7 project OMEGA [7] and nique, also known as phase conjugation in the frequency
through the ETSI Specialist Task Force 410 [8]. domain, was first used in the fields of acoustics [16, 17].
In order to avoid interference between PLC systems and More recently, this concept has been successfully extended
other users of the spectrum, regulation authorities impose to electromagnetic waves, where the rich multipath channel
strict emission masks for the transmission of electromagnetic provides excellent conditions for its application [13]. The
signals on the power lines. In the USA, the Federal Commu- basic concept of TR is simple. Let () be an ideal, Dirac
nications Commission (FCC) Part 15 [9] specifies a maximum impulse emitted by a transmitter (Tx) antenna (Figure 2(a)).
level of radiated field for carrier-current systems (including By definition, at any receiver (Rx) location r0 , the received
PLC), leading system specifications to define constrained signal is given by the Channel Impulse Response (CIR)
power transmission masks. Figure 1 represents an example of (, r0 ). The CIR is composed of multiple echoes reflecting
Power Spectral Density (PSD) limits provided in the IEEE the multiple propagation paths of the propagation channel.
1901 standard for North America. The observed notches are TR uses this Channel State Information (CSI) at the Tx
defined to protect specific systems, such as HAM bands. to prefilter the signal to be transmitted. More specifically,
In Europe, CENELEC is currently developing a draft on the CIR (, r0 ) is time reversed and normalized to serve as
regulation standard applying to in-home PLC systems [10]. an input filter for the signal to be transmitted (Figure 2(b)).
Regardless of the regulation limits in place, the research Physically, each delayed echo constituting the TR filter travels,
presented in this paper focuses on the mitigation of unin- among other multiple paths, through its original propagation
tentional radiation due to PLC systems. Several attempts to path. As a result, the multiple echoes sum up coherently at the
solve this problem have been presented in the literature. receiver, hence focusing the received energy in time.
Reference [11] presents a method to reduce radiated emissions Mathematically, applying TR leads for any Rx situated
by applying an auxiliary signal cancelling the electromagnetic at an arbitrary location r to the equivalent perceived CIR
field on a given point in space. Simulations demonstrated TR (, r) (note that this formulation holds for a real valued
good performance, with the drawback that the Electro CIR) [18]:
Magnetic Interference (EMI) could only be mitigated at a
single location. The authors in [12] used additional hardware (, r0 )
TR (, r) = (, r) , (1)
(, r0 )2
connected at the wall outlets, in order to reduce asymmetries
on power lines, at the cost of an increased complexity of the
PLC network.
where the symbol denotes time domain convolution.
In our approach, we tried to simultaneously reach two
Formulating (1) in the frequency domain leads to
complementary goals by the means of digital signal pro-
cessing. First, we intend to focus the transmitted signal at (, r0 )
the receiver location. The power gain linked to this energy TR (, r) = (, r) , (2)
focalization allows in turn relaxing the required power level (, r0 )2
( ,r0) ( ,r0) ( ,r0)

( ) ( , r0)
Tx Rx Tx Rx
(a) (b)
Figure 2: Transmission over an ideal multipath propagation channel (a) of a Dirac impulse, (b) of the time-reversed CIR.
where (, r) represents the complex valued Channel Trans- 2.2. Extension of Time Reversal to Wired Transmission. As
fer Function (CTF), and the superscript denotes the com- observed through the study of TR for wireless transmission,
plex conjugate operation. For this reason, TR is sometimes the TR scheme provides two main features, namely, an
called frequency domain phase conjugation. increase of the Rx power at the intended Rx location and a
Two conclusions can be drawn from (1) and (2). First, at decrease of the Rx power at any other location. These features
the intended Rx location r0 , the perceived CIR simplifies to are highly desirable in the context of wired transmission,
where the level of Tx power is constrained by the uninten-
1
TR (, r0 ) = (, r0 ) , tional radiation from the wires, causing possible EMI to other
(, r0 )2
(3)
systems.
Based on this observation, we conducted experimental
where (, r0 ) denotes the time domain autocorrelation of studies to analyze the potential of TR to mitigate unwanted
the function (, r0 ). Similarly, the perceived CTF simplifies emissions for PLC systems. The main principles of the
to extension of TR to PLC transmission can be explained with
1 2 the help of Figure 3.
TR (, r0 ) = (, r0 ) . (4)
(, r0 )
2 We assume an intended transmission between a Tx PLC
modem and an Rx PLC modem, over an LV indoor electrical
In the time domain, the effect of the TR filter is to trans- network. Different experimental investigations reported the
form the CIR into its autocorrelation. For a rich multipath PLC channel as a rich multipath propagation channel, due to
environment, the autocorrelation of the CIR presents a large the multiple branches present in a classical electrical network
peak at = 0, with reduced side echoes. Experimental and to the impedance mismatch occurring at the network
studies demonstrated that the resulting channel is less spread terminations (outlets) and nodes [2023]. This similarity of
in time [14, 15], hence reducing the possible Intersymbol the PLC channel with wireless channels suggests promising
Interference (ISI). In the frequency domain, the perceived results when applying TR to PLC.
CTF is proportional to the square of the magnitude of the With reference to Figure 3, the Rx modem is situated at
actual CTF. Besides the fact that TR provides a real valued the intended location r0 , and the Tx modem is situated at the
CTF, which could be exploited at the receiver, this also leads origin. By applying TR filtering at the Tx, the Rx power will
to a significant gain in terms of Rx power, due to a better be increased at location r0 ; hence, the Rx modem will benefit
exploitation of the frequency selective nature of the channel. from an increased Signal-to-Noise Ratio (SNR). This power
It was demonstrated in [18] that the application of TR in a flat increased can also in turn be applied as a reduction of the Tx
channel (i.e., without frequency domain power decay) under power to achieve similar performance. At other outlets in the
Rayleigh fading leads to a gain of 3 dB in the total received network, situated, for example, at locations r1 or r2 , the Rx
power. This gain was increased to 5 dB when considering the power will be reduced. This effect can be further exploited in
frequency domain power decay observed in practical UWB the design of multiuser transmission schemes.
radio channels. For our purpose of radiation mitigation, let us now
The second conclusion drawn from (1) and (2) is that, for consider a location r3 , situated at any point in space in the
any other location r different from r0 , TR creates a mismatch vicinity of the electrical network. The level of radiated field
between the Tx filter and the channel. This is particularly at this location can be evaluated, for instance, by the means
observable in the frequency domain representation of TR of an equivalent transfer function (, r3 ) between the Tx
given in (2). The perceived CTF corresponds to the product modem and an ideal antenna situated at location r3 . By the
of two independent CTF, (, r) and (, r0 ), with possibly virtue of the TR scheme (2), the perceived transfer function
very different frequency fading structures. More precisely, at location r3 after applying TR will be proportional to the
minima of the first CTF can happen randomly at maxima of product (, r0 ) (, r3 ). As the functions (, r0 ) and
the second CTF. Hence, averaging over all frequencies, the (, r3 ) are not correlated, their frequency fading structures
total received power at untargeted locations is reduced. In are different. In particular, the deep notches due to frequency
wireless TR analysis, this effect is called spatial focusing and selective fading do not appear at the same frequencies. As
is generally assessed as the ratio between the maximum of a result, the product (, r0 ) (, r3 ) will provide more
TR (, r) and the maximum of TR (, r0 ) for a given distance average attenuation when compared to (, r3 ) alone, and
r r0 . Spatial focusing factors of 10 dB have been reported therefore, the total power radiated at location r3 will be
in [18, 19]. reduced. A similar observation is made in previous studies
r1 Mains
Balun1 Balun2
r2 Signal generator Coaxial 3
Oscilloscope
Coaxial 2
Coaxial 1
Attenuator
Tx Rx r0 Amplifier Biconical antenna
Coaxial 4
Figure 5: Experimental setup calibration.
r3
3.2. Calibration. With reference to Figure 5, two measure-
ment paths can be distinguished. Path 1 serves for the
Figure 3: Principle of the extension of TR to wired transmission. measurement of the CTF and is composed of the following
elements: the transmitter, coaxial cable 1, a 30 dB amplifier
IFIM50, coaxial cable 2, balun 1, mains, balun 2, coaxial cable
Signal generator Oscilloscope 3, a 20 dB attenuator Radial R412720000, and the receiver. The
Mains baluns were considered as part of the channel. The calibration
for measurement path 1 was made by directly connecting
( ,r0) Balun2
Balun1 Rx coaxial cable 2 and coaxial cable 3.
Tx Position r0
() ()
Path 2 is composed by the transmitter, coaxial cable 1,
a 30 dB amplifier IFIM50, coaxial cable 2, the wire to free
( ,r3) space propagation channel (represented by the CTF (, r3 )),
the biconical antenna, coaxial cable 4, and the DSO. The
calibration for this second path was made by connecting
Biconical antenna coaxial cable 2 and coaxial cable 4. The gain of the antenna
Position r3
was removed from the measurements by postprocessing.
Figure 4: Equipment used in the experimental setup. The signal generator and the DSO were synchronized by
a direct connection of their 10 MHz reference clocks. The
sampling rates of both devices were set to = 100 MHz.
dedicated to wireless channels [14, 15, 18]. Hence, TR appears 3.3. Signal Processing. After initial system calibration, the
as an efficient method to mitigate EMI for wired communi- measurements are performed in 3 steps.
cations.
The theoretical basis for the application of TR to wired (1) The CTF (, r0 ) is evaluated using a specific Tx
transmission being set, we will now describe the experimental frame (see below). The considered frequency band
assessment of this method in the next Sections. extends from 2.8 MHz to 37.5 MHz.
(2) The TR filter is generated using the phase and magni-
tude of (, r0 ).
3. Experimental Setup
(3) The CTFs (, r0 ) and (, r3 ) as well as the per-
3.1. Equipment. In order to experimentally assess the use of ceived CTFs TR (, r0 ) and TR (, r3 ) are measured
TR as a method to mitigate EMI for wired communication, using a single Tx frame (see below). Measurements
we used the experimental setup presented in Figure 4. In this at the location of the Rx outlet (r0 ) and at arbitrary
setup, a signal generator Tektronix AWG7082C was used as a locations in space (r3 ) are made simultaneously using
generic Tx. A Digital Sampling Oscilloscope (DSO) LeCroy two ports of the oscilloscope, one connected to the
WaveRunner 715Zi-A was used to sample the received signal Rx balun and the other connected to the biconical
at Rx. The signal generated at Tx is denoted (), and the antenna.
signal received at the DSO is denoted (). Two Universal
PLC Couplers were used as baluns to couple the Tx and Rx The Tx signal is generated according to the HomePlug
signals with the power lines. The Tx coupler is situated at the standard [25]. The used frame is called PHY Protocol Data
origin, and the Rx coupler is connected to a plug at location Unit (PPDU) and it is composed by a preamble (including a
r0 . These couplers were developed within the ETSI Specialist Frame Control) and a number of payload symbols as shown
Task Force 410 [24]. In order to measure the power density in Figure 6. Each payload symbol consists of a 3072-sample
of the radiated emission received at any arbitrary location r3 , OFDM symbol as defined in the HomePlug specification [25].
a biconical antenna Schwarzbeck EFS921 was connected to a In order to estimate the CTFs (, r) and TR (, r), the
second port of the DSO. OFDM symbols were loaded with predefined constellations.
Preamble Symbol 1 Symbol 2 Symbol 3 Symbol 4 Symbol 5 Symbol 6 Preamble () () () () () ()
Figure 6: HomePlug frame. Figure 8: HomePlug frame for CTF and EMI measurement after
application of TR.
Preamble () () () () () ()
Figure 7: HomePlug frame for calibration and CTF measurement

before computation of the TR filter.
The frame used for the calibration and the initial mea-
surement of the CTF before computation of the TR filter is
represented in Figure 7.
The frame used for the measurement of the CTF and
EMI after computation of the TR filter is represented in
Figure 8. Note that in this frame, the TR filter is applied on the
3 last symbols only (represented by the signal ()). From this Figure 9: Picture of experimentation.
particular frame scheme, the CTF and EMI can be evaluated
with and without application of TR quasi-simultaneously.
Recomputing the CTF at this stage also allows monitoring electrical outlets (between 4 and 10). About half of the outlets
any possible temporal evolution of the channel between were connected to classical office appliances (lamps, desktops,
calibration and measurement. etc.). For each topology, one CTF was measured, first without
In essence, the overall channel estimation process is applying TR and then after applying TR filtering. In addition,
similar to the channel sounding procedure defined in the for each topology, between 3 and 5 locations were selected
HomePlug AV specification. In our experiment, the compu- to measure the received electrical field with the help of the
tation of the CTF from the received signal uses a classical biconical antenna. In total, 13 CTF and 43 measurements of
Zero Forcing (ZF) channel estimation procedure. This is the electrical field were collected for statistical analysis.
suitable in our experiment involving high levels of Signal-to-
Noise Ratio (SNR) at the Rx. In practical systems operating 4. Results and Statistical Analysis
at lower SNR, more sophisticated methods such as the
Minimum Mean Square Error channel estimation would be 4.1. Preliminary Results. The preliminary results for an exem-
more efficient against noise enhancement. Note that we chose plar network will be presented in two parts: first, the CTF at r0
to actually implement the TR filter in the time domain at and second the electric field and its associated power density
the transmitter using a programmable waveform generator, at r3 = r0 .
in order to measure results as close as possible to a realistic For the CTF at r0 , we show its attenuation characteristics
implementation of the TR scheme in practical PLC modems. in Figure 10.
In order to estimate channel attenuation continuously Let us first consider the measured CTF (black curve). The
over the measured frequency band, no spectral notches deep notches at some frequencies are due to reflections at the
were implemented for this study. In practical systems, a Tx terminations of the network and reflect the multipath nature
PSD mask is defined, where the Tx power is notched at of the PLC network. We define the average attenuation before
predefined frequencies, to protect existing services using the
TR (r0 ) in dB as:
same spectrum. The HomePlug specification uses windowing
in order to better exploit the power allocated within the max
1 2
(r0 ) = 10log10 ( (, 0 ) ) ,
max min min
PSD mask while protecting out-of-band services. The present
study concentrates on reducing the EMI within the band
(5)
effectively used by PLC systems. Therefore, the results also
hold for practical systems including notches. where min and max respectively represent the minimum
and maximum sounded frequencies. The average attenuation
3.4. Measured Environment. The measurement campaign corresponds to the signal attenuation perceived by a receiver
was conducted using 13 different topologies of 230 V mains capable of exploiting all the power received in the frequency
networks within the premises of Orange Labs in Lannion. The band from min to max . Typically, an OFDM system like
campaign took place in different rooms of about 5 4 m2 . the HomePlug AV specification is able to exploit the total
Figure 9 presents a picture of the experimental setup in an received power over a wide frequency band. In our example,
exemplar location. the average attenuation is about 11 dB.
The Tx and Rx modems were connected to two outlets Let us now focus on the perceived CTF after applying
in the same room, with distances varying between 2 m and TR. Owing to the mathematical definition of the TR filter,
8 m. In general, the rooms are equipped with several other TR allocates more power to frequencies showing minimal
5 transmission with some degree of channel attenuation. On

0 the contrary, the CTF perceived at Rx after application of TR
5 (red curve) corresponds to a logical channel, where the effects
of the measured CTF are combined with the effects of the TR
Attenuation (dB)
10
15 filter at Tx. Hence, the perceived CTF may exhibit some gain
20 over a limited frequency range.
25
We now consider the electrical field and its associated
power density at r3 = r0 . The value of the electrical field
30
(, r3 ) in dBV/m was computed from the CTF (, r3 )
35
measured between the Tx balun and the antenna connector
40 assuming an injected PSD feed = 55 dBm/Hz and using the
45 following formula [24]:
0 5 10 15 20 25 30 35

Frequency (MHz) (, r3 ) = feed + 20 log10 ( (, r3 )) + 107 + AF () ,
Measured CTF
(7)
Perceived CTF after TR
where AF() represents the antenna factor and 107 represents
Figure 10: Channel attenuation before and after TR. the conversion from dBm to dBV. In addition, we computed
the average radiated power density (r3 ) in dB(W/m2 ) as
40 max
1 1 2
(r3 ) = 10log10 ( (, r3 ) ) ,
35 max min min 120
(8)
Electric field (dBuV/m)
30
25
wheremin and max , respectively, represent the minimum
and maximum sounded frequencies, and (, r3 ) is expressed
20 in V/m. In (8), the term 120 provides the value of the
impedance of free space in Ohm.
15
Note that both (, r3 ) and (r3 ) can also be computed
10 after applying TR filtering, using the following equations:

5 TR (, r3 ) = feed + 20 log10 (TR (, r3 ))+ 107 +AF ()
0 5 10 15 20 25 30 35 (9)
Frequency (MHz)
Before TR AR power density: 119.9dB(W/m2) TR (r3 )
After TR AR power density: 127.4dB(W/m2)
max
1 1 2
= 10log10 ( (, r3 ) ) .
Figure 11: Electric field before and after TR. max min min 120 TR
(10)
attenuation, while strongly attenuated frequencies are more An example of radiated emission measurement is given
power constrained. In particular, for all frequencies where the in Figure 11. For this particular electrical network, the mit-
attenuation of the channel (, r0 ) is higher than the average igation of radiated emissions is clear in the frequency band
attenuation (r0 ), the perceived channel is more attenuated from 26 MHz to 37.5 MHz, and the average radiated power
after applying TR. This can be clearly seen in the frequency density has reduced by about 7.5 dB. Thus, we can observe
range from 26 MHz to 37.5 MHz. For the frequencies where with this example that the application of TR filtering can
the attenuation of the channel is less than (r0 ), the response reduce significantly the level of undesired radiated power.
of the channel is improved using TR. We can see this clearly
in the frequency range from 11 MHz to 22 MHz. If we define 4.2. Statistical Analysis. In this section, we present the statis-
the average attenuation after TR TR (r0 ) in dB as tical analysis of the measurement database collected within
13 rooms of the office building at Orange Labs in Lannion.
max
1 2 The total measurement set is composed of 13 CTF and 43
TR (r0 ) = 10log10 ( (, 0 ) ) ,
max min min TR measurements of the electrical field.
(6) We first computed the channel gain TR observed on the
perceived CTF TR (, r0 ) after application of TR filtering. As
we observe that the average attenuation after applying TR is OFDM systems can exploit the total power received over a
about 7.5 dB. Hence, the application of TR provided a gain in given frequency band, we computed this gain in dB for the
the total received power of 3.5 dB in this particular example. total received power as
Note that only the measured CTF (black curve) is phys-
ical. Therefore, the measured CTF will always degrade the TR = (r0 ) TR (r0 ). (11)
1 1
0.9 0.9
0.6 0.8
0.8 0.7
0.7 0.6
CDF
CDF
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
0 0
1 2 3 4 5 6 7 0 2 4 6 8 10 12
Gain of channel (dB) Reduction of total average power density + gain of channel (dBW/m2)
Figure 12: CDF of the channel gain TR . Figure 14: CDF of the effective EMI mitigation, TR .
1 electrical network topologies, where the rich multipath envi-

0.9
ronment results in different frequency fading structures for
different Rx locations.
0.8
Observing the statistics of the channel gain TR and of
0.7
the EMI reduction coefficient TR , an optimal strategy can be
0.6 proposed in order to minimize EMI for a PLC system. Indeed,
CDF
0.5 Figure 12 shows that the application of TR provides better

0.4 system performance due to the reduced channel attenuation.
0.3 This gives us in turn the flexibility to reduce the level of the
0.2 Tx power to further reduce EMI while keeping the system
0.1 performance constant. More precisely, when TR is applied,
0 a Tx power back-off of TR dB can be applied without
2 0 2 4 6 8 modifying the total received power. Finally, following this
Reduction of total average power density (dBW/m 2) power backoff strategy, the effective EMI mitigation factor
TR can be computed as the sum of the power backoff and
Figure 13: CDF of the EMI reduction coefficient TR .
the EMI reduction coefficient:
The cumulative distribution function (CDF) of TR is TR = TR + TR . (13)

given in Figure 12. This parameter shows always a positive The CDF of the effective EMI mitigation factor TR is
gain, between 1.4 dB and 6.6 dB in our experiment, which is depicted in Figure 14. Several conclusions can be drawn from
in line with results reported in similar wireless experiments this statistical result.
[18]. In about 60% of cases, the channel gain is higher than
3 dB. This means that at Rx we have always a better reception (i) First, the TR method is able to mitigate EMI generated
using TR. This channel gain can in turn be used to reduce by PLC transmission in 100% of our experimental
the injected PSD at Tx, hence reducing by the same factor the observations. To this respect, one can conclude that
unwanted EMI. the gain TR provided by the application of TR allows
We then computed the EMI reduction coefficient TR , a Tx power backoff that largely compensates for the
corresponding to the reduction of the undesired radiated possible EMI increment observed in Figure 13.
power due to the application of a TR filter. This figure of (ii) Second, in 40% of the cases, the undesired radiated
merit of the reduction of radiated signal is computed in dB power is reduced by more than 3 dB.
as follows:
(iii) Finally, in the most favorable configurations, a reduc-
TR = (r3 ) TR (r3 ) (12) tion of the EMI by more than 10 dB can be observed.
Such configurations correspond to cases where the
The CDF of the EMI reduction coefficient TR is given CTF between the Tx and Rx modem and the EMI
in Figure 13. Results show that the simple application of TR spectrum are particularly decorrelated.
reduces the EMI in more than 60% of cases. In the best
case, the EMI reduced by more than 7 dB using TR. In the 5. Conclusion
worst case, the EMI increased by 2 dB. The observations
of particular cases indicates that the reduction of EMI is In this paper, we proposed for the first time the application of
more effective when the CTFs (, r0 ) and (, r3 ) are TR in order to mitigate EMI generated by wired communi-
highly decorrelated. This is more likely to happen in complex cation systems. TR was originally used in the field of wireless
transmission as a mean to focus the transmitted signal in both the IEEE International Symposium on Power Line Communica-
time and space around the intended receiver. We proposed to tions and Its Applications (ISPLC 06), pp. 314318, Orlando, Fla,
use the same property to focus the signal injected in a wired USA, March 2006.
medium, such as the electrical network in the case of PLC [7] Seventh Framework Programme: Theme 3 ICT-213311 OMEGA,
for instance. As a result, the energy lost through undesired Deliverable D3.3, Report on Electro Magnetic Compatibility of
radiation is expected to decrease significantly. Power Line Communications, 2009.
We presented an experimental setup in order to demon- [8] A. Schwager, W. Baschlin et al., European MIMO PLC Field
strate this method. The experiment was conducted in the Measurements: overview of the ETSI STF410 Campaign &
time domain using signal frames similar to the industrial EMI Analysis, in IEEE International Symposium on Power Line
Communications and Its Applications (ISPLC 12), pp. 304309,
specification HomePlug AV. In addition, the TR filter was
Beijing, China, March 2012.
actually implemented using an arbitrary wave generator, thus
[9] Federal Communications Commission, Title 47 of the code of
providing results encompassing the possible drawbacks of a federal regulations part 15, 2007.
practical implementation.
[10] CENELEC, Final Draft European Standard FprEN 50561-1,
Results demonstrated that, on the wired medium, TR Power line communication apparatus used in low voltage
could provide a transmission channel gain between 1 dB and installationsRadio disturbance characteristicsLimits and
7 dB, which is similar to the gain observed in wireless methods of measurementPart 1: Apparatus for in-home use
transmission. In addition, the application of a TR filter alone 2011.
could effectively reduce EMI for 60% of the observations, [11] A. Vukicevic, M. Rubinstein, F. Rachidi, and J. L. Bermudez,
with a maximum EMI mitigation of 7 dB. Finally, by com- On the impact of mitigating radiated emissions on the capacity
bining the channel gain and the EMI reduction features, we of PLC systems, in Proceedings of the IEEE International
experimentally demonstrated that TR was efficient in 100% Symposium on Power Line Communications and Its Applications
of the observed cases to reduce EMI. In 40% of the cases, (ISPLC 07), pp. 487492, Pisa, Italy, March 2007.
the EMI mitigation was larger than 3 dB, with maxima higher [12] P. Favre, C. Candolfi, and P. Krahenbuehl, Radiation and
than 10 dB. TR is, thus, seen as a promising technique to help disturbance mitigation in PLC networks, in Proceedings of
resolving EMC issues related to PLC and other wired media. the 20th International Zurich Symposium on Electromagnetic
In the TR strategy presented in this paper, the CTF Compatibility (EMC Zurich 09), pp. 58, Zurich, Switzerland,
January 2009.
gain provided by the TR technique was fully dedicated as a
power backoff to minimize the EMI. Another strategy could [13] G. Lerosey, J. De Rosny, A. Tourin, A. Derode, G. Montaldo,
and M. Fink, Time reversal of electromagnetic waves, Physical
target an increase of the offered capacity while maintaining a
Review Letters, vol. 92, no. 19, Article ID 193904, 4 pages, 2004.
constant EMI. Further analyses will, therefore, be dedicated
[14] A. E. Akogun, R. C. Qiu, and N. Guo, Demonstrating the lever-
to analyze the tradeoff between channel throughput increase
ages of time reversal in ultra-wideband communications using
and EMI reduction. In addition, future research will focus time domain measurements, in Proceedings of the 51st Inter-
on the study of wired TR at higher frequencies and on other national Instrumentation Symposium, pp. 737742, Knoxville,
media, such as DSL cables. Finally, optimal protocols will be Tenn, USA, May 2005.
developed to practically implement TR in future standards. [15] A. Khaleghi, G. El Zein, and I. H. Naqvi, Demonstration
In particular, the application of TR to multicast or broadcast of time-reversal in indoor ultra-wideband communication:
scenarios, involving one Tx modem and several Rx modems, time domain measurement, in Proceedings of the 4th IEEE
could be further investigated. International Symposium on Wireless Communication Systems
(ISWCS 07), pp. 465468, October 2007.
References [16] A. Derode, P. Roux, and M. Fink, Acoustic time-reversal
through high-order multiple scattering, in Proceedings of the
[1] H. C. Ferreira, L. Lampe, J. Newbury, and T. G. Swart, Eds., IEEE International Ultrasonics Symposium, vol. 2, pp. 10911094,
Power Line Communications: Theory and Applications for Nar- Seattle, Wash, USA, November 1995.
rowband and Broadband Communications Over Power Lines, [17] D. R. Jackson and D. R. Dowling, Phase conjugation in under-
Wiley, Chichester, UK, 2010. water acoustics, Journal of the Acoustical Society of America, vol.
[2] IEEE, 1901-2010, IEEE standard for broadband over power line 89, pp. 171181, 1991.
networks: medium access control and physical layer specifica- [18] P. Pajusco and P. Pagani, On the use of uniform circular arrays
tions, 2010. for characterizing UWB time reversal, IEEE Transactions on
[3] ITU-T G. 9960, Unified high-speed wireline-based home net- Antennas and Propagation, vol. 57, no. 1, pp. 102109, 2009.
working transceiverssystem architecture and physical layer [19] C. Zhou and R. C. Qiu, Spatial focusing of time-reversed
specification, 2010. UWB electromagnetic waves in a hallway environment, in
[4] S. Galli, A. Scaglione, and Z. Wang, For the grid and through Proceedings of the 38th Southeastern Symposium on System
the grid: the role of power line communications in the smart Theory, pp. 318322, Cookeville, Tenn, USA, March 2006.
grid, Proceedings of the IEEE, vol. 99, no. 6, pp. 9981027, 2011. [20] M. Tlich, A. Zeddam, F. Moulin, and F. Gauthier, Indoor
[5] V. Oksman and J. Zhang, G.HNEM: the new ITU-T standard power-line communications channel characterization up to
on narrowband PLC Technology, IEEE Communications Mag- 100 MHzpart I: one-parameter deterministic model, IEEE
azine, vol. 49, no. 12, pp. 3644, 2011. Transactions on Power Delivery, vol. 23, no. 3, pp. 13921401,
[6] M. Ishihara, D. Umehara, and Y. Morihiro, The correlation 2008.
between radiated emissions and power line network compo- [21] M. Tlich, A. Zeddam, F. Moulin, and F. Gauthier, Indoor
nents on indoor power line communications, in Proceedings of power-line communications channel characterization up to 100
MHz-part II: time-frequency analysis, IEEE Transactions on

Power Delivery, vol. 23, no. 3, pp. 14021409, 2008.
[22] A. M. Tonello and F. Versolatto, Bottom-up statistical PLC
channel modeling-part I: random topology model and efficient
transfer function computation, IEEE Transactions on Power
Delivery, vol. 26, no. 2, pp. 891898, 2011.
channel modelingpart II: inferring the statistics, IEEE Trans-
actions on Power Delivery, vol. 25, no. 4, pp. 23562363, 2010.
[24] ETSI TR 101 562-1 V2.1.1, Powerline telecommunications
(PLT), MIMO PLT, part 1: measurements methods of MIMO
PLT, Tech. Rep., 2012, Chapter 7.1.
[25] HomePlug, HomePlug AV specification, version 1.1, 2007.
http://dx.doi.org/10.1155/2013/892628
Research Article
An Overview of the HomePlug AV2 Technology
Larry Yonge,1 Jose Abad,2 Kaywan Afkhamie,1

Lorenzo Guerrieri,3 Srinivas Katar,1 Hidayat Lioe,4 Pascal Pagani,5
Raffaele Riva,6 Daniel M. Schneider,7 and Andreas Schwager7
1
Qualcomm Atheros, 203 East Silver Springs Boulevard, Ocala, FL 34470, USA
2
Broadcom Corporation, Llacuna 5670, 8th Floor, Building A, 08005 Barcelona, Spain
3
DORA S.p.A., STMicroelectronics Group, Via Lavoratori Vittime del Col Du Mont 28, 11100 Aosta, Italy
4
Marvell Semiconductors, 5488 Marvell lane, Santa Clara, CA 95054, USA
5
Orange Labs Networks, 2 Avenue Pierre Marzin, 22300 Lannion, France
6
STMicroelectronics S.r.l., Via Olivetti 2, 20864 Agrate Brianza, Italy
7
Sony, Hedelfinger Strae 61, 70327 Stuttgart, Germany
Correspondence should be addressed to Pascal Pagani; pascal.pagani@ieee.org
Received 3 August 2012; Accepted 21 November 2012
Academic Editor: Andrea M. Tonello
Copyright 2013 Larry Yonge et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
HomePlug AV2 is the solution identified by the HomePlug Alliance to achieve the improved data rate performance required by the
new generation of multimedia applications without the need to install extra wires. Developed by industry-leading participants in
the HomePlug AV Technical Working Group, the HomePlug AV2 technology provides Gigabit-class connection speeds over the
existing AC wires within home. It is designed to meet the market demands for the full set of future in-home networking connectivity.
Moreover, HomePlug AV2 guarantees backward interoperability with other HomePlug systems. In this paper, the HomePlug AV2
system architecture is introduced and the technical details of the key features at both the PHY and MAC layers are described. The
HomePlug AV2 performance is assessed, through simulations reproducing real home scenarios.
1. Introduction wires to deploy wide band services is the main source of

attractiveness of the in-home PLC technologies. Another
The convergence of voice, video, and data within a variety major advantage of PLC is the ubiquity of the power lines
of multifunction devices, along with the evolution of High which can be used to provide whole-home connectivity
Definition (HD) and 3-Dimensional (3D) video, are today solutions. However, the power line medium has not been
driving the demands for home connectivity solutions. Home originally designed for data communication; the frequency
networks are required to support high throughput connec- selectivity of the channel and different types of noise (back-
tivity guaranteeing at the same time a high level of reliability ground noise, impulsive noise, and narrow band interferers)
and coverage (the percentage of links that are able to sustain make the power line a very challenging environment and
a given throughput in two-node or multinode networks). require state of the art design solutions.
Applications such as HD Television (HDTV), Internet Pro- In 2000, the HomePlug Alliance [1], an industry-led
tocol Television (IPTV), interactive gaming, whole-home organization, was formed with the scope to promote power
audio, security monitoring, and Smart Grid management line networking through the adoption of HomePlug specifica-
have to be supported by the new home networks. tions. In 2001, the HomePlug Alliance released the HomePlug
During the last decade, in-home power line communica- 1.0.1 specification followed in 2005 by a second release named
tion (PLC) has received increasing attention from both the HomePlug AV [2]. The letters AV abbreviate Audio, Video.
industry and research communities. The reuse of existing HPAV rapidly became the most widespread adopted solution
for in-home power line communication. To meet the future the performance of the HomePlug AV2 system. The adoption
market needs, in January 2012 the HomePlug Alliance pub- of all the above listed features provides a significant gain of
lished the HomePlug AV2 specification [3]. HomePlug AV2 HomePlug AV2 compared to HomePlug AV. The remainder
enables Gigabit-class connection speeds by leveraging on the of this paper is organized as follows. Section 2 presents
existing power line wires while remaining fully interoperable the overall HomePlug AV2 system architecture preceded
with other technologies for in-home connectivity (HomePlug by a brief overview of the HomePlug AV technology. The
AV [2], HomePlug Green PHY [4], and IEEE 1901 [5]). The new PHY and MAC layer technical features are detailed in
Alliances AV Technical Working Group (AV TWG) defined Sections 3 and 4, respectively. Section 5 explains how the
new features at both the PHY and MAC layers. These have technology coexistence is dealt with. Section 6 summarizes
been introduced in the HomePlug AV2 specification based on the performance gain of HomePlug AV2 compared to Home-
extensive field tests conducted in real home scenarios across Plug AV and Section 7 concludes this introduction to the
different countries. The field results obtained by the AV TWG HomePlug AV2 technology.
validated the HomePlug AV2 claimed performance both in
terms of achievable data rate and coverage.
In this paper, we highlight the key differentiating Home- 2. System Architecture
Plug AV2 features compared to the HomePlug AV technol-
2.1. A Brief Overview of HomePlug AV. HomePlug AV
ogy. At the physical (PHY) layer HomePlug AV2 includes
employs PHY and MAC technology that provides a 200-
the following. (i) Multiple-Input Multiple-Output (MIMO)
Mbps class power line networking capability. The PHY
signaling with Beamforming to offer the benefit of improved
operates in the frequency range of 228 MHz, uses windowed
coverage throughout the home, especially on highly atten-
Orthogonal Frequency Division Multiplexing (OFDM) and
uated channels. MIMO enables HomePlug AV2 devices to
a powerful Turbo Convolutional Code (TCC) that provides
transmit on any two-wire pairs within three-wire configura-
robust performance within 0.5 dB of Shannons limit. Win-
tions comprising Line (L), Neutral (N) and Protective Earth
dowed OFDM provides more than 30 dB spectrum notching.
(PE) (the coupling is done in the MIMO Analog Front End
OFDM symbols with 917 usable carriers (tones) are used
(AFE) blocks in Figure 1). (ii) Extended Frequency Band up
in conjunction with a flexible guard interval. Modulation
to 86 MHz to increase the throughput especially at the low
densities from BPSK to 1024 QAM are adaptively applied to
to mid coverage percentages (part of IFFT/FFT (1024, 8192)
each carrier based on the channel characteristic between the
and Mapper blocks in Figure 1). (iii) Efficient Notching allows
transmitter (Tx) and the receiver (Rx).
transmitters to create extremely sharp frequency notches. If
On the MAC layer, HomePlug AV provides a Quality of
Electromagnetic Compatibility (EMC) regulation will require
Service (QoS) connection oriented, contention-free service
a fragmented communication spectrum, the throughput loss
on a periodic Time Division Multiple Access (TDMA) alloca-
by excluded frequencies can be minimized (part of Power
tion, and a connectionless, prioritized contention-based ser-
Allocation and Window and Overlap blocks in Figure 1).
vice on a Carrier Sense Multiple Access/Collision Avoidance
(iv) Power back-off to increase the HomePlug AV2 data
(CSMA/CA) allocation. The MAC receives MAC Service
rate while reducing the electromagnetic emissions (part of
Data Units (MSDUs) and encapsulates them with a header,
Power Allocation blocks and the AFEs in Figure 1). (v) EMC
optional Arrival Time Stamp (ATS), and a checksum to create
Friendly Power Boost to optimize the transmit power by
a stream of MAC frames. The stream is then divided into 512-
monitoring the input port reflection coefficient (known as the
octet segments, encrypted and encapsulated into serialized
11 parameter) at the transmitting modem. (vi) Additional
PHY Blocks (PBs), and packed as MAC Protocol Data Units
PHY Improvements, comprising higher order Quadrature
(MPDUs) to the PHY unit which then generates the final PHY
Amplitude Modulation (4096-QAM) (part of Mapper and
Protocol Data Unit (PPDU) to be transmitted onto the power
Demodulator blocks in Figure 1), higher Code Rates (8/9
line [2].
code rate) (part of Turbo Convolutional Encoder in Figure 1)
HomePlug AV2 enhances significantly the capability
and smaller guard intervals (part of Cyclic Prefix blocks in
above to accommodate new generation of multimedia appli-
Figure 1), to assist better peak data rates. At the Medium
cations with Gigabit performance.
Access Control (MAC) layer, HomePlug AV2 includes the
following. (i) Power Save mode to improve energy efficiency
when the device is in standby. (ii) Short Delimiter to reduce 2.2. The HomePlug AV2 Technology. The HomePlug system is
the overhead of the transmission shortening the Preamble specified for in-home communication adapted to the power
and Frame Control symbols. (iii) Delayed Acknowledgements line channel. The in-home communications are established
to increase the overall transmission efficiency by reducing the with Time Division Duplexing (TDD) mechanism to allow
Inter Frame Spacings. (iv) Immediate Repeating to expand the for symmetric communication between peers, as opposed
coverage by repeating the signal on paths with better Signal to the classical access system in ADSL with two different
to Noise Ratio (SNR) characteristics. downstream and upstream throughputs.
All the features listed above improve the Quality of The PHY layer employs OFDM modulation scheme for
Service of AV2 power line modems, by improving coverage better efficiency and adaptability to the channel impairments
and robustness of communication links. (such as being frequency selective and suffering from narrow-
The following sections provide technical details for all the band interference and impulsive noise). The HomePlug AV2
above mentioned features. Moreover, the AV TWG evaluated OFDM parameters correspond to a system with 4096 carriers
Tx
Insert
1.0.1 Frame Turbo Frame Preamble
Control Product Control 2-Stream MIMO Transmit Encoder
Data in Encoder Interleaver
Power Allocation
MIMO Precoding
IFFT Cyclic Prefix,
1.0.1 Frame Control Encoder Mapper (1024, Window and
8192) Overlap
AV2 Frame Turbo Frame Control MIMO IFFT Cyclic Prefix,

Control Convolutional Diversity Phase (1024, Window and
Stream Mapper Shifter
Data in Encoder Copier Parser 8192) Overlap
AV2 Frame Control Encoder
MIMO
AFE
AV2 Turbo Channel and
Payload Scrambler Convolutional ROBO
Data in Encoder Interleaver 2-Transmit
channels
AV2 Payload Data Encoder
Power line
Rx
MIMO
AFE
Receive 1.0.1 Frame Control Decoder
Channels 1024 Frame Frame Turbo 1.0.1 Frame
AGC Point Control Control Product Control
FFT Demodulator deinterleaver Decoder Data out
Time Frame Control Turbo AV2 Frame

Sync Combine Convolutional Control
Copies Decoder Data out
AV2 Frame Control Decoder
8192
Point AV2
FFT(1) Channel and Turbo
Demodulator ROBO Convolutional Descrambler Payload
deinterleaver Decoder Data out
MIMO MIMO
.. Demultiplexer
. Equalizer AV2 Payload Data Decoder
Demodulator
8192 Point
FFT( ) -input 2-Stream MIMO Decoder
Figure 1: HomePlug AV2 transmitter and receiver PHY layer.
in 100 MHz, but only carriers from 1.8 to 86.13 MHz are Besides, two OFDM paths are shown in order to imple-
supported for communication (3455 carriers). The subcarrier ment the MIMO capabilities with 2 transmission ports.
spacing of 24.414 kHz was chosen in the HomePlug AV Apart from the Hybrid or AV-only Frame Control sym-
system according to the power line coherence bandwidth bols, the payload data can be sent using adaptive bit loading
characteristic and is maintained in HomePlug AV2 for per carrier or robust modes (ROBO) with fixed Quadrature
interoperability. Phase Shift Keying (QPSK) constellation and several copies
More significantly, HomePlug AV2 incorporates MIMO of data interleaved in both time and frequency.
capability (see details in the next sections) to improve
throughput and coverage. Looking at the data path details in the block diagram
A block diagram of the PHY layer is shown in Figure 1. (Figure 1), the following can be seen: at the transmitter
The HomePlug AV2 system is capable of supporting two side, the PHY layer receives its inputs from the MAC layer.
network operating modes: AV-only mode and Hybrid mode Three separate processing chains are shown because of the
(1.0.1 Frame Control Encoder in Figure 1). Hybrid mode different encoding for HomePlug 1.0.1 Frame Control (FC)
is used for coexistence with HomePlug 1.0.1 stations. For data, HomePlug AV2 FC data, and HomePlug AV2 payload
that purpose, a 1.0.1 Frame Control Encoder is included for data. AV2 Frame Control data is processed by the AV2 Frame
Hybrid modes and an AV2 Frame Control Encoder for both Control Encoder, which has a Turbo Convolutional Encoder
Hybrid and AV-only modes. The AV-only mode is used for and Frame Control Diversity Copier, while the HomePlug
communications in networks where only HomePlug AV and AV2 payload data stream passes through a Scrambler, a Turbo
HomePlug AV2 stations are involved. Convolutional Encoder, and an Interleaver. The HomePlug
1.0.1 Frame Control data passes through a separate HomePlug L

1.0.1 Frame Control Encoder. Tx 1 1 Rx
PLC N
The outputs of the FC Encoders and Payload Encoder PLC PLC
2 3 channel 2 3
lead into a common MIMO OFDM modulation structure, modem PE modem
consisting of a MIMO Stream Parser that provides up to 4
two independent data streams to two transmit paths which Ground
include two Mappers, a phase shifter that applies a 90-
degree phase shift to one of the two streams (to reduce the Figure 2: MIMO-PLC channel: different feeding and receiving
coherent addition of the two signals), a MIMO precoder options.
to apply transmitter Beamforming operations, two Inverse
Fast Fourier Transform (IFFT) processors, preamble and 10
Cyclic Prefix insertion, and symbol Window and Overlap
blocks, which eventually feed the AFE module with one or 20
two transmit ports that couple the signal to the power line
medium. 30
At the receiver, an AFE with one, two, three, or four ( )
40
receive ports operates with individual Automatic Gain Con-
(dB)
trol (AGC) modules and one or more time-synchronization 50
21
modules to feed separate Frame Control and payload data
recovery circuits. Receivers plugged to power outlets which 60
are connected to the three wires Line, Neutral, and Protective
Earth might utilize up to three differential mode Rx ports and 70
one common mode Rx port.
80
The Frame Control data is recovered by processing the
received signals through a 1024-point FFT (for HomePlug 90
1.0.1 delimiters) and multiple 8192-point FFTs, and through 0 10 20 30 40 50 60 70 80
separate Frame Control Decoders for the HomePlug AV2/AV Frequency (MHz)
and HomePlug 1.0.1 modes. The payload portion of the sam-
L-PE L L-N L
pled time domain waveform, which contains only HomePlug
L-PE N L-N N
AV2 formatted symbols, is processed through the multiple L-PE PE L-N PE
8192-point FFT (one for each receive port), a MIMO Equal- L-PE CM L-N CM
izer that receives signals, performs receive Beamforming,
and recovers the two transmit streams, two Demodulators,
a Demultiplexer to combine the two MIMO streams, and
Figure 3: Magnitude of the transfer functions (21 , forward scatter-
a channel deinterleaver followed by a Turbo Convolutional
ing parameter) of all MIMO paths of a 2 4 MIMO configuration.
Decoder and a Descrambler to recover the AV2 payload data.
The numbers of used transmit ports and used receive

3. PHY Layer Improvements of HomePlug AV2 ports define the MIMO configuration, which is called
MIMO. For example, using L-N and L-PE to feed
3.1. Multiple-Input Multiple-Output (MIMO) Capabilities with and receive signals results in a 2 2 MIMO configuration.
Beamforming. The HomePlug AV2 specification incorpo- HomePlug AV2 specification supports MIMO configurations
rates Multiple-Input Multiple-Output (MIMO) capabilities with up to 2 Tx ports and up to Rx ports.
with Beamforming, which offers the benefit of improved Some regions and maybe homes with older electrical
coverage throughout the home, particularly for hard to reach installations do not have the third wire installed in the private
outlets. MIMO technology enables HomePlug AV2 devices buildings. In this scenario, HomePlug AV2 automatically
to transmit and receive on any two-wire pairs within a three- switches to Single-Input Single-Output (SISO) operating
wire configuration. Figure 2 shows a three-wire configuration mode. HomePlug AV2 incorporates also selection diversity in
with Line (L), Neutral (N), and Protective Earth (PE). SISO mode. The ports used for feeding and receiving might be
Whereas HomePlug AV always transmits and receives on different from the traditional L-N feeding. If, for example, the
the Line-Neutral pair (port 1 in Figure 2), HomePlug AV2 path from L-PE to L-N offers better channel characteristics
modems can transmit and receive signals on the other pairs than L-N to L-N, the transmitter can choose to use the L-PE
as well. Any two pairs formed by the Line, Neutral, or port for feeding.
Protective Earth wires (i.e., L-N, L-PE, and N-PE) can be Figure 3 shows the magnitude of several transfer func-
used at the transmitter. At the receiver, up to four receive tions of a typical MIMO-PLC channel. The figure illustrates
ports may be used according to Figure 2. The common mode all eight possible paths of a 2 4 MIMO configuration where
(CM) signalindicated as port 4 in Figure 2is the voltage L-N and L-PE pairs are used as feeding and all four receive
difference between the sum of the three wires and the ground. ports are used as receiving. The example shows that the signal
L-N L-N
1 1
VH
L-PE L-PE
2 2
D
U
N-PE 0 3
0 4
CM
(a) (b)
Figure 4: Schematic MIMO channel and decomposition into parallel and independent MIMO streams.
fed into one port is visible at all receive ports. Crosstalk Feeding: L-N and L-PE; receiving: L, N, PE, and CM
caused by the coupling of the wires results in the presence of 0
all possible MIMO paths. At most frequencies the 21 (for- 10
ward scattering parameter of the network analyzer) curves
20
are quite independent. As less correlation the 21 curves
show, as higher is the gain by MIMO. For some frequencies, 30
(dB)
for example in the region of the fading degradation around 40
21
16 MHz, the shapes of the transfer functions are similar due
50
to the same underlying topology of the different MIMO
paths. Usually, the wires are located in parallel within the 60
walls facing a similar multipath propagation. This results in 70
a higher spatial correlation of the MIMO-PLC channel than
80
compared to radio MIMO channels. 0 10 20 30 40 50 60 70 80
More information about the MIMO-PLC channel char- Frequency (MHz)
acteristics and channel modeling may be found in [618].
The channels presented in Figure 3 were recorded with a Stream 1
Stream 2
delta coupler at the transmitter (L-PE, L-N) and a star coupler
SISO: L-N P
at the receiver (L, N, PE, CM). This coupler combination
allows 2 4 MIMO [19].
The MIMO-PLC channel is described by a
channel matrix for each OFDM subcarrier : Figure 5: Attenuation of the decomposed PLC-MIMO channel of
Figure 3.
11 () 1 ()
[ .. .. .. ]
H () = [ . . . ] (1)
() () is a diagonal matrix containing the singular values of H.
[ 1 ]
Figure 4(b) illustrates this decomposition.
with belonging to the set of used subcarriers (depending, Figure 5 shows the decomposition of the two streams of
e.g., on the frequency band and tone mask). the MIMO-PLC channel shown in Figure 3. As a comparison,
Figure 4(a) shows the schematic MIMO channel for a the attenuation of the SISO channel is also depicted.
2 4 MIMO configuration with the overall 8 MIMO paths. The decomposition into parallel and independent streams
HomePlug AV2 supports 1 or 2 streams (see Section 3.1.1). As by means of the SVD illustrates the MIMO gain; instead of
HomePlug AV2 supports MIMO configurations with up to having one spatial stream in SISO, two independent spatial
2 Tx ports and up to Rx ports, the maximum number of streams are available in a 2 MIMO configuration with
supported streams in AV2 is 2. The number of the underlying 2 doubling in average the MIMO capacity [9, 12, 13, 21].
MIMO streams depends on the rank of the channel matrix. This fact is also verified by means of Figure 5; the first spatial
If the channel matrix has full rank, the number of MIMO stream shows less attenuation compared to the SISO channel;
streams is = min( , ). The underlying MIMO streams are the second stream is slightly higher attenuated compared to
obtained by a Singular Value Decomposition (SVD) of the SISO. Hence, in this example, the MIMO link would result in
channel matrix [20]: more than twice the capacity of the SISO link.
H () = U () D () V(), (2) 3.1.1. MIMO Stream Parser. Depending on how many streams
are used in the transmission, the payload bits have to be split
where V and U are unitary matrices, that is, V1 = V into different spatial streams. This task is performed by the
and U1 = U (with the Hermitian operator) and D MIMO Stream Parser (MSP). The MSP splits the incoming
bits into one or two streams based on the MIMO mode 2-stream MIMO Encoder
MIMO Pre-Coding
of operation and the Tone Map information (see Figure 1).
1 1 1 1 IFFT
To transmit a single-stream payload with Spot-Beamforming Mapper = V2 2 (1024,
MIMO 2 2 8192)
(see Section 3.1.3) or for SISO transmissions, the MSP sends Power
all the data at its input to the first Mapper. In this case, Allocation Phase 2 V 2 IFFT
Mapper Compute V (1024,
the MSP operates as if it had only one output and it was Shifter 8192)
connected to the first mapper in Figure 1. To transmit a two-
stream payload with Eigen-Beamforming (see Section 3.1.3), index, index
the MSP allocates the bits to the two streams.
3.1.2. Precoding. Precoded Spatial Multiplexing or Beam- Figure 6: Precoding at the transmitter.
forming was chosen as the MIMO scheme as it offers
the best performance by adapting the transmission in an
optimum way to the underlying Eigen modes of the MIMO-
PLC channel. The best performance is achieved for various Note that the subcarrier index is omitted in (3) and in the
channel conditions. On the one hand, the full spatial diversity following section to simplify the notation and to allow better
gain is achieved in highly attenuated and correlated channels legibility. However, it has to be kept in mind that the following
when each symbol is transmitted via each available MIMO vector and matrix operations are applied to each subcarrier
path. On the other hand, a maximum bit rate gain is achieved separately if not stated otherwise.
for channels with low attenuation when all available spatial P is a diagonal matrix that describes the power allocation
streams are utilized. Beamforming also offers flexibility with of the total transmit power to each of the transmit streams.
respect to the receiver configuration. Only one spatial stream Power allocation is considered in more detail in Section 3.1.4.
may be activated by the transmitter when only one receive V is the right hand unitary matrix of the SVD of the channel
port is available, that is, if the outlet is not equipped with the matrix (see (2)). Precoding by the unitary matrix V is often
3rd wire or if a simplified receiver implementation is used that referred to as unitary precoding or Eigen-Beamforming.
supports only one spatial stream. Since Beamforming aims Figure 6 shows the basic MIMO blocks of the transmitter.
to maximize one MIMO stream, the performance loss of not The Mappers of the two streams follow the MIMO stream
utilizing the second stream is relatively small compared to parser (see Figure 1). The two symbols of the two streams
the Spatial Multiplexing schemes without precoding. This is are then weighted by the power allocation coefficients and
especially true for highly attenuated and correlated channels, multiplied by the matrix V (the computation of V is explained
where the second MIMO stream carries only a small amount in Section 3.1.3 in more detail). Finally, each stream is OFDM
of information. These channels are most critical for PLC modulated separately.
and adequate MIMO schemes are important. A comparison
and analysis of different MIMO schemes may be found in
[14, 21, 22]. 3.1.3. Beamforming and Quantization of the Beamforming
Beamforming requires knowledge about the channel state Matrix. The precoding matrix V is derived by means of the
information at the transmitter to apply the optimum precod- SVD (see (2)) from the channel matrix H.
ing. Usually, only the receiver has channel state information, There are two possible modes of operation. If only one
for example, by channel estimation. Thus, the information spatial stream is utilized, single-stream Beamforming (or
about the precoding has to be fed back from the receiver to Spot-Beamforming) is applied. In this case, the precoding
the transmitter. The HomePlug AV2 specification supports is described by the first column vector of V; that is, the
adaptive modulation [23, 24]. The application of adaptive precoding simplifies to a column vector multiplication. Note
modulation also requires feedback about the constellation that despite the fact that only one spatial stream is used, both
of each subcarrier (Tone Map); that is, the feedback path is transmit ports are active as the precoding vector splits the
required anyway. The information about the Tone Maps and signal to two transmit ports. If both spatial streams are used,
the precoding are updated simultaneously upon a change in two-stream Beamforming (or Eigen-Beamforming) is used
the PLC channel (realized by a channel estimation indication and the full precoding matrix V is applied in the MIMO
message). The information about the precoding is quantized precoding block in Figure 6.
very efficiently and the amount required for the precoding Since the information about the precoding matrix has to
information is in the same order of magnitude as the be fed back from the receiver to the transmitter, an adequate
information about the Tone Maps. Thus, the overhead in quantization is required. To achieve this goal, the special
terms of management messages and required memory can be properties of V are utilized.
kept low. The unitary property of V consequences that the columns
The optimum linear precoding matrix F for a precoded v ( = 1, 2) of V are orthonormal; that is, the column vectors
Spatial Multiplexing system can be factored into the two are orthogonal and the norm of each column vector is equal to
matrices V and P [25]: 1 [26]. There is more than one unique solution to the SVD; the
column vectors of V are phase invariant; that is, multiplying
each column vector of V by an arbitrary phase rotation results
F = VP. (3) in another valid precoding matrix.
50 50
45 45
40 40
35 35
SNR (dB)
SNR (dB)
30 30
25 25
20 20
15 15
10 10
5 5
0 0
5 10 15 20 25 30 5 10 15 20 25 30
Frequency (MHz) Frequency (MHz)
Spatial Multiplexing Spatial Multiplexing

Beamforming Beamforming
(a) (b)
SNR1 (, , H) at 13 MHz SNR2 (, , H) at 13 MHz
8 10 12
8 14 1
4 4 12 16 16 14 2
24 2220 18
26
28
30
32
18
30 3 10 20 22 18 30
3 14
10
16 14
8 12 12
2 8 2 10
10 8 8 10
12 12 10 25 8 25
1 14 16 1
14
18 22 20
11 dB 20 22 2242368032 8 dB
12
14
18
16
0 0
18
16 20 20
14 12
10 8
1 12
10
18 8 10
8 10
8
10 12
2 8
15 2 10 14 12 15
12 16 24
26 2
2 0
2 1816 14
18 28
30
32
3 3
18
20 22
14 16 14
4 8 10 4 12 10
8
10 12
10 10
10 8
8
0 0.5 1 1.5 0 0.5 1 1.5
SNR1 ( , , H) SNR2 ( , , H)
Optimum precoding: 35 dB Optimum precoding: 6 dB
No precoding: 11 dB No precoding: 8 dB
(c) (d)
Figure 7: Influence of precoding on the SNR: SNR of the first (a) and second (b) streams with and without precoding, transmit power to
noise power of = 65 dB, 40 dB average channel attenuation; SNR depending on the precoding matrix at 13 MHz of the first (c) and second
(d) streams.
These properties allow to represent the complex 2 2 The norm of the column vectors is one: |11 |2 + |21 |2 =
matrix V by only the two angles and : |12 |2 + |22 |2 = sin2 () + cos2 () = 1. Also, the two columns
are orthogonal:
11 12 cos sin
V = [v1 v2 ] = [ ] = [ ], (4)
21 22 sin cos
v1v2 = sin () cos () sin () cos () = 0. (5)
where the range of and to represent all possible Beam-
forming matrices is 0 /2 and .
According to the phase-invariance property, the first In both modes, Spot-Beamforming and Eigen-
entry of each column (11 , 12 ) may be set to be real without Beamforming, the Beamforming vector or Beamforming
loss of generality, as defined in (4). It is easy to prove that matrix, respectively, is described by both angles and .
the properties of the unitary precoding matrix are fulfilled. Thus, the signaling of and is the same in both modes.
If the MIMO Equalizer is based on zero-forcing (ZF) way not to feed back additional information from the receiver
detection, the detection matrix to the transmitter. The power allocation evaluates the Tone
1 Maps of the two streams to set the power allocation coeffi-
W = H = (HH) H (6) cients of the two streams. It was shown that the performance
is very close to the optimum power allocation (mercury water
is the pseudo inverse HP of the channel matrix H. In case filling [27]). Power allocation improves especially highly
of Eigen-Beamforming with the precoding matrix V, H can attenuated channels at the high coverage point.
be replaced by the equivalent channel HV in (6) and the
detection matrix can be expressed by
3.1.5. Precoding Grouping. HomePlug AV2 devices support

W=V H =D U . 1
(7) an expanded frequency spectrum (up to 86.13 MHz, see
Section 3.2) and one additional stream (utilizing MIMO)
The SNR of the MIMO streams after detection is calcu- compared to the 30 MHz for HomePlug AV devices. Thus,
lated as HomePlug AV2 devices need to store two Tone Maps (one for
1 1 each stream) and one precoding matrix (PCM) per carrier.
SNR1 = 2 , SNR2 = 2 , (8) This significantly increases the memory requirements for
w1 w2
an AV2 modem. In order to save memory, HomePlug AV2
devices transmit and store the PCM only on a subset of
with being the ratio of transmit power to noise power and
carriers called precoding pilot carriers. At the transmitter,
the norm of the ith row of the detection matrix W.
the PCMs for the carriers between two adjacent precoding
Figure 7(a) shows the SNR of the first stream of a MIMO-
pilot carriers are obtained via interpolation. The spacing
PLC channel. The median attenuation of this link is 40 dB.
between the precoding pilot carriers is selected out of a
The ratio of the transmit power to noise power is = 65 dB.
set of predefined values and may be adjusted depending
The blue line represents the SNR of Spatial Multiplexing
on the memory capabilities or channel conditions. The
without precoding and the green line represents the Eigen-
more memory is embedded in the modems, the finer the
Beamforming. The SNR of the 2nd stream is displayed in
granularity is. The performance loss of precoding grouping
Figure 7(b).
compared to the quantization of each subcarrier separately
The two markers X in Figure 7(a) mark a frequency
is marginal due to the high correlation of the precoding
(13 MHz) where good Beamforming conditions are found.
matrices of neighbored subcarriers. An investigation about
Different precoding matrices influence the SNR of the two
different precoding grouping algorithms for MIMO-PLC
MIMO streams according to (6) and (8).
may be found in [14].
Figures 7(c) and 7(d) show the level of the gain or signal
elimination due to Beamforming for the frequency marked by
X in Figures 7(a) and 7(b). The color lines in Figures 7(c) and 3.2. Extended Frequency Band up to 86 MHz. During the
7(d) indicate the SNR. Depending on the precoding matrix specification development, the AV TWG realized a mea-
the SNR varies between 6 dB and 35 dB. No Beamforming surement campaign, where power line channel and noise
( = 0 and = 0) would result in a SNR of 11 dB. As measurements were performed in 30 homes located in
seen in Figure 7(c), there is one SNR maximum in the area different countries, from Europe to USA. Such variability
spanned by and . Due to both streams being orthogonal was a key ingredient to obtain insight into the power line at
one shows an SNR minimum at the location where the other frequencies above those used in HomePlug AV (1.830 MHz).
one has its maximum. The SNR plot in Figures 7(c) and 7(d) In particular, in every home, all the possible links among at
is 2 periodic in , where the black horizontal lines indicate least 5 different nodes were measured.
= . Examples of recorded channel attenuation and noise PSD
The Beamforming matrix is quantized efficiently to are reported in Figures 8 and 9, respectively.
reduce the amount of feedback required to signal V. This Based upon the results of the measurement campaign, the
quantization is chosen such that the loss of the SNR after present EMC regulations, coverage, and complexity targets,
detection (refer also to (6) and (8)) is within 0.2 dB compared the final outcome of the analysis was the selection of the 30
to optimum Beamforming without quantization. 86 MHz band for the following reasons.
(i) The FM band region 87.5108 MHz shall be avoided.

3.1.4. Power Allocation. MIMO power allocation is applied
In fact this frequency region presents higher attenu-
to two-stream MIMO transmissions. The power allocation
ation and higher noise compared to the 3086 MHz
adjusts the power of a carrier on one stream relative to
band (see also Figures 8 and 9). Since, probably,
the other stream. For SISO transmissions and MIMO Spot-
lower transmit level requirements will be needed in
Beamforming transmissions, MIMO power allocation can
order to not interfere with the FM radio service, very
be bypassed since there is only one transmit stream. In this
low operating SNRs are obtained in this band with
case, the only available option is to allocate all the power to
negligible coverage increase;
the single stream. The power allocation module is located
between the Mapper and the precoding block (see Figure 1) (ii) The 3086 MHz frequency band appears to offer a
and performs on the two MIMO streams before Eigen- throughput increase especially at the low to mid
Beamforming. The power allocation in AV2 is designed in a coverage percentages; the reason is that while the
20 0
13 dB
10
30 42 dB 32 dB 27 dB
20
Attenuation (dB)
40
30
58 dB
Magnitude (dB)
50 40 55 dB
50
60
60
70
70
80 80
2 4 6 8 10
Frequency (Hz) 107 90
100
Figure 8: HomePlug AV2 measurement campaign. Example of 0 10 20 30 40 50 60 70 80 90 100
channel measurement. Frequency (kHz)
Rectangular Hamming
100 Barlett Raised Cosine
Blackmann Kaiser
110
Figure 10: OFDM side lobes of a single carrier. Comparison of
various window functions.
120
PSD (dBm/Hz)
130
3.3. Efficient Notching. HomePlug AV2 increases throughput
140
by allowing devices to minimize the overhead incurred
due to EMC notching requirements. While in HomePlug
150 AV the mechanism (windowed OFDM) for creating the
PSD notches is fixed and relatively conservative, HomePlug
160 AV2 devices may gain up to 20% in efficiency if they
2 4 6 8 10 implement additional techniques to accommodate sharper
Frequency (Hz) 107 PSD notches. The 20% includes the gain of guard carriers
which were excluded by HomePlug AV modems and the
Figure 9: HomePlug AV2 measurement campaign. Example of
noise measurement.
reduced Transition Interval in time domain. Such devices
gain additional carriers at the band edges and may utilize
shorter cyclic extensions, which reduces the duration of the
OFDM symbols.
attenuation is greater compared to the 1.830 MHz
band, noise is lower. A concern derives from the fact 3.3.1. Influence of Windowing on Spectrum and Notch Shape.
that EMC requirements for the 3086 MHz frequency The FFT process uses a rectangular window to cut data
band are generally tighter than in the 1.830 MHz out of a continuous stream to convert them from time to
band. However, this observation could be mitigated frequency domain. The FFT of a rectangular function in the
in view of the fact that, for instance, the upcoming time domain is a sin()/ function in the frequency domain.
standard FprEN 50561-1 appears to be more restrictive The sin()/ becomes 0 at integer multiples of . Some parts
in the 1.830 MHz frequency band. of the signal remain in-between the zeros, which results in
(iii) Within the HomePlug AV2 specification it has been the unwanted side lobes of an FFT OFDM system. Figure 10
possible to provide flexibility in choosing the stop shows a sin()/ function as green curve. The frequency
frequency in the 3086 MHz interval. In particular, axis is shown horizontally. The level in a logarithmic view is
an AV2 device implementing a frequency band 1.8- presented on the vertical axis.
MHz band (with 30 < < 86) will be The process of multiplying a window with an OFDM
interoperable with a device implementing a frequency symbol (see Figure 11) in the time domain aims to suppress
band 1.8- MHz (with 30 < < 86, =/ ). the sharp corners at the beginning and the end of the OFDM
symbol in order to get smooth transitions. This affects the
(iv) The 3086 MHz frequency band extension allows shape and distances of the side lobes in the frequency domain
devices to be fully interoperable with the IEEE 1901 [28]. There are numerous types of window functions available
devices that use the 1.850 MHz frequency band. for implementation such as Hamming, Barlett (triangular),
In fact, HomePlug AV2 devices that implement a Kaiser, Blackman, Raised Cosine (Hann), and so forth. A
frequency band extension shall support at least the comparison of various window waveforms with the achieved
IEEE 1901 bandwidth. side lobe attenuation is shown in Figure 10 which also serves
Flat
guard Copy and window
interval Raised Cosine
Raised Cosine
guard interval guard interval
FFT section
Copy and window
Figure 11: OFDM symbol with GI and window.
beginning and at the end by copying the bits as done for

TI TI TI TI
the Guard Interval (GI). This expansion is multiplied with
2 2 2 2
the smoothly descending window. The more smoothly the
OFDM symbol
signal approaches zero in the time domain, the lower the side
lobes in the frequency domain. Of course, this expansion of a
I RI
2 2
symbol is a waste of communication resources as it does not
RI GI RI carry useful information and has to be cut by the receiver.
The two descending slopes in the time domain could
overlap, to save communication resources at consecutive
=0 = RI
OFDM symbols, as shown in Figure 12. The new symbol time
Figure 12: Consecutive OFDM symbols, guard interval (GI), roll-off is measured between the middles of the roll-offs before and
interval (RI), and windowing. after the symbol. The overlap region is the roll-off interval
(RI) (see in Figure 12) which is related to the OFDM symbol
10 duration via the parameter :
5
RI = s (9)
0
and the total symbol length :
Power (dB)
5
10 Single 2 = (1 + ) s . (10)
15 carrier 3
notch 4
20
5 The HomePlug AV specification allows the usage of sev-
eral Guard intervals [30, 31]. If the shortest guard interval
25 is selected, HomePlug AV uses a of RI/( + GI) =
10 carriers notch
30 4.96 /(40.96 + 5.56 ) = 0.1066.
11.5 12 12.5 13 13.5 14
With HomePlug AV2, the Transition Interval TI is intro-
Frequency (MHz)
duced. The shape of the windowing is transmitter implemen-
=0 = 0.0625 tation dependent; it does not affect interoperability. The pulse
= 0.01 = 0.121 shaping window and Guard interval might be reduced even
= 0.03125 to zero to minimize overhead in time domain and to notch
Figure 13: OFDM spectra with different notch widths and depths
efficiently. In order to guarantee backward compatibility with
achieved with different roll-off factors. previous HomePlug versions the definition and timings of
the parameter RI as shown in Figure 12 have to stay stable.
In contrast to RI, the new parameter TI might be reduced to
zero.
to illustrate the disadvantage of windowing. The side lobes are There is a balance between this intervals and obtaining a
suppressed, but the main lobe stays wider. The first time the better side lobe attenuation.
spectrum reaches zero, the distance for all windows is at least Figure 13 shows the notch of any OFDM system spectrum
twice the frequency of the rectangular window used by the which could be obtained by varying the roll-off factors
pure FFT. Windowing is considered to be state of the art in (the values) of the Raised Cosine window. Increasing the
signal processing [29]. roll-off factor directly increases the amount of attenuation
The process of how a window is applied is shown in achieved inside the notches. However, this has the drawback
Figure 11. The original OFDM symbol or the output of the of increased symbol length.
IFFT at the transmitter is marked with FFT section. As To create notches in the OFDM spectrum, a model
described above, the guard interval is copied from the tail is implemented using QAM modulation and notches of
samples to the beginning of the symbol. In order to create omitting a various number of carriers: 1, 2, 3, 4, 5, and 10. A
a window, the symbol has to be expanded further at the max-hold function is implemented in order to create these
figures. The 10 carrier notch shows the spectral benefit of affecting performance. Conversely, the second technique,
windowing. The influence of windowing is hardly visible up called EMC Friendly Power Boost is a technique that allows
to the point of the 5-carrier notch. At the 5-carrier notch, the transmitter to increase the power on some carriers with
the difference between no window and the highest simulated the knowledge that this can be done without exceeding
is around 5 dB. The trade-off for this is a 27% longer regulatory limits.
symbol time. At the 10-carrier notch, the center frequency
is suppressed by more than 15 dB without windowing, but
the adjacent side slopes of the used corner carriers are only 3.4.1. Power Back-Off. In power line communications the
improved by 5 dB. transmit power limit is typically defined as a power spectral
The additional overhead in the time domain is extremely density (PSD) mask applicable over the range of frequencies
large compared to the improved notch depth. Windowing used in the standard. And since power line modems are
with a small roll-off factor is sufficient in order to suppress directly connected to the electrical wiring, they are tradi-
the side lobes outside the used spectrum, but it is not tionally designed to transmit with the maximum allowed
recommended to increase the depth of a single carrier notch. transmit PSD on each frequency (i.e., they do not need
In the case of the HomePlug AV specification using a to be sensitive to a limited battery supply). In many cases
of 0.1066, some guard carriers on the left and right side of maximizing transmit power leads to the best performance;
a protected frequency range have to be omitted to guarantee however, certain definitions of PSD masks combined with
the depth of the notch. The North American Carrier Mask certain channel conditions can produce cases where modems
requests 10 notches in-between 1.7 MHz and 30 MHz. The can benefit from transmitting at less than the maximum
first carriers as well the last carriers of this spectrum are allowed power level.
notched. These two notches have only one slope to the carriers We illustrate the benefits of transmit power control using
allocated for communications. All other notches have two the North American regulatory limits as an example. FCC
slopes resulting in 18 notch slopes in total. The spectrum loses regulations that are applicable to power line devices in
almost 6% of communication resources in frequency domain. North America are commonly interpreted to allow a transmit
Additionally the causes almost 13% of wasted resources PSD of 50 dBm/Hz from 1.8 to 30 MHz, and 80 dBm/Hz
in time domain. Assuming an ideal implementation with from 30 MHz up to 86 MHz. This 30 dB drop in the PSD
maximal sharpness in time as well as frequency domain, it (at 30 MHz) causes the signal from the higher frequency
would be possible to regain these communication resources carriers (above 30 MHz) to have much smaller amplitude
when applying the North American Carrier Mask. If the than the signal for the low band carriers (up to 30 MHz).
frequency spectrum becomes more fragmented because of Consequently, when the overall signal is represented in a
additional notches like requested by the newly upcoming quantized digital domain, the high band signal has lower
European regulations [32] these losses become even more resolution than the signal in the lower band and will therefore
obvious. also have a limited SNR. This will be evident at the transmitter
where the 30 dB drop will result in a reduction of 5 bits of
resolution for the high band signal. If the transmit power
3.3.2. Digital, Adaptive Band Stop Filters Improve the Notchs is backed off in the low band, so that the PSD drop is
Depth and Slopes. Of course an ideal implementation as reduced, then the high band signal will be represented with
described above is not possible, but digital band stop filters an increased number of bits of resolution and enjoy increased
increase the sharpness of the notches as well the imple- SNRs out of the transmitter.
mentation efforts in hardware. Shrinking the semiconductor
Another limiting factor is the limited dynamic range
manufacturing process to smaller structures, allowing to
of the analog to digital converter (ADC). We illustrate its
integrate additional functions on the same die size, shifts the
impact using once again the example of the North American
balance towards hardware implementation efforts.
PSD limits. For the sake of simplicity, we assume a flat
HomePlug AV2 specification gives maximum freedom to
power line channel, and flat noise spectrum in Figure 14. For
the chip implementer. Filter algorithm, order, and structure
convenience, the power line channel (cyan curves labeled
are implementation dependent. An example is documented
Rx signal: the transmitted signal after channel attenuation)
in [19]. As higher the filtering efforts, as better is the sharpness
and noise (green curves labeled Rx noise) contributions
of the notch slopes, as shorter might be GI, as higher the
to the received signal are shown separately. Moreover, the
resources available for communication. A reduction of the
different signals are shown before (dashed line curves) and
Guard Interval down to zero (see Section 3.5.1) is possible
after (continuous line curves) the analog amplifier. On the left
at short PLC channels without multipath reflections or
of Figure 14 the scenario without power back-off is shown;
intersymbol interference.
the transmitted signal is 30 dB greater in the 1.830 MHz
band compared to the 3086 MHz band. The analog amplifier
3.4. HomePlug AV2 Power Optimization Techniques. The brings the received signal to a level A tailored to optimize
HomePlug AV2 standard introduces two novel techniques the ADC conversion. Since the dominating noise is the ADC
that can be used to optimize the use of transmit power. The noise (black curve), after the ADC converter SNRs of 35 dB
first one, named transmit power back-off , is a technique that and 5 dB are found below and above 30 MHz, respectively.
reduces the transmitted power spectral density for a selected On the right of Figure 14 the scenario with power back-
set of carriers when this can be done without adversely off is shown: the transmitted signal is reduced by 10 dB
A A
SNR = 35 dB
SNR = 30 dB
A 10 A 10
A 20 A 20
SNR = 10 dB
A 30 A 30
SNR = 5 dB
PSD (dBm/Hz)
PSD (dBm/Hz)
A 40 A 40
A 50 A 50
A 60 A 60
A 70 A 70
A 80 A 80
A 90 A 90
A 100 A 100
0 2 4 6 8 10 0 2 4 6 8 10
Frequency (Hz) 107 Frequency (Hz) 107
Rx signal (after analog amplifier) Rx signal (after analog amplifier)
Rx noise (after analog amplifier) Rx noise (after analog amplifier)
Rx signal (before analog amplifier) Rx signal (before analog amplifier)
Rx noise (before analog amplifier) Rx noise (before analog amplifier)
ADC noise ADC noise
(a) (b)
Figure 14: Benefits of power back-off. (a) No power back-off, (b) 10 dB power back-off.
in the 1.830 MHz frequency band so that it is only 20 dB 11 parameter. PSD limits as proposed in the HomePlug
greater compared to the 1.886 MHz band. Again, the analog AV2 specification are based on representative statistics of
amplifier brings the received signal to the A level. However the impedance match at the interface between the device
in this case, the dominating noise is no longer the ADC noise port and the power line network. In practice, the input
and the obtained SNRs are 30 dB and 10 dB in the 1.830 MHz impedance of the power line network is frequency selective
and 3086 MHz, respectively. and varies for different network configurations. This leads
In this example, the power back-off technique results in to part of the transmit power being dissipated within the
a 5 dB SNR reduction for the carriers in the lower frequency transmitter. The input port reflection coefficient, or input
band, and a 5 dB SNR increase for the carriers in the upper return loss, is characterized by the -parameter 11 , and
frequency band. Given the larger bandwidth of the upper the part of transmit power dissipated at the transmitter is
frequency band there will be an overall throughput gain due given by the amount 20 log 10(|11 |) in dB. In situations
to transmit power back-off. where the impedance mismatch is large (and thus the 11
Transmit power back-off is also an effective interference parameter is large), only a small part of the input power
mitigation technique. For instance, in Europe, the ability of is effectively transferred to the power lines. In those cases,
a PLC transmitter to reduce the transmit power depending the Electromagnetic Interference (EMI) induced by the PLC
on the attenuation link is a possible requirement considered modem is reduced and can be much lower than the values
in [32], though the procedure described in [32] does not recommended by EMC regulation limits.
consider Quality of Service (QoS) requirements of PLC
modems. In order to compensate for the frequency selective
impedance mismatch at the interface between the device
port and the power line network, HomePlug AV2 modems
3.4.2. EMC Friendly Power Boost Using 11 Parameter. The adapt their transmission mask, upon measurement of the 11
EMC Friendly Power Boost is a mechanism introduced in parameter at the transmitter. The transmitted signal power is
the HomePlug AV2 specification to optimize the transmit increased by an Impedance Mismatch Compensation (IMC)
power by monitoring the input port reflection coefficient at factor, leading a more effective power transmission to the
the transmitting modem. This coefficient is known as the power line medium. While increasing the Tx power leads to
an increase of the radiated EMI, the design of the IMC factor 1

ensures that the resulting EMI continuously falls below the
targeted EMC regulation limits. 0.8
A statistical analysis was conducted on the practical
11
values of the 11 parameter and the effectiveness of the 0.6
CDF of
EMC Friendly Power Boost technique based on a series
of measurements performed by the ETSI Specialist Task 0.4
Force 410 [610, 33]. 11 parameter and EMI measurements
0.2
were taken over the 1 MHz100 MHz range in 6 countries:
Germany, Switzerland, Belgium, UK, France, and Spain. The
0
modem used for measurements is described in [68]. 70 60 50 40 30 20 10 0
The 11 measurements considered in this study consist of 11 (dB)
3 differential feeding possibilities (L-N, N-PE, and PE-L), and
1 common mode feeding (CM). For the EMI measurements, Figure 15: CDF of 11 parameter.
we consider the measurements taken with different feeding
possibilities, that is, differential feeding with the other differ- 1
ential ports being unterminated or terminated with 50 Ohm,
and CM feeding. 0.8
The measurement set used in this analysis consists of 478
CDF of IMC
frequency sweeps, that can be categorized as follows: 0.6
(i) 6 different locations in Germany and 3 different 0.4

locations in France.
(ii) 264 measurements outdoor at 10 m, 43 measurements 0.2
outdoor at 3 m, and 171 measurements indoors.
0
0 2 4 6 8 10 12 14 16
Statistical analysis of this experimental data allowed IMC (dB)
designing the practical implementation of the EMC Friendly
Power Boost technique. In the following, we define the Figure 16: CDF of IMC parameter.
Impedance Mismatch Compensation (IMC) factor as (in dB)
1 (ii) For 10% of the records, the Tx power could be

IMC () = min (max (10 log10 ( 2 )
1 11 () increased by more than 4 dB to compensate for the
(11) impedance mismatch.
() , 0) , IMCmax ) , We then focused on the effect of the EMC Friendly Power

Boost technique on the radiated EMI statistics. The recorded
values allowed the computation of two statistics:
where 11 () is the carrier dependent estimate of the 11
parameter, is a margin accounting for possible estimate (i) the statistical CDF of the recorded EMI in terms of
uncertainties in the measurement of parameter 11 , and E-field for all frequencies and feeding possibilities
IMCmax is the maximum allowed value of the IMC factor in without applying any power boost,
dB. (ii) the statistical CDF of the recorded EMI in terms of E-
Figures 15 and 16 represent the statistical CDF of the field for all frequencies and feeding possibilities when
11 parameter as well as the corresponding IMC0 factor, applying the EMC Friendly Power Boost.
computed with a margin M of 0 dB. These statistics are based
on the experimental data collected during the ETSI STF 410 Figure 17 presents, for each percentile of the CDF, the
measurement campaign. difference in dB between the E-Field CDFs for the two
On Figure 15, one can read that the 11 parameter is larger methods of transmission.
than 10 dB for 80% of the cases. This means that for 80% Different observations can be made from this figure. First,
of the records, more than 10% of the energy is reflected back the application of the EMC Friendly Power Boost leads to an
towards the transmitter. increase of the radiated field CDF comprised between 0 and
More interestingly, Figure 16 gives an idea of the potential 6 dB. Note that the extreme value of 6 dB arises for one of
power increase offered by the EMC Friendly Power Boost the lowest values of radiated field, and, hence, is not relevant.
technique. Secondly, in general, the application of the EMC Friendly
Power Boost increases the radiated power CDF by about 2 dB.
(i) For 40% of the records, the Tx power could be More importantly, the increase of the radiated power CDF
increased by more than 2 dB to compensate for the is lower than 2 dB for the 25% most radiating cases. This
impedance mismatch. practically means that in the worst case scenarios where the
1 (i) The Transition Interval defines the part of the Roll-

High EMI off Interval dedicated to the transition window, allow-
0.9
values ing more flexibility in the choice of the window
0.8
(see Section 3.3)
0.7
(ii) A new Guard Interval has been defined for the
0.6
HomePlug AV2 Short Delimiter (see Section 4.2.1)
Percentile
0.5
(iii) The payload symbol Guard Interval has been made
0.4 variable and can be as short as 0 s. It can also be
0.3 increased up to 19.56 s. This allows adaptation to a
0.2 wide range of channel conditions and removes the
0.1
Low EMI overhead of the Guard Interval for channels that
values have either very low multipath dispersion, or that are
0 completely limited by the receive noise and not by ISI.
1 1.5 2 2.5 3 3.5 4 4.5 5
Difference between CDF ( -field + IMC0) and CDF ( -field) (dB)
3.5.2. Additional Constellations. In HomePlug AV, the max-
Figure 17: Difference in dB between the CDFs of the radiated EMI imum constellation size is 1024-QAM, corresponding to 10
before and after applying the EMC Friendly Power Boost.
coded bits per carrier. HomePlug AV2 also provides support
for 4096-QAM, which corresponds to 12 bits per carrier. The
higher constellation size increases the peak PHY rates by
modems produce the largest EMI, the application of the IMC 20%. Practically, the increased throughput will be available
factor does not increase the EMI by more than 2 dB. This mostly on average to very good channels, but even some of the
value can be compared with the CDF of the IMC factor given poorer channels sometimes have frequency bands in which
in Figure 16. Although the IMC factor is larger than 2 dB in high SNRs can be achieved, and the increased constellation
40% of the cases and larger than 4 dB in 10% of the cases, size can be used.
the application of the EMC Friendly Power Boost techniques
does not increase the radiated power CDF extreme values 3.5.3. Forward Error Correction (FEC) Coding. HomePlug
by more than 2 dB. Of course, even an increase of the EMI AV2 uses the same duobinary Turbo Code as HomePlug AV.
by 2 dB is not acceptable. Therefore, a margin M of 2 dB is In addition to the code rates of 1/2 and 16/21, HomePlug
applied when increasing the Tx power using the IMC factor. AV2 also provides support for a 16/18 code rate. This allows
Based on this study, we conclude that the application more granularity in the compromise between robustness
of the EMC Friendly Power Boost technique provides a and throughput degradation. For this new code rate, a new
significant gain in terms of transmit power increase for a large puncturing structure is defined, as well as a new channel
number of configurations, where the impedance mismatch interleaver. In addition, a new Physical Block size of 32
causes the dissipation of the signal at the transmitter. In octets is defined, which includes specification of a new
addition, the statistical analysis shows that this technique termination matrix for the FEC as well as a new interleaver
will not lead to an increase of the undesirable radiated seed table. The 32-byte octet PBs are used in the PHY level
interference, in particular in the worst EMI scenarios, as long acknowledgements and allow for the acknowledgement of
as a margin M of 2 dB is used in the computation of the IMC much larger packet sizes that are supported with the increased
factor, as specified in (11). Finally, a recommended limit for PHY rates possible in HomePlug AV2.
the maximum allowed value of the IMC factor is IMCmax =
6 dB. 3.5.4. Line Cycle Synchronization. The HomePlug AV2 spec-
ification describes also the device operation in scenarios
3.5. Additional PHY Improvements. In addition to the MIMO where there is no alternating current (AC) line cycle (e.g.,
technology, the frequency band extension, the Efficient a direct current (DC) power line) or when the AC line
Notching, and power optimization techniques such as the cycle is different from 50 Hz or 60 Hz. In this case, the
power back-off and the EMC friendly power boost, other Central Coordinator is preconfigured to select a Beacon
elements of the PHY layer were modified as presented in the Period matching either 50 Hz (i.e., Beacon Period is 40 msec)
paragraphs below. or 60 Hz (i.e., Beacon Period is 33.33 msec). One key use case
where this feature is useful is the transfer of data towards a
3.5.1. New Time Domain Parameters. In the HomePlug AV2 multimedia equipped electrical vehicle during the electrical
specification, a number of time domain parameters were charging phase (using DC power).
refined. As the sampling frequency has increased from
75 MHz to 200 MHz, the number of time samples for a given 4. MAC Layer Improvements of
symbol duration is increased by a factor 8/3; the IFFT interval HomePlug AV2
is 8192 samples in length, and the number of samples in the
HomePlug AV Guard Interval has increased accordingly. In 4.1. Power Save Modes. HomePlug AV2 stations improve
addition, new features have been added: their energy efficiency in standby mode through the adoption
of the specific Power Save Mode already defined in the Beacon period (BP)
HomePlug Green PHY [4] specification. In Power Save Awake Start of first
Window CSMA allocation
mode stations reduce their average power consumption by in beacon period
periodically transitioning between Awake and Sleep states.
Stations in the Awake state can transmit and receive packets Station A
over the power line. In contrast, stations in Sleep state Time
temporarily suspend transmission and reception of packets Station B
over the power line. Time
We introduce some basic terms useful to describe the
Power Save Mode. Station C
Time
(i) Awake Window: period of time during which the sta-
tion is capable of transmitting and receiving frames. Station D
The Awake Window has a range from a few millisec- Time
BPCnt = 0 010
BPCnt = 0 011
BPCnt = 0 012
BPCnt = 0 013
BPCnt = 0 014
onds to several Beacon Periods (a Beacon Period is
two times the AC line cycle: 40 ms for a 50 Hz AC line
and 33.3 ms for a 60 Hz AC).
(ii) Sleep Window: period of time during which the
station is not capable of transmitting or receiving
frames. Figure 18: Example of Power Save operation in HomePlug AV2 and
HomePlug Green PHY.
(iii) Power Save Period (PSP): interval from the beginning
of one Awake Window to the beginning of the next
Awake Window. Power Save Period is restricted to 2
multiples of Beacon Periods (i.e., 1 Beacon Period, 2 OFDM symbol 1 OFDM symbol 2 OFDM symbol N
Beacon Periods, 4 Beacon Periods, . . .).
(iv) Power Save Schedule (PSS): the combination of the

values of the PSP and of the Awake Widow duration.
To communicate with a station in Power Save mode,
other stations in the logical network (AVLN) need to
know its PSS.
Potentially, the specification allows aggressive PSSs con- Preamble Carrier
stituted by an Awake Window duration of 1.5 ms and a PSP Frame Control Carrier
of 1024 Beacon Periods, that will cause over 99% energy
Data Carrier
saving compared to HomePlug AV. In practice, some in-home
applications will require lower latency and response time, Figure 19: Short Delimiter.
and a balance will take place reducing the mentioned gain.
This is particular appealing for applications that foresee a
PLC utilization that is variable during the day (for instance
large utilization in daylight time and small utilization during structured the protocol insuring that at least one superposi-
night period). It is worth highlighting that the HomePlug tion of all the Awake Windows occurs). This overlap interval
AV2 specification is flexible in allowing each station in a can also be used for transmission of information that needs
network to have a different PSS. Given these remarks, in order to be received by all stations within the AVLN.
to enable efficient Power Save operation without causing Figure 18 shows an example of Power Save Schedule of
difficulties to regular communication, all the stations in a the four stations {A, B, C, and D}. All stations save more
network need to know the PSSs of the other stations. The than 75% energy compared to HomePlug AV. Note that in
network Central Coordinator (CCo) has a key role since it this example, stations A and B can communicate once every
grants the requests of the different stations to enter and exit 2 Beacon Periods. Moreover, all the stations are always awake
from Power Save mode operation. Moreover, it distributes the at the same time once every 4 Beacon Periods thus preserving
different PSSs to all the stations in the network. When needed communication possibility.
a CCo can
(i) optionally disable Power Save mode for all stations of 4.2. Short Delimiter and Delayed Acknowledgement. The
the AVLN, Short Delimiter and Delayed Acknowledgement features
(ii) optionally wake up a station in Power Save mode. were added to HomePlug AV2 to improve efficiency by
reducing the overhead associated with transmitting payloads
The shared knowledge of the PSS allows stations commu- over the power line channel. In HomePlug AV, this overhead
nicating during the common Awake Windows (the Home- results in relatively poor efficiency for transmission control
Plug AV2 and HomePlug Green PHY specifications have protocol (TCP) payloads. One goal that was achieved with
Long MPDU using a AV1.1 delimiter AV1.1

acknowledgement
PRS
AV1.1 AV1.1 Payload AV1.1

delimiter Delimiter delimiter
RIFS RIFS CIFS

AV1.1 Payload SD
delimiter
Long MPDU using a AV1.1 delimiter Short Delimiter based

acknowledgement
Figure 20: Short delimiter efficiency improvement.
Acknowledge segments that end

Acknowledge segments in all but the
in the last data symbol in the
last data symbol in the immediately next SACK transmission
following SACK transmission
RIFS RIFS
Preamble + Data (2nd) SACK Preamble + Data (2nd)

FC + data Data (last) FC + data Data (last) SACK
Figure 21: Delayed Acknowledgement.
these new features was TCP efficiency that improved to be the length of the delimiter reduced from 110.5 to 55.5 s, the
relatively close to that of UDP. Response Inter-Frame Space (RIFS) and Contention Inter-
In order to send a packet carrying payload data over a Frame Space (CIFS) can also be reduced to 5 s and 10 s,
noisy channel, signaling is required for a receiver to detect respectively. Reduction of RIFS requires Delayed Acknowl-
the beginning of the packet and to estimate the channel so edgement, which is described in Section 4.2.2. Backward
that the payload can be decoded, and additional signaling is compatibility when contending with HomePlug AV devices
needed to acknowledge the payload was received successfully. is maintained by indicating the same length field for virtual
Interframe spaces are also required between the payload carrier sense in both cases so that the position of the priority
transmission and the acknowledgement for the processing resolution symbols (PRS) contention remains the same. A
time at the receiver to decode and check the payload for field in the Frame Control of the Long MPDU indicates the
accurate reception and to encode the acknowledgement. This Short Delimiter format to a HomePlug AV2 device so that it
overhead is even more significant for TCP payload since the can correctly determine the length of the payload.
TCP acknowledgement payload must be transmitted in the
reverse direction. 4.2.2. Delayed Acknowledgement. The processing time to
decode the last OFDM symbol and encode the acknowl-
4.2.1. Short Delimiter. The delimiter specified in HomePlug edgement can be quite high, thus requiring a rather large
AV contains the Preamble and Frame Control symbols and is Response Inter-Frame Space (RIFS). In HomePlug AV, since
used for the beginning of data PPDUs as well as for imme- the preamble is a fixed signal, the preamble portions of
diate acknowledgements. The length of the HomePlug AV the acknowledgement can be transmitted while the receiver
delimiter is 110.5 s and can represent a significant amount is still decoding the last OFDM symbol and encoding the
of overhead for each channel access. A new single OFDM payload for the acknowledgement. With the Short Delimiter,
symbol delimiter is specified in HomePlug AV2 to reduce the the preamble is encoded in the same OFDM symbol as the
overhead associated with delimiters by reducing the length Frame Control for the acknowledgement, so the RIFS would
to 55.5 s. Figure 19 shows that every fourth carrier in the need to be larger than for HomePlug AV, eliminating much
first OFDM symbol is assigned as a Preamble Carrier, and the of the gain the Short Delimiter provides.
remaining carriers encode the Frame Control. The following Delayed Acknowledgement solves this problem by
OFDM symbols encode data the same as in HomePlug AV. acknowledging the segments ending in the last OFDM
Figure 20 demonstrates the efficiency improvement when symbol in the acknowledgement transmission of the
the HomePlug AV2 Short Delimiter is used for the acknowl- next PPDU, as shown in Figure 21. This permits practical
edgement of a CSMA Long MPDU (MAC Protocol Data implementations with a very small RIFS, reducing the RIFS
Unit) compared to the HomePlug AV delimiter. Not only is overhead close to zero. HomePlug AV2 also allows the
B acknowledges The TDM synchronization period for the in-home and

SOF indicates
MPDU burst
directly to A Normal channel access systems is defined with the parameter in Figure 23.
contention There are four ISP windows within a single period.
PRSCRS Each ISP window is further divided into three TDM units
SOF Payload (AR) SOF Payload (RB) SACK
(TDMU), so there are a total of twelve TDMUs in each
Segments from A are retransmitted by R to B
period, labeled TDMU#0 through TDMU#11. Each TDMU
Bad segments are replaced with dummy segments is further divided into eight TDM slots (TDMS), labeled
TDMS#0 through TDMS#7.
Figure 22: Immediate repeating channel access for CSMA. Figure 23 illustrates the TDM partitioning relative to the
AC line cycles. The ISP window is used to generate and detect
the ISP signal that is allocated within the first TDMS#0 in
option of delaying acknowledgement for segments ending TMDU#0, TMDU#3, TMDU#6, and TDMU#9. The purpose
in the second to last OFDM symbol to provide flexibility for and nature of the ISP signals will be described later. Note:
implementers. the term Beacon Period in the HomePlug AV2 specification
is similar to the 2 AC line cycles of the TDMU.
4.3. Immediate Repeating. HomePlug AV2 supports repeat-
ing and routing of traffic to not only handle hidden nodes 5.2. ISP Signals. Coexistence signaling is carried out by
but also to improve coverage (i.e., performance on the worst the use of periodically repeating ISP signals within the
channels). ISP windows. The signals are used to convey information
With HomePlug AV2, hidden nodes are extremely rare. on coexisting system presence, resource requirements, and
However, some links may not support the data rate required resynchronization request. Each PLC system category is
for some applications such as a 3D HD video stream. In a allocated a particular ISP window in a round-robin fashion,
network where there are multiple HomePlug AV2 devices, as illustrated in Figure 23.
the connection through a repeater typically provides a higher The ISP signal is transmitted using a range of designated
data rate than the direct path for the poorest 5% of channels. phases that convey a range of information to be used by the
Immediate Repeating is a new feature in HomePlug AV2 system. This set of instantaneous information is termed the
that enables high efficient repeating. Immediate Repeating Network Status that defines the allocation of resources to each
provides a mechanism to use a repeater with a single chan- coexisting system.
nel access, and the acknowledgement does not involve the The ISP signal is merely detected. By monitoring the ISP
repeater. As shown in Figure 22, station A transmits to the signal transmitted within the ISP windows allocated to other
repeater R. In the same channel access, repeater R transmits systems, a coexisting system is able to determine the number
all payload received from station A to station B. B sends an and type of coexisting systems present on the line and their
acknowledgement directly to A. With this approach, latency resource requirements. Similarly, by monitoring the signal
is actually reduced with repeating, assuming the resulting within its own ISP window, a coexisting system is able to
data rate is higher, the obvious criteria for using repeating detect a resynchronization request from one of the other
in the first place. Also, resources required by the repeater coexisting systems.
are minimized since the repeater uses and immediately frees The ISP signal consists of 16 consecutive OFDM sym-
memory it would require for receiving payload destined for bols. Each OFDM symbol is formed by a set of all-one
it. Also the receiver has no retransmission responsibility for binary phase shift keying (BPSK), modulated onto the carrier
failed segments. waveforms using IFFT, and multiplied by a window function
to reduce out-of-band energy complying with the transmit
spectrum requirement. Since all devices send the signal
5. Coexistence with Other PLC Technologies simultaneously, the ISP signal must be sent with 8 dB less
power than the normal transmission.
5.1. Inter System Protocol. Intersystem Protocol (ISP) allows
Timing parameter that is used for generating an ISP signal
coexistence between noninteroperable devices sharing the
is described in Figure 24.
same power line medium. Using the current ISP protocol,
noninteroperable devices are able to coexist. The HomePlug
AV2 will operate in a 1901-FFT mode [5] in order to coexist 5.3. Startup and Resynchronization Procedures. The TDM
with the other systems, in this case the TDM slot allocated for synchronization scheme mentioned above is used such that
the 1901-FFT is to be used by the HomePlug AV2 system. each PLC system shares the medium without interfering with
The ISP protocol allows a TDM scheme to be imple- one another. However, it is possible that two or more systems
mented between coexisting in-home systems and between are synchronized to two or more different, mutually visible
coexisting in-home and access systems. Each of the PLC ISP sequences [35]. In such cases, in order to prevent mutual
system categories is allocated a particular ISP window in a interference, it is important that they resynchronize to the
round-robin fashion. The allocation is determined by (1) the same ISP sequence.
number of systems on the power line, (2) the type of the In other words, whenever a HomePlug AV2 device starts
systems present, and (3) the access system bandwidth request, up or restarts, it needs to be aware of the presence of any other
as defined in the 1901 standard [34]. systems with which it is able to coexist. Accordingly, a startup
ACC IH-W IH-O IH-G
ISP
TH
ISP window
TMDU#0 TDMU#1 TDMU#2
TDMS TDMS TDMS TDMS TDMS TDMS TDMS TDMS

#0 #1 #2 #3 #4 #5 #6 #7
1 TDMU = 2 AC cycles
Figure 23: Time Division Multiplexing on ISP.
Table 1: Improvement of HomePlug AV2 in a 2-node network.
Percentage of throughput
Coverage based on UDP
improvement of HomePlug AV2
throughput
compared to HomePlug AV
95% >136%
16 OFDM symbols
5% >220%
Table 2: Improvement of HomePlug AV2 in a 4-node network.

Timing
Parameters Time ( s) Percentage of throughput
Coverage based on UDP
improvement of HomePlug AV2
throughput
2 compared to HomePlug AV
99% >131%
Windowing duration 10.24
5% >173%
81.92
Figure 24: ISP Signal Timing Parameter.

than 136% compared to HomePlug AV (which means a
performance enhancement by a factor nearly equal to 2.4!).
Benefits are even higher when considering the most favorable
and resynchronization procedure is defined within the ISP connections (see the improvement at the 5% coverage value).
protocol that allows the starting system to synchronize with Table 2 considers a 4-node scenario where 3 streams carrying
an existing system. different data are transmitted from one source (e.g., a Set-
Top Box) to 3 different destinations (e.g., TVs). In this case,
6. Gain of HomePlug AV2 Compared to the improvement in the aggregate throughput is relevant for
HomePlug AV the 99% of networks compared to HomePlug AV (more than
131%).
The AV TWG has evaluated the performance of the Home- Note that the benefits of the HomePlug AV2 technology
Plug AV2 specification; this activity has been fundamental in are expected to be greater (which explains the > symbol)
order to see if the produced specification meets the require- than the ones shown in Tables 1 and 2 since, for instance, a
ments of all the stakeholders. The following tables show the 2 2 MIMO was tested in Florida; 2 3 or 2 4 MIMO would
performance improvement as compared to HomePlug AV in likely provide better performance.
terms of coverage. These preliminary results are based on a Another interesting figure is the theoretical maximum
6-home field test in Florida, home sizes 19003300 sq. ft. PHY throughput for the system, for different options of the
Table 1 presents the results in a 2-node network scenario; standard (Table 3). This number represents the throughput
95% of nodes experiment a throughput improvement greater of transmitted bits on the PHY layer for optimum channel
Table 3: Maximum PHY rate computation. [8] ETSI TR 101 562, PowerLine Telecommunications (PLT);
MIMO PLT; Part 3: Setup and Statistical Results of MIMO
System configuration PLT Channel and Noise Measurements, http://www.etsi.org/
Max PHY rate
(North American tone mask) deliver/etsi tr/101500 101599/10156203/01.01.01 60/tr 10156203
HomePlug AV (1.830 MHz) v010101p.pdf, 2012.
197 Mbps
(917 carriers, 10 bits/carr, 5.56 s GI) [9] D. Schneider, A. Schwager, W. Baschlin, and P. Pagani, Euro-
IEEE 1901 (1.850 MHz) pean MIMO PLC field measurements: channel analysis, in
556 Mbps
(1974 carriers, 12 bits/carr, 1.6 s GI) Proceedings of the IEEE International Symposium on Power
HomePlug AV2 SISO (1.886.13 MHz) Line Communications and its Applications (ISPLC 12), Beijing,
1012 Mbps
(3455 carriers, 12 bits/carr, 0.0 s GI) China, March 2012.
HomePlug AV2 MIMO (1.886.13 MHz) [10] P. Pagani, R. Hashmat, A. Schwager, D. Schneider, and W.
2024 Mbps
(3455 carriers, 12 bits/carr, 0.0 s GI, 2 streams) Baschlin, European MIMO PLC field measurements: noise
analysis, in Proceedings of the IEEE International Symposium
on Power Line Communications and its Applications (ISPLC 12),
Beijing, China, March 2012.
conditions and gives an idea of the benefits of different
features. It can be seen that if the full frequency range is [11] D. Veronesi, R. Riva, P. Bisaglia et al., Characterization of in-
used, HomePlug AV2 provides a 1 Gbps throughput in SISO home MIMO power line channels, in Proceedings of the IEEE
International Symposium on Power Line Communications and
configuration and 2 Gbps in a MIMO configuration.
Its Applications (ISPLC 11), pp. 4247, Udine, Italy, April 2011.
[12] D. Rende, A. Nayagam, K. Afkhamie et al., Noise correlation
7. Conclusion and its effect on capacity of inhome MIMO power line chan-
nels, in Proceedings of the IEEE International Symposium on
In this paper, an overview of HomePlug AV2 has been Power Line Communications and Its Applications (ISPLC 11), pp.
presented. The overall system architecture and the key tech- 6065, Udine, Italy, April 2011.
nical HomePlug AV2 improvements introduced at PHY and [13] R. Hashmat, P. Pagani, A. Zeddam, and T. Chonavel, MIMO
MAC layers have been described. It has been also shown communications for inhome PLC Networks: measurements
the related performance improvements were achieved by and results up to 100 MHz, in Proceedings of the 14th IEEE
HomePlug AV2 while ensuring both backward compatibility International Symposium on Power Line Communications and
versus HomePlug AV and the coexistence with other power its Applications (ISPLC 10), pp. 120124, Rio de Janeiro, Brazil,
line technologies. March 2010.
The HomePlug AV2 performance presented in this work [14] D. Schneider, Inhome power line communications using multiple
has been assessed by AV TWG through simulations based on input multiple output principles [Doctoral thesis], University of
field measurements. Stuttgart, Stuttgart, Germany, 2012.
The results show the significant benefits introduced by [15] R. Hashmat, P. Pagani, A. Zeddam, and T. Chonavel, A channel
the new set of HomePlug AV2 features, both in terms of model for multiple input multiple output in-home power line
achievable data rate and coverage. networks, in Proceedings of the IEEE International Symposium
on Power Line Communications and Its Applications (ISPLC 11),
pp. 3541, Udine, Italy, April 2011.
References [16] R. Hashmat, P. Pagani, A. Zeddam, and T. Chonavel, Analysis
and modeling of background noise for inhome MIMO PLC
[1] HomePlug Alliance, http://www.homeplug.org/. channels, in Proceedings of the 14th IEEE International Sympo-
[2] HomePlug Alliance, HomePlug AV White Paper, http://www sium on Power Line Communications and its Applications (ISPLC
.homeplug.org/. 12), pp. 120124, Beijing, China, March 2012.
[3] HomePlug Alliance, HomePlug AV Specification Version 2.0, [17] F. Versolatto and A. M. Tonello, A MIMO PLC random channel
January 2012. generator and capacity analysis, in Proceedings of the IEEE
[4] HomePlug Alliance, HomePlug GreenPHY, the Standard International Symposium on Power Line Communications and
for In-Home Smart Grid Powerline Communications, Its Applications (ISPLC 11), pp. 6671, Udine, Italy, April 2011.
http://www.homeplug.org/. [18] R. Hashmat, P. Pagani, T. Chonavel, and A. Zeddam, A time
[5] IEEE Standard 1901-2010, IEEE Standard for Broadband over domain model of background noise for In-home MIMO PLC
Power Line Networks: Medium Access Control and Physical networks, IEEE Transactions on Power Delivery, vol. 27, no. 4,
Layer Specifications, http://standards.ieee.org/findstds/stand- pp. 20822089, 2012.
ard/1901-2010.html, 2010. [19] A. Schwager, Powerline communications: significant technologies
[6] ETSI TR 101 562, PowerLine Telecommunications (PLT); 19 to become ready for integration [Doctoral thesis], University of
MIMO PLT; Part 1: Measurement Methods of MIMO PLT, Duisburg-Essen, Essen, Germany, 2010.
http://www.etsi.org/deliver/etsi tr/101500 101599/10156201/ [20] A. Paulraj, R. Nabar, and D. Gore, Introduction To Space-Time
01.03.01 60/tr 10156201v010301p.pdf, 2012. Wireless Communications, Cambridge University Press, New
[7] ETSI TR 101 562, PowerLine Telecommunications (PLT); York, NY, USA, 2003.
MIMO PLT; Part 2: Setup and Statistical Results of MIMO [21] L. Stadelmeier, D. Schneider, D. Schill, A. Schwager, and J.
PLT EMI Measurements, http://www.etsi.org/deliver/etsi tr/ Speidel, MIMO for inhome power line communications, in
101500 101599/10156202/01.02.01 60/tr 10156202v010201p.pdf, Proceedings of the 7th International ITG Conference on Source
2012. and Channel Coding (SCC 08), Ulm, Germany, January 2008.
[22] A. Canova, N. Benvenuto, and P. Bisaglia, Receivers for

MIMO-PLC channels: throughput comparison, in Proceedings
of the 14th IEEE International Symposium on Power Line
Communications and its Applications (ISPLC 10), pp. 114119,
Rio de Janeiro, Brazil, March 2010.
[23] S. Katar, B. Mashburn, K. Afkhamie, H. Latchman, and R.
Newman, Channel adaptation based on cyclo-stationary noise
characteristics in PLC systems, in Proceedings of the IEEE
International Symposium on Power Line Communications and Its
Applications (ISPLC 06), pp. 1621, Orlando, Fla, USA, March
2006.
[24] A. M. Tonello, J. A. Cortes, and S. DAlessandro, Optimal
time slot design in an OFDM-TDMA system over power-line
time-variant channels, in Proceedings of the IEEE International
Symposium on Power Line Communications and its Applications
(ISPLC 09), pp. 4146, Dresden, Germany, April 2009.
[25] A. Scaglione, P. Stoica, S. Barbarossa, G. B. Giannakis, and H.
Sampath, Optimal designs for space-time linear precoders and
decoders, IEEE Transactions on Signal Processing, vol. 50, no. 5,
pp. 10511064, 2002.
[26] A. R. Horn and R. C. Johnson, Matrix Analysis, Cambridge
University Press, New York, NY, USA, 1985.
[27] A. Lozano, A. M. Tulino, and S. Verdu, Optimum power
allocation for parallel gaussian channels with arbitrary input
distributions, IEEE Transactions on Information Theory, vol. 52,
no. 7, pp. 30333051, 2006.
[28] S. D. Alessandro, A. M. Tonello, and L. Lampe, Adaptive
pulse-shaped OFDM with application to in-home power line
communications, Telecommunication Systems Journal, vol. 51,
no. 1, pp. 311311, 2011.
[29] F. J. Harris, On the use of windows for harmonic analysis with
the discrete Fourier transform, Proceedings of the IEEE, vol. 66,
no. 1, pp. 5183, 1978.
[30] K. H. Afkhamie, H. Latchman, L. Yonge, T. Davidson, and R.
Newman, Joint optimization of transmit pulse shaping, guard
interval length, and receiver side narrow-band interference
mitigation in the HomePlugAV OFDM system, in Proceedings
of the IEEE 6th Workshop on Signal Processing Advances in
Wireless Communications (SPAWC 05), pp. 9961000, New
York, NY, USA, June 2005.
[31] A. M. Tonello, S. DAlessandro, and L. Lampe, Cyclic prefix
design and allocation in bit-loaded OFDM over power line
communication channels, IEEE Transactions on Communica-
tions, vol. 58, no. 11, pp. 32653276, 2010.
[32] CENELEC, Power line communication apparatus used in low
voltage installationsradio disturbance characteristicslimits
andmethods of measurementpart 1: apparatus for inhome
use, Final Draft European Standard FprEN 50561-1, 2011.
[33] A. Schwager, W. Baschlin, J. L. Gonzalez Moreno et al., Euro-
pean MIMO PLC field measurements: overview of the ETSI
STF410 campaign & EMI analysis, in Proceedings of the IEEE
International Symposium on Power Line Communications and
its Applications (ISPLC 12), Beijing, China, March 2012.
[34] IEEE Standard 1901-2010, IEEE Standard for Broadband over
Power Line Networks: Medium Access Control and Physical
Layer Specifications, Chapter 16, Inter System Protocol (ISP),
2010.
[35] IEEE Standard 1901-2010, IEEE Standard for Broadband over
Power Line Networks: Medium Access Control and Physical
Layer Specifications, Annex R, Resynchronization example,
2010.
http://dx.doi.org/10.1155/2013/635086
Research Article
A Study on the Optimal Receiver Impedance for
SNR Maximization in Broadband PLC
Massimo Antoniali, Andrea M. Tonello, and Fabio Versolatto

Dipartimento di Ingegneria Elettrica Gestionale e Meccanica (DIEGM), Universita di Udine, 33100 Udine, Italy
Correspondence should be addressed to Andrea M. Tonello; tonello@uniud.it
Received 27 July 2012; Revised 31 October 2012; Accepted 13 December 2012
Academic Editor: Moises V. Ribeiro
Copyright 2013 Massimo Antoniali et al. This is an open access article distributed under the Creative Commons Attribution
cited.
We consider the design of the front-end receiver for broadband power line communications. We focus on the design of the input
impedance that maximizes the signal-to-noise ratio (SNR) at the receiver. We show that the amplitude, rather than the power,
of the received signal is important for communication purposes. Furthermore, we show that the receiver impedance impacts the
amplitude of the noise term. We focus on the background noise, and we propose a novel description of the noise experienced at
the receiver port of a PLC network. We model the noise as the sum of four uncorrelated contributions, that is, the active, resistive,
receiver, and coupled noise components. We study the optimal impedance design problem for real in-home grids that we assessed
with experimental measurements. We describe the results of the measurement campaign, and we report the statistics of the optimal
impedance. Hence, we study the best attainable performance when the optimal receiver impedance is deployed. We focus on the
SNR and the maximum achievable rate, and we show that power matching is suboptimal with respect to the proposed impedance
design approach.
1. Introduction corresponds to the input impedance of the antenna. Indeed,

the absence of reflections and stationary waves is obtained
The communication technology that exploits the power deliv- under simple (not complex) matching conditions, that is,
ery network to convey data is commonly referred to as power when the impedance of the transmitter and the receiver are
line communication (PLC). PLC is broadly deployed and, equal to the characteristic impedance of the transmission
recently, it has been recognized as a key technology to enable medium. Reflections and stationary waves are not desired.
the communication within the smart grid. The last node of the The reflected waves yield to multiple delayed echoes in the
smart grid is the home, where PLC is suitable for both home channel impulse response and they can be either due to the
entertainment, with datarates of about 200 Mbps, and home- multipath nature of the channel or due to the unmatched
automation, with lower data-rates but higher robustness and termination nodes. Matching ensures the absence of the latter
reliability. type of reflections. The stationary waves yield to voltage
The design of the PLC transceiver is a challenging task values along the line that are higher than necessary and that
due to the severe attenuation, fading effects, and noise may damage the transmission medium, especially in wireline
impairments that characterize the communication media. In communications as, for instance, over coaxial cables.
other application scenarios, as wireless, the transceivers are When the characteristic impedance of the medium is
designed to fulfill the maximum power transfer condition. real, the maximum power transfer and the absence of reflec-
Basically, the maximum power transfer is achieved under tions are ensured by the same matching conditions. In the
complex matching conditions, that is, when the internal following, we simply indicate with impedance matching the
impedance of the transmitter and the input impedance of receiver design that enables the maximum power transfer.
the receiver are the complex conjugate of the characteristic In wireless, impedance matching can be easily satisfied by
impedance of the transmission medium. In wireless, this letting the characteristic impedance of the antenna and the
cables, the internal impedance of the transmitter and the premises. For each site, we performed measurements
input impedance of the receiver be equal to the reference between all couples of available outlets in the 1.5100 MHz
value of 50 . frequency range. The measured database is useful to provide
In PLC, impedance matching is typically achieved a realistic description of the frequency response and the line
through the use of impedance matching networks, so that the impedance that characterizes the signal paths.
received signal power is maximized. Basically, the matching In this work, we focus on the time invariant description
network is a loss-less network of lumped elements whose of the network because, from the experimental evidence, we
input impedance is matched to that of the power delivery observed that very little (or inexistent) time variation was
network. The latter is frequency dependent and it varies sig- present in the sites that we considered, namely, in the order of
nificantly from outlet to outlet. Furthermore, it may exhibit few dBs. More in general, the PLC channel can be periodically
a time-varying behavior [1]. Several matching networks time variant. In such a case, an extension of our analysis can
were presented in the literature for PLC, and their use was be obtained under the assumption that both the channel and
demonstrated to be beneficial in terms of an increase of the the noise exhibit a slow periodic variation that is synchronous
received signal power [2, Ch. 4] [3, 4]. However, from a data with the mains. Thus, the mains period can be divided into
transmission point of view, it is important to maximize the short time intervals (slots) during which we can reasonably
signal-to-noise ratio (SNR) and not only the signal power. assume both the channel and the noise to be time invariant.
The effective formulation of the SNR is in signal ampli- Hence, we can apply the analysis that we propose in this work
tude rather than in power terms. The reason is that the analog to each single slot.
front-end of the PLC receiver is designed to convert the We study the attainable SNR improvement when the
analog amplitude of the received signal into a digital sample optimal receiver impedance is used. We compare the results
stream. In this respect, the amplitude of the received signal is to the case of impedance matching, when the receiver
more important than its power, though the two quantities are impedance is constant and equal to the reference value of
related once the receiver impedance is given. 50 , and when it is constant and equal to 1 k. The latter case
In PLC, the maximum power transfer condition does is representative of a receiver with high input impedance.
not imply the maximization of the SNR in amplitude terms The remainder of the work is as follows. In Section 2,
because power matching may turn into a higher noise contri- we describe the model of the power delivery network. In
bution. Concerning an in-home network, the noise injected Section 3, we overview the transceiver design, with emphasis
by the household devices is attenuated by the insertion loss of on the receiver side. Then, in Section 4, we formulate the
the path followed to reach the receiver outlet. In general, the optimization problem, and, in Section 5, we study its con-
latter is different from that followed by the communication vexity. In Section 6, we present several numerical results.
signal. The receiver impedance affects the insertion loss, and Basically, we study the optimal receiver impedance for a
common impedance matching techniques do not take into real scenario. Furthermore, we provide some details on the
account the impact of the receiver impedance on the insertion measurements. Then, we study the performance in terms of
loss experienced by the noise. It follows that impedance SNR and achievable rate and we compare the results to the
matching may reduce the attenuation of the noise path, ones that we would obtain for the other receiver impedance
thus increasing the noise amplitude at the receiver port. designs. Finally, the conclusions follow.
Furthermore, the receiver impedance itself contributes to the
increase of the amplitude of the noise.
In this work, we discuss the optimal design of the receiver 2. Power Delivery Network Model
impedance that enables achieving the maximum SNR in
broadband PLC. We formulate the SNR in terms of amplitude PLC experiences high attenuation and deep fading effects that
and we focus on in-home networks. are a function of the loads and the layout of the network. We
Firstly, we propose an analytic description of the noise as focus on in-home networks. In [5], we presented an accurate
the sum of multiple contributions. We consider the resistive model of the in-home power delivery network. Basically, in-
noise of the network by itself, the noise injected by the home networks exhibit a treelike structure and the outlets are
household devices, and the noise introduced by the receiver the termination nodes. The outlets are fed by the power line
impedance. From the experimental evidence, we show that cables which resemble the branches of the network. Between
the noise injected by the household devices, namely, the active any pair of outlets, only one electrical path is possible.
noise, dominates among all noise contributions. From a data transmission perspective, we can abstract
Then, we formulate the SNR as a function of the the power delivery network to obtain an equivalent represen-
receiver impedance. To this aim, we assume the transmitter tation that is suitable for the SNR analysis of the following
impedance to be constant and known. We study the convexity sections. In Figure 1, we show the model. We represent the
of the optimization problem and we derive the optimal power delivery network as an O-port network, where O is
receiver impedance that maximizes the SNR at each single the number of outlets. The transmitter and the receiver are
frequency. connected to two ports of the network, namely, ports and
Finally, we validate the results for real-life in-home . We model the transmitter as a real voltage source with its
networks. We exploit the results of an experimental own internal impedance, and the receiver as a passive load.
measurement campaign that we carried out in Italy, where The household appliances are connected to the remaining
we collected more than 1200 channel responses in different ports of the network. They inject noise, and we model them as
IMN VGA BPF AGC ADC
1PSU 1PSU

Figure 2: Block diagram of the PLC receiver AFE [6].

on the amplitude of the received signal, rather than its power,
and this motivates the formulation of the SNR in amplitude
terms. Furthermore, to preserve the signal amplitude, the
input impedance of the analog-to-digital converter (ADC)
Figure 1: Equivalent model of the power delivery network. circuit is typically high [7].
At the receiver, the AFE consists of several blocks. In
Figure 2, we show the schematic diagram [6]. The variable
gain amplifier (VGA) is followed by a band-pass filter (BPF)
noise sources, namely, real voltage generators with their own
and an adaptive gain control (AGC). A low-noise amplifier
internal impedance. In general, the number of household
(LNA) can be used to substitute the VGA, and an additional
appliances is O because some outlets may be not in use,
ad hoc amplifier can be deployed to drive the AD converter
as shown in the uppermost part of Figure 1.
instead of the AGC [8]. The gain of the AGC is adjusted
The signal path between two ports of the network is
dynamically to ensure always the optimum signal level for AD
characterized by the channel frequency response (CFR). We
conversion and the maximum dynamic range [6].
refer to the CFR as the ratio, in the frequency domain,
Practical receiver schemes adopt matching techniques
between the voltage at the output and input ports.
to interface the blocks of the AFE. In wireless, where the
We introduce the following notation. We denote the com-
communications are in the range of GHz, the reference
plex amplitude of the source signal, the internal impedance
impedance value is 50 [9], and the advantages provided
of the transmitter, the complex amplitude of the voltage at
by impedance matching are the following. Firstly, since
the transmitter port, the complex amplitude of the voltage
the reference impedance is real, matching enables both the
at the receiver port, and the input impedance of the receiver
maximum transfer of power and the absence of reflections
with (), (), (), (), and (), respectively.
at the receiver port. The latter condition is desired to avoid
All the quantities are defined in the frequency domain, at
multipath fading exceeding that amenable closely to the
frequency . Concerning the noise, we use the subscript {}
channel. Secondly, matching simplifies the design of the LNA
to denote the th source of noise, where = 1, . . . , , and
and the BPF. The noise figure (NF) of the LNA is a function
we denote the complex amplitude of the th noise source and
of the impedance that is seen at the input and output port
its internal impedance with () and (), respectively,
of the amplifier [10]. By letting the impedance be equal to
and the voltage amplitude at the port where the th noise
a reference value, the LNA can be optimized to exhibit the
source is connected with (). Furthermore, we note that
lowest NF. Similarly, the behavior of the BPF is a function of
all impedances are, in general, noisy and thus they generate
the impedance seen by the filter at the input and output ports.
a thermal noise contribution. Finally, we denote the CFR
The BPF is designed to operate in matching conditions, and
between the transmitter and the receiver port with (),
its frequency behavior, that is, the pass-band and the stop-
and between the port where the th noise source is connected
bands, may vary significantly otherwise [9].
and the receiver port with ().
In PLC, the transmission interests the frequency range
below 100 MHz, where the presence of reflected waves can
3. PLC Transceiver Design be tolerated to preserve the amplitude of the received signal.
Therefore, the interface between the blocks can be designed
We consider the analog front-end (AFE) of the transceiver. in high-impedance mode.
At the transmitter side, the AFE amplifies the signal. The The use of an impedance matching network (IMN)
final amplification stage is the line driver. In this work, we before the AFE provides some benefits. Strictly, the matching
do not consider the impairments related to the design of the network allows the receiver to be matched to the complex
line driver and we model it as a real voltage generator with and frequency-dependent impedance of the power delivery
an internal impedance . To obtain the numerical results of network, regardless of the input impedance of the first stage of
Section 6, we simply neglect it; that is, we let = 0 . The the AFE. In the literature, the design of the matching network
assumption is validated by the fact that the output impedance was aimed at obtaining the maximum transfer of power from
of the transmitter stage of commercially available broadband the network [3, 4]. Alternative solutions are possible. In this
PLC transceivers is typically low, for example, 3 [6]. respect, we propose the use of the matching network to
At the receiver side, the AFE processes the received signal obtain the optimal receiver impedance that lets the SNR be
to make it suitable for the analog-to-digital (AD) conversion. maximum, as we describe in the following sections. Basically,
The final goal is to convert the (analog) amplitude of the we model the IMN as an impedance in parallel with the
received signal into a digital sample stream. Thus, the focus is first stage of the AFE, and we assume the first stage to exhibit
an infinite input impedance. Therefore, the resultant input

impedance of the transceiver is equal to .

4. SNR Formulation
We follow the notation of Figure 1. We study the SNR at the

receiver port when the transmitter is connected to port ,
and noise sources are connected to the remaining ports (a)
of the network. At the receiver port, () is the sum of the
signal of interest () and the noise (), that is, () =
() + (). We formulate the SNR in amplitude terms, as
the power spectral density (PSD) of () divided by the PSD

of (). It reads
()
() = . (1)
() (b)

To obtain (1), we assume the signals to be stationary and
continuous in time, and we note that the PSD of the generic
signal (), limited in power, can be computed as [11]
2
( () )
() = lim [V2 /Hz] , (2)
(c)
where () is the Fourier transform of () on a finite time Figure 3: Equivalent models for the active, resistive, and receiver
interval of length , and () denotes the statistical expectation. noise terms.
Now, we study the terms in (1). Firstly, let us consider the
signal of interest. According to the notation of Figure 1, we
explicit it as We show the equivalent model in Figure 3(a). The active noise
at the receiver port reads
2
() = ( , ) () [V2 /Hz] . (3)
() = ()
Hence, let us focus on the noise term. In this work, we =1

limit the study to the stationary noise components, and we

explicit the PSD of the noise at the receiver port, namely, 2
() as the sum of contributions due to four noise terms. = ( , )
=1
They are the active noise (), the resistive noise (), the
receiver noise (), and the coupled noise (). The latter 2
( , )
component models the radio disturbances that are captured () [V2 /Hz] ,
( , ) + ()
by the wiring through coupling. We assume the coupled noise
to be independent from the receiver impedance. Thus, we (5)
can neglect the impact of the coupled noise on the design
of the optimal receiver impedance, and we let () = 0. where ( , ) is the input impedance of the power delivery
Furthermore, we assume the noise terms to be uncorrelated network at the port where the th appliance is connected.
with zero mean. It follows All other quantities in (5) are consistent with the notation
of Figure 1. Furthermore, we note that we account for the
() = () + () + () [V2 /Hz] . (4) thermal noise contribution due to the real part of () in
the resistive noise term. Toward the receiver, the active noise
In the next sections, we describe the noise terms in (4) injected by the th device is attenuated by ( , ) and,
and we propose an analytical expression to model them, we further, by an attenuation factor that is a function of the input
introduce and motivate the approximations that allow us to impedance ( , ). Both ( , ) and ( , ) depend
formulate and solve the SNR maximization problem and, on the load impedance (). It follows that the design of the
finally, we present the final SNR expression. load impedance has an impact on the received active noise.
Now, we introduce the following approximations. We
assume that () is identical for all appliances, that is,
4.1. Active Noise. The active noise is generated by the power-
supply circuitry of the household appliances that are con- () = () for all {1, . . . , }, and we neglect

nected to the power delivery network. We model the house- the source impedance, that is, we let () = 0. In this
hold appliances as real and independent voltage generators. respect, we point out that by letting = 0, we follow a
conservative approach because, in such a case, we experience further details, we refer to [14]. In Figure 3(b), we show the
the highest level of noise at the receiver. equivalent network model from which we obtain
In Section 5, we show that the optimal receiver
impedance is independent from (). However, the
value of () is fundamental to obtain realistic values 2
of SNR. To determine (), we explicit the quantity as ()
() = 4R { ()} [V2 /Hz]
follows: () + ()
(10)
() () ()
() = [V2 /Hz] ,
( , )2
=1 and we remark that () is a function of the internal
(6) impedance of the transmitter and the loads.
and we exploit the experimental results in [12]. Basically,
is the total additive background noise experienced at
4.3. Receiver Noise. We isolate the thermal noise due to the
the receiver port. In [12], the total additive background noise
receiver from the contribution due to the rest of the network
experienced at the receiver port was measured and modeled
and we refer to it as receiver noise, namely, (). The receiver
in power terms for the reference load of 50 . The model
noise is due to the parallel impedance that we propose to
reads
use as IMN before the first amplification stage of the front-
end, that is, the VGA. The parallel impedance provides an
1 SNR improvement as shown in Section 6, but it can be, in
bck () = + 1015.5 [mW/Hz] . (7)
2 general, resistive and thus it may increase the thermal noise
contribution.
To be consistent with the measurement setup in [12], in We do not account for the noise contribution due to the
(6), we let be constant and equal to 50 and we substitute VGA because the input impedance of the VGA is modeled
with as noiseless. In fact, all noise contributions due to the
amplification stage are described by the noise figure of the
AFE [15].
() = 0.05 bck () [V2 /Hz] , (8) In Figure 3(c), we show the equivalent model of the noisy
receiver impedance. We model the PSD of () as
that is, the V2 /Hz-representation of the measured noise PSD
in (7), where the coefficient 0.05 allows for the conversion
from dBm/Hz to V2 /Hz. Furthermore, in (6), we compute 2
() and () as described in Sections 4.2 and 4.3, ()
() = 4 R { ()} [2 /] .
respectively. () + ()
(11)
4.2. Resistive Noise. Resistive elements introduce thermal
noise [13]. The noisy impedance () can be modeled as a
thermal noise voltage source and a noiseless impedance. The From (11), we note that the PSD of the amplitude of the
PSD of the amplitude of the thermal noise reads receiver noise is lower than the actual noise generated by the
impedance () due to the presence of the voltage divider
term.
() = 4R { ()} [2 /] , (9)
where = 1.38 1023 J/K is the Boltzmann constant, = 4.4. Final Model. We now explicit the dependencies of the
290 K is the absolute temperature, and R{} denotes the real SNR from the impedance of the receiver. To this aim,
operator. The branches and the nodes of the power delivery we proceed as follows. Firstly, we formulate the CFR and
network are not ideal and they show a resistive component the output impedance of the power delivery network as a
that contributes to the thermal noise at the receiver. Further function of the chain-matrix parameters, that is, [16]
sources of thermal noise are the internal impedance of both
the transmitter and the household appliances. Now, we point
out that the power delivery network between the transmitter 1
and the receiver port can be modeled as a passive two-port ()
( , ) = ( () + ) , (12)
network. We can gather the thermal noise contributions into ()
the resistive noise PSD that can be obtained from the network
impedance at the receiver port (). In this respect, we point
out that an alternative description of the noise generated by () + () ()
() = [] , (13)
the network can be obtained in terms of noise figure. For () + () ()
2
1
( + )

( ) = ,
2 + 2

2
1
4R { + } + + 4R { }
(14)
( + ) +

+
=1
+ +
+
+
+

+
active noise resistive noise receiver noise
where = , , and , , and are the parameters of the first term, while the contributions of the resistive and
the transmission matrix that describes the link . Then, we receiver noise are gathered in the second term. Secondly,
substitute (3), (4), and (12) in (1), and we replace the PSD the receiver noise contribution in (, ) does not increase
of the noise terms with the correspondent expressions (5), monotonically as a function of . Furthermore, (, )
(10) and (11), where we explicit () according to (13). In can be neglected because it is always significantly lower
(14), we report the resultant SNR expression. We neglect the than (, ) for realistic noise scenarios. In fact, (, )
frequency dependency for notation simplicity. is dominated by the Boltzmann constant , and it assumes
values in the order of 1023 . Indeed, assumes values in
5. SNR Optimization Problem the order of 1017 , according to the derivations of Section 4.1.
Therefore, we can approximate the SNR as
We formulate the SNR optimization problem as follows:
1
opt () = argmax { ( , )} (, ) . (22)
() =1 (, ) / (, )
(15) We substitute (22) into (15) and we study the concavity of
s.t. () 0,
the maximization problem. In general, the expression in (22)
() R, is not concave in the domain of and that we consider. In
detail, the functions (, ) are convex. The ratio of convex
where R denotes the set of real numbers, and we introduce the functions is, in general, not convex. Similarly, the sum of
notation () = R{ ()} and () = I{ ()}, with I{} nonconvex functions is, in general, not convex. Therefore, the
being the imaginary operator. We also highlight that the SNR denominator term in (22) is, in general, not convex. It follows
is dependent on and . We start from (14) and, according that the maximization problem is not concave [17].
to the derivations that we report in the Appendix, we obtain Consequently, we performed the exhaustive search to
identify the optimal pair (, ) for each frequency bin, as
(, ) = , (16) described in Section 6. From simulation results, we obtained
(, ) + (, )
that the optimal receiver resistance is identically null in the
where entire frequency range, that is, = 0. Therefore, by letting
= 0 in (15), we note that the domain of the maximization
(, ) problem can be limited to the set of values = {} for which
(, ) = , (17)
(, )
=1 ( = 0, )
= 0. (23)
2

(, ) = 4 ( + 2 2 ) , (18)
+
6. Numerical Results
2 2
(, ) = (2 + 2 ) + 2 + 2 + , (19)
We have studied the optimal receiver impedance in real
networks. To this aim, we carried out a measurement cam-
= R { } , (20)
paign in Italy and we collected more than 1200 channel
= I { } , (21) responses. We considered three sites, with 11, 23, and 26
outlets, respectively. We followed an exhaustive approach. For
and = , . Furthermore, we remark that all quantities in each site, we performed measurements between all pairs of
(16)(21) are a function of the frequency and we neglect the available outlets, where no loads were connected.
dependence only for notation simplicity. Experimentally, we We performed measurements in the frequency domain.
have observed that 0, while can be either positive or We deployed a vector network analyzer (VNA) in combi-
negative. nation with broadband couplers and extension cables. We
Now, we focus on the denominator in (16) and we removed the effect of the couplers and the cables from the
note the following. Firstly, the active noise is described by measures to obtain the actual scattering parameters of the
PLC channel. We connected the VNA to the network through 0

coaxial cables and broadband couplers. Couplers protect the
20
equipment from the mains and they show an attenuation
CFR (dB)
of 50 dB at the mains frequency, and lower than 5 dB up to 40
100 MHz.
60
We calibrated the VNA when only the cables were
connected and we removed the effect of the couplers by 80
0 20 40 60 80 100
exploiting the chain rule of the ABCD matrices. To this aim,
we characterized the couplers in terms of ABCD matrices.
The procedure proved to be the most reliable. (a) Frequency (MHz)
From the measured scattering parameters, we computed
0
the CFR. In Figure 4, we provide the results. In different
subplots, we show the channels from the three sites. Further- 20
CFR (dB)
more, we report the average profile (dashed line). In all cases 40
the CFR exhibits a frequency decreasing behavior. In sites 2
60
and 3, that is, the largest premises, the average attenuation
is greater than in site 1, that is, a small urban flat. For the 80
statistical characterization of the channels, we refer to [18], 0 20 40 60 80 100
where we describe the channels in terms of the main metrics,
and we study the probability density function of the CFR as a (b) Frequency (MHz)
function of frequency. Furthermore, we point out that herein 0
we consider a restricted set of the measurements in [18]. 20
From the measured scattering parameters, we also com-
CFR (dB)
40
puted the output impedance. We define it as in (13), and we
60
note that it can be interpreted as the network impedance at
the receiver port when the impedance is connected to the 80
transmitter port [18]. Hence, is a function of the source 100
impedance . In our case, = 0 . We denote the real 0 20 40 60 80 100
and imaginary component of () with () and (),
respectively. To analyze the results, in Figure 5, we report the (c) Frequency (MHz)
statistics in quantile terms. We focus on three probability
Figure 4: Measured channel frequency responses. From top to
values, that is, = 10, 50, and 90%, and we plot the quantiles
bottom, sites 1, 2, and 3. In all cases, the mean profile is also shown
, () that corresponds to Pr[() , ()] = , where (dashed line).
{ , } and Pr[] denotes the probability. Interestingly,
we can note that the resistive component is more spread in the
lower frequency range, while the reactive component exhibits the search domain to {0, 1 k} and to {1 k, 1 k} for the
statistically an inductive-like behavior; that is, it increases resistive and the reactive component, respectively. In both
with frequency. cases, the resolution is 1 . In Figures 6 and 7, we show
The measured values of , , and have been substi- the quantiles of the components of opt () as a function of
tuted in (14). By exploiting measurements, we obtain realistic the frequency. We limit the plot to three probability values,
results, although we introduced the following approxima- that is, = 20, 50, and 80%. Figure 6 shows that the
tions. We neglected the impact of the internal impedance resistive component of the optimal receiver impedance is
of the noise sources on and , and the impact of the equal to 0 for all realizations and frequencies. We speculate
internal impedance of the transmitter on . In fact, when that this result is determined by the nonnegative behavior of
we performed measurements, no loads were connected to the the coefficients in (18)(20). Indeed, Figure 7 shows that
available outlets. The household appliances were connected the reactive component of the optimal receiver impedance
to the remaining outlets. Hence, according to the model exhibits a strong frequency-dependent behavior. Further-
of Figure 1, the measures resemble the case of no loads more, the median value of the reactive component is close
connected to the transmitter and the noise ports. Similar to 0. It follows that the optimal impedance does not exhibit a
considerations hold true for the measurements of the output prominent inductive-like or capacitive-like behavior.
impedance. In Figures 6 and 7, we compare the statistics of the
For the analysis of the effect of the optimal impedance, we optimal receiver impedance to the statistics of the matched
focus on the 1.5100 MHz frequency range and the resolution impedance, that is, the impedance that is designed to ful-
of the measurements is = 500 kHz. In the following, we fil the power matching conditions. To this aim, we show
adopt the discrete frequency representation, and we denote the quantiles of the receiver impedance under the power
with the frequency sample = . Furthermore, 1 = matching assumption, namely, pmi = . We focus on
1 = 1.5 MHz and 2 = 2 = 100 MHz. the probability values = 20, 50 and 80%. Differently from
We carried out an exhaustive search of the optimal the optimal case, the resistive component of the matched
receiver impedance for the measured channels. We limited impedance is nonzero and, by definition, it is equal to the
400 300

,90
Resistance ( )
,50
300
,10
200
200
100
100
3FBDUBODF
0 0
0 20 40 60 80 100
(a) Frequency (MHz) 100
300
200
Reactance ( )
200
100 300
0
,90 400
100 ,10 ,50 0 20 40 60 80 100
200 Frequency (MHz)
0 20 40 60 80 100
Optimal reactance
(b) Frequency (MHz)
Matched reactance
Figure 5: Quantiles of the resistive (a) and reactive (b) component Figure 7: Quantiles of the reactive component of the optimal and
of the output impedance. Three probability values are considered, power matched receiver impedance. Three probability values are
that is, 10, 50 and 90%. considered, that is, 20, 50, and 80%.
200
180 ,80
that is, the values that yield to a transmitted power
160 of 50 dBm/Hz below 30 MHz, and 80 dBm/Hz beyond
140 30 MHz on a reference load of 50 . Basically, we shape
Resistance ( )
120
as described in [12], and not as directly specified by the EMC
norms because the latter do not target the frequency range
100 beyond 30 MHz yet. Furthermore, we determine as
,50
80 described in Section 4.1.
60
,20 We compute the SNR when the receiver impedance is
optimal, matched in power, or constant in frequency and
40 equal to 50 = 50 . We assume the latter to be the reference
20 Optimal resistance
case, and we obtain the SNR improvements provided by
the use of opt and pmi with respect to 50 , that is,
0
Frequency (MHz) ( , ) = (( , ))dB ((50 , ))dB where {opt, pmi}.
Note that is formulated as the difference of the SNR
Optimal resistance in dB terms. Hence, we study the statistics of ( , ).
Matched resistance In Figure 8, we show the fitted profiles of the quantiles of
Figure 6: Quantiles of the resistive component of the optimal and
( , ) associated to the probability values 20, 50, and
power matched receiver impedance. Three probability values are 80%. Impedance adaptation techniques are more beneficial
considered, that is, 20, 50, and 80%. in the lower frequency range, where the optimal impedance
and the power matched impedance provide, in half of the
cases, up to 4 and 1 dB of improvement, respectively. In the
resistive component of the output impedance. Indeed, the higher frequency range and for half of the cases, the matched
statistics of the reactive component of pmi and opt are impedance does not provide any improvement, while the
similar, but, for the matched impedance, the high-probability optimal impedance ensures an SNR gain of approximately
area is better confined around the median and the profiles of 1 dB. Finally, we note that, according to the formulation in
the quantiles associated to probabilities = 20 and 80 are (14), the matched impedance can even reduce the SNR with
smoother in frequency. Finally, the median of the reactive respect to the case of using 50 (see Pr[(pmi , ) ] =
component of the matched impedance overlaps the median for = 20% in Figure 8).
of the reactive component of the optimal impedance. Now, we study the performance in terms of achievable
Now, we study the SNR, namely, ( , ). We compute it rate. We assume the transmitted signal and the noise to be
according to (14). We let the transmitted power (in dB terms) Gaussian, and we define the achievable rate as follows:
be
2
63 dBV/Hz for 30 MHz () = log2 (1 + ( , )) [bps] ,
() = { (24) (25)
93 dBV/Hz otherwise =1
8 1
1
7 0.9
1S< PQU
>
6 0.9
0.8
5
0.7
1S< QNJ
> 0.8
4
E#
0.6
3
CCDF
0.5 0.7
2
1 0.4
0.6
0 0.3 500 1000
1 0.2
0 20 40 60 80 100
Frequency (MHz) 0.1
0
1000 2000 3000 4000

PQU

Figure 8: Best cubit fit of the SNR quantile profiles associated to the QNJ
IJHI
probability values = 20, 50, and 80%.

(a) Achievable Rate (Mbps)
1
where {opt, pmi, 50}, and we limit the transmitted PSD
0.9
according to (24). In Figure 9(a), we show the complementary
cumulative distribution function (C-CDF) of the achievable 0.8
rate for the three receiver impedances that we consider.
We magnify the plot for the probability values greater than 0.7
0.6. As expected, the optimal impedance outperforms the
0.6
power matched impedance. As an example, (opt) exceeds
800 Mbps in 83% of the cases, while (pmi) achieves the same
CDF
0.5
rates in 77% of the cases. In a dual manner, with probability
0.9, (opt) > 745 Mbps while (pmi) > 695 Mbps, with an 0.4
improvement of 7%.
Now we quantify the improvement with respect to 50 . 0.3
We compute the achievable rate improvement as follows:
0.2
() (50)
() = 100 [%] , (26) 0.1
(50)
0
where {opt, pmi}. In Figure 9(b), we show the cumulative 0 5 10 15 20 25
distribution function (CDF) of (). The use of the optimal
impedance turns into achievable rate improvements of up to PQU
IJHI
QNJ
24%. Interestingly, the power matching approach does not

provide significant benefit with respect to 50 . In this case, the (b) Improvement (%)
improvement is lower than 5% in 90% of the cases. Further-
more, the matched impedance can even reduce the achievable Figure 9: (a) C-CDF of the achievable rate when the receiver
rate. The reason is that the matched impedance maximizes the impedance is optimal, matched in power or constant and equal to
50 or 1 k. (b) CDF of the achievable rate improvement with
power of the received signal, but not necessarily the SNR.
respect to the case of 50 .
6.1. High Receiver Impedance Design. In the previous sec-

tion, we have shown the performance when the receiver
impedance is optimal, matched in power or constant and the study to the achievable rate. We compute it as in (25),
equal to 50 . It is also interesting to address the performance and we denote it with (high). Similarly, we compute the
when the receiver impedance is high. In such case, we denote achievable rate improvement as in (26), and we denote it with
the receiver impedance with high and we let it be purely real (high). In Figure 9, we show both the C-CDF of (high)
and equal to 1 k, that is, the largest value that we admitted and the CDF of (high). As it can be noted, high does not
for the exhaustive search of the optimal impedance. We limit provide any benefit with respect to 50 and the two solutions
are equivalent in terms of achievable rate. Similar results can and we proceed as follows. Firstly, we introduce the notation
be obtained by letting high be large, and purely imaginary = R{ } and = I{ }. Then, we substitute in (12)
either positive or negative. with + , where is the imaginary unit, to obtain
As a final remark, we note that the optimal impedance
provides the best gain in terms of achievable rate. However, it 2 2 + 2
() = 2 , (A.1)
is rather complex to be implemented in practice. Therefore, (2 + 2 ) + 2 + 2 + 2
more simpler though suboptimal solutions, as 50 () and
high (), are often used. where = , , and are defined as in (20)-(21), and we
exploited the following relations:
7. Conclusions
= R { } R { } + I { } I { } , (A.2)
We have investigated the optimal design of the receiver input
impedance for the maximization of the SNR in broadband = R { } I { } I { } R { } . (A.3)
PLC. We have firstly presented a comprehensive description
of the power delivery network. We have modelled the Now, we focus on the output impedance . When =
contribution of the household appliances on the noise at the 0, the real part of reads
receiver port. Then, we have discussed the front-end design
R { }
for broadband PLC transceivers. We have pointed out that R { ( = 0 )} = R { } = 2 . (A.4)
the amplitude (and not the power) of the received signal is

important for data communication purposes.
Therefore, we have formulated the SNR in amplitude We exploit (A.4), and we explicit the resistive and the
terms, and we have modelled the noise as the sum of four con- receiver noise in (14) as
tributions, namely, the active, resistive, receiver, and coupled
+ 2
noise. Basically, the active noise is the noise that is injected by
=
the household appliances, the resistive noise accounts for the / + +
thermal noise due to the resistive components of the network, (A.5)
2 2
the receiver noise is the thermal noise due to the receiver ( + 2 )
= 2
impedance, and the coupled noise is the noise that couples (2 + 2 ) + 2 + 2 + 2
into the wirings, for example, due to broadcast radios.
We have studied the dependency of the SNR from 2
/
the receiver impedance and we have highlighted the latter =
impacts not only on the amplitude of the useful signal, but / + +
(A.6)
also on the amplitude of the noise at the receiver port. 2
Hence, we have found the optimal receiver impedance that = 2 ,
maximizes the SNR. To this aim, we have exploited the results (2 + 2 ) + 2 + 2 + 2

of an experimental measurement campaign where we have
collected the scattering parameters of PLC channels in real respectively. Finally, we substitute (A.1) and (A.5)-(A.6) in
home grids. From the measurement results, we have found (14); we divide the numerator and the denominator by (A.1)
that the optimal impedance is purely reactive and not equal and we obtain (16).
to that obtained according to the power matching approach,
namely, the matched impedance. We have compared the References
performance of the optimal impedance to that of the matched
impedance in terms of SNR and achievable rate for real- [1] F. J. Canete, J. A. Cortes, L. Dez, and J. T. Entrambasaguas,
life scenarios. We have assumed the performance of a 50- Analysis of the cyclic short-term variation of indoor power line
receiver as the reference case and we have found that while channels, IEEE Journal on Selected Areas in Communications,
the matched impedance can even reduce the performance, vol. 24, no. 7, pp. 13271338, 2006.
the optimal impedance provides improvements up to 24% in [2] H. C. Ferreira, L. Lampe, J. Newbury, and T. G. Swart, Power
terms of achievable rate. Finally, we have shown that when the Line Communications: Theory and Applications for Narrowband
receiver exhibits a high input impedance, the achievable rate and Broadband Communications over Power Lines, Wiley &
is close to that obtained with a 50- receiver. Therefore, the Sons, New York, NY, USA, 2010.
two solutions are equivalent in performance. [3] N. Taherinejad, R. Rosales, S. Mirabbasi, and L. Lampe, On
the design of impedance matching circuits for vehicular power
line communication systems, in Proceedings of the IEEE Inter-
Appendix national Symposium on Power Line Communications and Its
Applications (ISPLC 12 ), pp. 322327, March 2012.
SNR Derivations [4] P. A. Janse van Rensburg and H. C. Ferreira, Design of a bidi-
rectional impedance-adapting transformer coupling circuit for
We aim to simplify the SNR formulation in (14) to obtain (16). low-voltage power-line communications, IEEE Transactions on
We neglect the frequency dependency for notation simplicity, Power Delivery, vol. 20, no. 1, pp. 6470, 2005.

channel modelingpart I: random topology model and effi-
cient transfer function computation, IEEE Transactions on
Power Delivery, vol. 26, no. 2, pp. 891898, 2011.
[6] Maxim, Integrated powerline communication analog front-
end transceiver and line driver, http://www.maximic.com/
datasheet/index.mvp/id/6333.
[7] P. G. M. Baltus and R. Dekker, Optimizing RF front ends for
low power, Proceedings of the IEEE, vol. 88, no. 10, pp. 1546
1559, 2000.
[8] J. Bauwelinck, E. de Backer, C. Melange et al., A 1024-QAM
analog front-end for broadband powerline communication up
to 60 MHz, IEEE Journal of Solid-State Circuits, vol. 44, no. 5,
pp. 14771485, 2009.
[9] B. Razavi, RF Microelectronics, Prentice Hall, 1998.
[10] C. Bowick, RF Circuit Design, Newnes, 1982.
[11] L. W. Couch, Digital and Analog Communication Systems,
Prentice Hall, 1997.
[12] R. Hashmat, P. Pagani, and T. Chonavel, MIMO communica-
tions for inhome PLC networks: measurements and results up to
100 MHz, in Proceedings of the IEEE International Symposium
on Power Line Communications and Its Applications (ISPLC 10),
pp. 120124, March 2010.
[13] A. Papoulis, Probability, Random Variables, and Stochastic
Processes, McGraw-Hill, 1991.
[14] H. T. Friis, Noise figures of radio receivers, Proceedings of the
IRE, vol. 32, no. 7, pp. 419422, 1944.
[15] W. F. Egan, Practical RF System Design, John Wiley & Sons, 2003.
[16] D. M. Pozar, Microwave Engineering, Wiley Academy, 2004.
[17] S. Boyd, Convex Optimization, Cambridge University Press,
2004.
[18] A. M. Tonello and F. Versolatto, PLC channel characterization
up to 300 MHz: frequency response and access impedance,
in Proceedings of the IEEE Global Communications Conference
(GLOBECOM 12), December 2012.
)JOEBXJ 1VCMJTIJOH $PSQPSBUJPO
+PVSOBM PG &MFDUSJDBM BOE $PNQVUFS &OHJOFFSJOH
7PMVNF "SUJDMF *% QBHFT
IUUQEYEPJPSH
3FTFBSDI "SUJDMF
&OIBODFNFOUT PG (1-$ 5FDIOPMPHZ GPS
4NBSU)PNF#VJMEJOH "QQMJDBUJPOT
-VDB %J #FSU 4BMWBUPSF %"MFTTBOESP BOE "OESFB . 5POFMMP

%JQBSUJNFOUP EJ *OHFHOFSJB &MFUUSJDB (FTUJPOBMF F .FDDBOJDB %*&(.
6OJWFSTJU EJ 6EJOF 6EJOF *UBMZ
$PSSFTQPOEFODF TIPVME CF BEESFTTFE UP "OESFB . 5POFMMP UPOFMMP!VOJVEJU
3FDFJWFE 4FQUFNCFS "DDFQUFE %FDFNCFS
"DBEFNJD &EJUPS "INFE ;FEEBN
$PQZSJHIU -VDB %J #FSU FU BM JT JT BO PQFO BDDFTT BSUJDMF EJTUSJCVUFE VOEFS UIF $SFBUJWF $PNNPOT "UUSJCVUJPO -JDFOTF
XIJDI QFSNJUT VOSFTUSJDUFE VTF EJTUSJCVUJPO BOE SFQSPEVDUJPO JO BOZ NFEJVN QSPWJEFE UIF PSJHJOBM XPSL JT QSPQFSMZ DJUFE
5P FOBCMF UIF TNBSU HSJE DPODFQU JU JT GVOEBNFOUBM UP DPOTJEFS UIF JOIPNFCVJMEJOH DPOUFYU XIFSF CFTJEF UIF DPOWFOUJPOBM IPNF
OFUXPSLJOH TFSWJDFT IPNF BVUPNBUJPO BOE TNBSU FOFSHZ NBOBHFNFOU TFSWJDFT IBWF UP CF PFSFE *O UIJT QBQFS XF DPOTJEFS UIF JO
IPNFCVJMEJOH TDFOBSJP GPS XIJDI XF QSPQPTF B DPOWFSHFOU OFUXPSL BSDIJUFDUVSF UP FOIBODF UIF QFSGPSNBODF PG UIF OBSSPXCBOE
QPXFS MJOF DPNNVOJDBUJPO 1-$
(1-$ UFDIOPMPHZ UISPVHI JUT JOUFHSBUJPO XJUI BO &UIFSOFUCBTFE OFUXPSL 5P UIJT FOE XF
EFOF UIF QSPUPDPMT DIBSBDUFSJ[JOH UIF OFUXPSL NPEVMFT OBNFMZ TXJUDIFT BOE SPVUFST XIJDI BMMPX GPS JOUFHSBUJOH UIF (1-$ XJUI
&UIFSOFU EFWJDFT 4JODF &UIFSOFU SFQSFTFOUT B DPOWFSHFOU TUBOEBSE GPS NBOZ DPNNVOJDBUJPO EFWJDFT CZ BEEJOH UIJT GVODUJPOBMJUZ UP
(1-$ JOUFSDPOOFDUJWJUZ XJUI PUIFS IFUFSPHFOFPVT OPEFT DBO CF PFSFE 'VSUIFSNPSF TJODF UIF (1-$ NFEJVN BDDFTT DPOUSPM
MBZFS JT CBTFE PO B DBSSJFS TFOTF NVMUJQMF BDDFTT TDIFNF JUT QFSGPSNBODF EFDSFBTFT XIFO UIF OVNCFS PG OFUXPSL OPEFT DPOUFOEJOH
GPS UIF DIBOOFM JODSFBTFT FSFGPSF XF FWBMVBUF UIF OFUXPSL QFSGPSNBODF XIFO BO PQUJNJ[FE UJNF EJWJTJPO NVMUJQMF BDDFTT TDIFNF
JT BEPQUFE F QSPQPTFE DPOWFSHFOU OFUXPSL BSDIJUFDUVSF IBT CFFO JNQMFNFOUFE JO UIF 0./F5 OFUXPSL TJNVMBUPS
*OUSPEVDUJPO B TNBMM BNPVOU PG EBUB %FTQJUF UIF TNBMM BNPVOU PG EBUB

TUSJDU SFRVJSFNFOUT BT DPWFSBHF MBUFODZ EFMBZ BOE SPCVTUOFTT
&OFSHZ FDJFODZ BOE QPXFS TBWJOH XFSF JEFOUJFE JO NVTU CF GVMMMFE GPS TPNF BQQMJDBUJPOT GPS FYBNQMF UIF
BT GVOEBNFOUBM PCKFDUJWFT UP DPOUSJCVUF UP UIF TVTUBJOBCMF BVUPNBUJPO PG XJOEPXT BOE EPPST UIF DPOUSPM PG IFBUJOH
HSPXUI TQFDJFE JO UIF i&VSPQF w TUSBUFHZ <> "T B WFOUJMBUJPO BOE BJS DPOEJUJPOJOH )7"$
UIF NPOJUPSJOH PG
DPOTFRVFODF JO UIF OFBS GVUVSF UIF QPXFS HSJE OFFET UP QSFTFODFUFNQFSBUVSFHBTTNPLF UIF NBOBHFNFOU PG MPBET
CFDPNF B EJTUSJCVUFE MBSHFTDBMF TZTUFN UIBU IBT UP TNBSUMZ BOE TP GPSUI
NBOBHF PXT PG FMFDUSJDJUZ QSPEVDFE CZ CJH PS TNBMM QMBOUT *O UIJT DPOUFYU QPXFS MJOF DPNNVOJDBUJPO 1-$
JT B
UIBU JT B TNBSU HSJE 4(
5P UIJT SFTQFDU EFNBOE TJEF LFZ UFDIOPMPHZ TJODF B
UIF QPXFS MJOF HSJE JT QFSWBTJWFMZ
BOE EFNBOE SFTQPOTF NFDIBOJTNT IBWF UP CF JNQMFNFOUFE EFQMPZFE XJUIJO IPVTFTCVJMEJOHT BOE C
JU EPFT OPU SFRVJSF
TP UIBU QSPTVNFST XJMM BDUJWFMZ DPMMBCPSBUF JO UIF VTF BOE UIF VTF PG OFX XJSFT XIJDI JO UVSO SFEVDFT EFQMPZNFOU
EFMJWFSZ PG FOFSHZ < > 5P FOBCMF UIF 4( DPODFQU JU CFDPNFT DPTUT
UIFSFGPSF GVOEBNFOUBM UP DPOTJEFS UIF JOIPNFCVJMEJOH 1-$ EFWJDFT BSF JO HFOFSBM HSPVQFE JOUP UXP DBUFHPSJFT
DPOUFYU XIFSF CFTJEF UIF DPOWFOUJPOBM IPNF OFUXPSLJOH UIBU JT OBSSPXCBOE /#
BOE CSPBECBOE ##
EFWJDFT
TFSWJDFTGPS FYBNQMF USJQMF QMBZ IJHI TQFFE JOUFSOFU BDDFTT BDDPSEJOH UP UIF CJUSBUF UIBU UIFZ DBO PFS ##1-$ EFWJDFT
UFMFWJTJPO UFMFQIPOF
JOGPUBJONFOU BOE SFTPVSDF TIBSJOH JO BSF XJEFMZ EFQMPZFE UP PFS IJHI TQFFE IPNF OFUXPSLJOH
MPDBM BSFB OFUXPSLT -"/T
IPNF BVUPNBUJPO BOE TNBSU TFSWJDFT FZ XPSL JO UIF GSFRVFODZ CBOE o .)[ BOE
FOFSHZ NBOBHFNFOU TFSWJDFT IBWF UP CF PFSFE CFZPOE
BOE NBLF VTF PG BEWBODFE NPEVMBUJPO UFDIOJRVFT
%JFSFOUMZ GSPN IPNF OFUXPSLJOH TFSWJDFT IPNF BVUP UP PFS CJUSBUFT JO UIF PSEFS PG IVOESFET PG .CQT 0O UIF
NBUJPO BOE FOFSHZ NBOBHFNFOU TFSWJDFT VTVBMMZ JOWPMWF PUIFS IBOE /#1-$ EFWJDFT IBWF CFFO EFWFMPQFE UP PFS MPX
B MBSHF OVNCFS PG OPEFT TFOTPSTBDUVBUPST
UIBU BSF QFS CJUSBUFT GPS IPNF BVUPNBUJPO JOEPPS
BOE TNBSU NFUFSJOH
WBTJWFMZ EFQMPZFE XJUIJO UIF IPVTFCVJMEJOH BOE USBOTNJU PVUEPPS
TFSWJDFT FTF EFWJDFT VTVBMMZ XPSL JO B GSFRVFODZ
CBOE CFMPOHJOH UP UIF SBOHF PG o L)[ EFQFOEJOH PO MBZFS XJMM QSPWJEF B DPNNPO EBUB BOE DPOUSPM 4FSWJDF
UIF SFHVMBUJPOT BQQMJFE JO B TQFDJD DPVOUSZ GPS FYBNQMF "DDFTT 1PJOU 4"1
GPS UIF IPNF OFUXPSLJOH UFDIOPMPHJFT
$&/&-&$ "3*# '$$ %FUBJMFE JOGPSNBUJPO SFHBSEJOH ## EFTDSJCFE JO UIF *&&& 1 <> <> BOE
BOE /# 1-$ UFDIOPMPHJFT DBO CF GPVOE JO <> .P$" TQFDJDBUJPOT <> 'JOBMMZ B OVNCFS PG TUBOEBSE
$MFBSMZ CFTJEFT 1-$ UIFSF JT BMTP B CSPBE WBSJFUZ PG EFWFMPQNFOU HSPVQTBNPOH XIJDI ;JH#FF <> )PNF1MVH
XJSFMFTT BOE XJSFE UFDIOPMPHJFT UIBU DBO CF VTFE GPS CPUI IJHI <> BOE 8J'J BMMJBODFTBSF XPSLJOH PO UIF EFWFMPQNFOU
BOE MPX CJUSBUF JOIPNF BQQMJDBUJPOT <> GPS FYBNQMF 8J'J PG UIF ;JH#FF 4NBSU &OFSHZ WFSTJPO <> UIBU XJMM PFS
;JHCFF #MVFUPPUI ;8BWF UXJTUFE QBJS &UIFSOFU .P$" *1CBTFE DPOUSPM GPS BEWBODFE NFUFSJOH JOGSBTUSVDUVSFT BOE
BOE TP GPSUI IPNF BSFB OFUXPSLT BMMPXJOH GPS JOUFSPQFSBCJMJUZ XJUI
%FTQJUF UIF FYJTUFODF JO UIF NBSLFU PG UIJT CSPBE WBSJ BOE )PNF1MVH EFWJDFT
FUZ PG DPNNVOJDBUJPO UFDIOPMPHJFT UIF CPUUMFOFDL UP UIFJS
QFSWBTJWF EFQMPZNFOU BOE UIVT UP UIF KPJOU EFMJWFSZ PG $POUSJCVUJPO *O UIJT QBQFS XF DPOTJEFS UIF JOIPNF
EJFSFOU TFSWJDFT JT UIBU UIFZ PFO DBOOPU JOUFSDPOOFDU BOE TDFOBSJP XIFSF XF QSPQPTF UIF VTF PG B DPOWFSHFOU OFUXPSL UP
JOUFSPQFSBUF BOE JO TPNF DBTFT FWFO DPFYJTU FOIBODF UIF QFSGPSNBODF PG /# (1-$ UFDIOPMPHZ < >
UISPVHI JUT JOUFHSBUJPO JO BO &UIFSOFU OFUXPSL 8F UIVT GPDVT
4NBSU )PNF#VJMEJOH &OBCMFE CZ B $POWFSHFOU /FUXPSL PO UIF DPOWFSHFODF BU UIF %-- $MFBSMZ UIF JOUFHSBUJPO PG
'SPN UIF QSFWJPVT EJTDVTTJPO XF EFFN JNQPSUBOU UP SFBMJ[F UIF QSPQPTFE OFUXPSL JO B 5$1*1 OFUXPSL DBO CF FBTJMZ
B DPOWFSHFOU DPNNVOJDBUJPO OFUXPSL JO PSEFS UP JOUFHSBUF B PCUBJOFE CZ BEEJOH UIF DPSSFTQPOEJOH USBOTQPSUOFUXPSL
CSPBE WBSJFUZ PG DPNNVOJDBUJPO UFDIOPMPHJFT UIVT UP FOBCMF MBZFST *U TIPVME CF OPUFE UIBU (1-$ XBT NPTUMZ EFTJHOFE
UIF TNBSU IPNF 4)
DPODFQU UIBU JT UIF KPJOU EFMJWFSZ PG GPS PVUEPPS 4( BQQMJDBUJPOT 'VSUIFSNPSF JU IBT CFFO VTFE
IPNF OFUXPSLJOH IPNF BVUPNBUJPO BOE FOFSHZ NBOBHF BT B CBTFMJOF UFDIOPMPHZ GPS UIF EFWFMPQNFOU PG UIF <>
NFOU TFSWJDFT *&&& 1 BOE UIF *565 (IOFN TUBOEBSE GPS 4(
F DPOWFSHFOU DPNNVOJDBUJPO OFUXPSL DBO CF PCUBJOFE BQQMJDBUJPOT F EFQMPZNFOU PG (1-$ JO IPNFCVJMEJOHT
CZ QSPWJEJOH B DPOWFSHFOU MBZFS XIFSF IFUFSPHFOFPVT DPN JT JOEFFE JOUFSFTUJOH )PXFWFS JO PVS QSFWJPVT XPSL <>
NVOJDBUJPO UFDIOPMPHJFT DBO DPFYJTU BOE CF JOUFSDPOOFDUFE XF IBWF GPVOE UISPVHI FYQFSJNFOUBM UFTU DBNQBJHOT UIBU
6TVBMMZ UIF DPFYJTUFODF UIBU JT UIF BCJMJUZ PG TIBSJOH UIF UIF (1-$ UFDIOPMPHZ DBO FYIJCJU QPPS QFSGPSNBODF XJUIJO
TBNF QIZTJDBM NFEJVN CZ UXP PS NPSF EFWJDFT JT PCUBJOFE MBSHF IPVTFTCVJMEJOHT FSFGPSF JO UIJT QBQFS XF QSPQPTF
CZ FYQMPJUJOH EJFSFOU GSFRVFODZ CBOET BU UIF QIZTJDBM 1):
QPTTJCMF FOIBODFNFOUT GPDVTJOH BU UIF ."$ %%- TVCMBZFS
MBZFS PS UISPVHI UIF VTF PG NFEJVN BDDFTT DPOUSPM ."$
F STU DPOUSJCVUJPO PG UIJT QBQFS JT UIF EFOJUJPO PG
QSPUPDPMT F JOUFSDPOOFDUJWJUZOBNFMZ UIF DBQBCJMJUZ PG UIF QSPUPDPMT DIBSBDUFSJ[JOH UIF OFUXPSL NPEVMFT OBNFMZ
EFWJDFT UP FYDIBOHF EBUB BU 1): EBUB MJOL MBZFS %--
BOE TXJUDIFT BOE SPVUFST XIJDI BMMPX UIF JOUFHSBUJPO PG (1-$
OFUXPSL MBZFSSFRVJSFT DPFYJTUFODF BOE JU DBO CF BDIJFWFE BOE &UIFSOFU EFWJDFT JO B DPOWFSHFOU OFUXPSL BSDIJUFDUVSF
XJUI B DPOWFSHFOU MBZFS BCPWF UIF EBUB MJOL MBZFS PS BU UIF 3BOHF FYUFOTJPO JT BDIJFWFE CZ QBSUJUJPOJOH UIF PWFSBMM
OFUXPSL MBZFS $PFYJTUFODF BOE DPOOFDUJWJUZ BNPOH EJFSFOU OFUXPSL JOUP TVCOFUXPSLT UIBU JT (1-$ TVCOFUXPSLT "O
UFDIOPMPHJFT UISPVHI 1): PS EBUB MJOL MBZFS %--
NFDIB &UIFSOFU CBDLCPOF JO QBSUJDVMBS VTJOH DBU XJSJOH
QSPWJEFT
OJTNT IBWF CFFO UIF GPDVT PG NBOZ TUVEJFT 3FQSFTFOUBUJWF DPOOFDUJWJUZ CFUXFFO UIF TVCOFUXPSLT
FYBNQMFT BSF BT GPMMPXT UIF JOUFS1): QSPUPDPM EFWFMPQFE F TFDPOE DPOUSJCVUJPO SFHBSET UIF FWBMVBUJPO PG UIF
XJUIJO UIF *&&& 1 XPSLJOH HSPVQ UIBU BMMPXT DPFYJTUFODF DPOWFSHFOU OFUXPSL XIFO BEPQUJOH B UJNF EJWJTJPO NVMUJQMF
BNPOH UXP EJFSFOU CSPBECBOE 1-$ EFWJDFT <> UIF *56 BDDFTT 5%."
TDIFNF BU UIF ."$ TVCMBZFS PG UIF (1-$
5 (IO TUBOEBSE UIBU IBT CFFO DPODFJWFE XJUI UIF BJN OPEFT *O GBDU UIF (1-$ ."$ TVCMBZFS JT CBTFE PO UIF
PG PFSJOH JOUFSDPOOFDUJWJUZ BNPOH JOIPNF IJHI TQFFE OPOCFBDPOFOBCMFE NPEF PG UIF *&&& TUBOEBSE
DPNNVOJDBUJPO EFWJDFT XPSLJOH PWFS UFMFQIPOF XJSFT QPXFS XIJDI JT B DBSSJFS TFOTF NVMUJQMF BDDFTT $4."
TDIFNF "T
MJOFT BOE DPBY < > (IO TQFDJFT UIF 1): MBZFS BOE JU JT LOPXO UIF QFSGPSNBODF PG $4." TDIFNFT EFDSFBTFT BT
UIF ."$ TVCMBZFS BOE BEESFTTFT UIF DPFYJTUFODF CFUXFFO UIF OVNCFS PG UIF OFUXPSL OPEFT DPOUFOEJOH GPS UIF DIBOOFM
QSPUPDPMT UIBU PQFSBUF PO EJFSFOU NFEJB
UIF TPMVUJPO JODSFBTFT JT TJUVBUJPO DBO PDDVS JO MBSHF IPVTFTCVJMEJOHT
EFWFMPQFE XJUIJO UIF &6'1 0.&(" QSPKFDU BDDPSEJOH UP XIFSF B MBSHF OVNCFS PG TFOTPSTBDUVBUPST JT EFQMPZFE 5P
XIJDI EFWJDFT CFMPOHJOH UP UIF 0.&(" OFUXPSL TIBSF UIF DPQF XJUI UIJT QSPCMFN XF JNQMFNFOU B NPEJFE WFSTJPO PG
TBNF JOUFS."$ TVCMBZFS BOE DPOTFRVFOUMZ UIFZ DBO DPFYJTU UIF CFBDPOFOBCMFE NPEF PG UIF *&&& <> XIJDI
BOE UIFZ DBO CF JOUFSDPOOFDUFE <> JT CBTFE PO B 5%." ."$ NFDIBOJTN BOE XF TIPX UIBU
$POOFDUJWJUZ BNPOH IFUFSPHFOFPVT EFWJDFT DBO CF BMTP TVCTUBOUJBM QFSGPSNBODF JODSFBTF DBO CF PCUBJOFE
SFBDIFE BU OFUXPSL MBZFS CZ FYQMPJUJOH UIF *OUFSOFU QSPUPDPM F OFUXPSL QFSGPSNBODF IBT CFFO FWBMVBUFE XJUI UIF VTF
*1
F VTF PG UIF 5$1*1 QSPUPDPM TUBDL UP JOUFHSBUF EJFS PG UIF 0./F5 OFUXPSL TJNVMBUPS <>
FOU DPNNVOJDBUJPO UFDIOPMPHJFT GPS 4) BOE 4( BQQMJDBUJPOT F SFNBJOEFS PG UIF QBQFS JT BT GPMMPXT *O 4FDUJPO XF
JT MBSHFMZ BEWPDBUFE 1SFWJPVT XPSL UIBU GPMMPXT UIJT BQQSPBDI CSJFZ EFTDSJCF UIF (1-$ BOE UIF &UIFSOFU UFDIOPMPHJFT
JT EFTDSJCFE JO < o> BOE SFGFSFODFT UIFSFJO "MTP TUBO FO JO 4FDUJPO XF QSFTFOU UIF JNQMFNFOUBUJPO PG UIF DPO
EBSEJ[BUJPO XPSLJOH HSPVQT BSF DPOTJEFSJOH JU GPS FYBNQMF WFSHFOU OFUXPSL BOE JO 4FDUJPO XF TUVEZ JUT QFSGPSNBODF
BU UIF FOE PG *&&& MBVODIFE UIF 1 XPSLJOH HSPVQ 'JOBMMZ JO 4FDUJPO XF DPOTJEFS UIF 5%." JNQMFNFOUBUJPO
<> UIBU JT GPDVTFE PO UIF EFOJUJPO PG BO BCTUSBDUJPO MBZFS BOE XF DPNQBSF JUT QFSGPSNBODF XJUI $4." $PODMVTJPOT
GPS NVMUJQMF IPNF OFUXPSLJOH UFDIOPMPHJFT F BCTUSBDUJPO GPMMPX JO 4FDUJPO
5 (1-$ TZTUFN TQFDJDBUJPOT

$PNNVOJDBUJPO 5FDIOPMPHJFT
/VNCFS PG 140'%. TZNCPMT !
*O UIJT TFDUJPO XF EFTDSJCF UIF (1-$ BOE UIF &UIFSOFU /VNCFS PG *''5''5 QPJOUT !
UFDIOPMPHJFT UIBU XJMM CF VTFE UP CVJME UIF DPOWFSHFOU JO /VNCFS PG VTFE TVCDIBOOFMT !
IPNF OFUXPSL JO 4FDUJPO /VNCFS PG PWFSMBQQFE TBNQMFT !
/VNCFS PG DZDMJD QSFY TBNQMFT $1 !
/VNCFS PG '$) TZNCPMT '$) !
(1-$ 0SJHJOBMMZ EFWFMPQFE GPS "VUPNBUJD .FUFSJOH
4BNQMJOH GSFRVFODZ .)[
!
.BOBHFNFOU "..
BQQMJDBUJPOT (1-$ JT QMBZJOH B /VNCFS PG QSFBNCMF TZNCPMT XJUIPVU '$)
QSF !
SFMFWBOU SPMF JOBTNVDI JU IBT CFFO VTFE GPS UIF EFWFMPQNFOU
PG UIF *&&& 1 BOE UIF *565 (IOFN TUBOEBSE GPS 4(
BQQMJDBUJPOT <> *O UIF GPMMPXJOH XF EJTDVTT JUT 1): MBZFS Data
Preamble FCH n bytes RS parity 2FCS
byte
BOE ."$ TVCMBZFS
' (1-$ 1): GSBNF GPSNBU
1): -BZFS "DDPSEJOH UP <> UIF (1-$ UFDIOPMPHZ
IBT CFFO EFTJHOFE UP TVQQPSU $&/&-&$ "3*# BOE '$$
CBOET JO UIF GSFRVFODJFT SBOHF CFUXFFO L)[ BOE L)[ CBDLP UJNF *U JT XPSUI OPUJOH UIBU (1-$ EPFT OPU QSPWJEF
*U NBLFT VTF PG QVMTF TIBQFEPSUIPHPOBM GSFRVFODZ EJWJ TQFDJDBUJPOT GPS UIF IJHIFS *4004* MBZFST
TJPO NVMUJQMFYJOH 140'%.
DBSSZJOH %214, PS %#14, F ."$ JNQMFNFOUBUJPO TUBSUT GSPN UIF BTTVNQUJPO PG
TZNCPMT JO iOPSNBMw PS iSPCVTUw NPEF F DPSSFTQPOEFOU B NJOJNVN PWFSIFBE PG CZUFT FSFGPSF BDDPSEJOH UP UIF
NBYJNVN EBUB QBDLFU JT CZUFT XJUI %214, CZUFT BNPVOU PG EBUB CZUFT EFOFE JO UIF 1): UIF NBYJNVN ."$
XJUI %#14, JO iOPSNBMw NPEF BOE CZUFT XJUI %#14, JO QBZMPBE JT CZUFT JO OPSNBM NPEF BOE CZUFT JO SPCVTU
iSPCVTUw NPEF F NBYJNVN BDIJFWBCMF CJUSBUF JT LCQT NPEF 'PS UIF TBLF PG JNQMFNFOUBUJPO TJNQMJDJUZ XF BTTVNF
VTJOH %214, *U TIPVME CF OPUFE UIBU UIF EJFSFOUJBM FODPE UIBU UIF USBOTNJTTJPO EPFT OPU XBJU GPS BOZ DVSSFOU SFDFQUJPO
JOH UBLFT QMBDF BDSPTT UIF EBUB TUSFBN PG FBDI TVCDIBOOFM BDLOPXMFEHFNFOU "$,

F 1): GSBNF GPSNBU TFF 'JHVSF
JT DIBSBDUFSJ[FE CZ
J
UIF QSFBNCMF XIJDI JT B NVMUJTZNCPM FME VTFE UP QFSGPSN &UIFSOFU &UIFSOFU JT XJEFMZ EFQMPZFE PO -"/T *U PFST
DBSSJFS TFOTF PQFSBUJPOT UP FOBCMF DPOUSPM GVODUJPOT BOE UP B DPOWFSHFOU MPHJDBM MJOL DPOUSPM --$
TVCMBZFS GPS NBOZ
TZODISPOJ[F UIF SFDFJWFS BOE UIF USBOTNJUUFS JJ
UIF GSBNF EJFSFOU 1): MBZFST BOE ."$ TVCMBZFST GPS FYBNQMF DPBY
DPOUSPM IFBEFS '$)
XIJDI DBSSJFT DPOUSPM JOGPSNBUJPO UXJTUFE QBJS BT XFMM BT PQUJDBM CFS BOE XJSFMFTT "MUIPVHI UIF
SFRVJSFE UP DPSSFDUMZ EFNPEVMBUF UIF SFDFJWFE TJHOBM JJJ
UIF &UIFSOFU QSPUPDPM XBT PSJHJOBMMZ QSPQPTFE GPS VTF PWFS DPBY
EBUB QBZMPBE F EBUB QBZMPBE MFOHUI
EFQFOET PO UIF DBCMFT NPTU PG UPEBZT &UIFSOFU OFUXPSLT NBLF VTF PG UXJTUFE
USBOTNJTTJPO NPEF UIBU JT OPSNBM BOE SPCVTU *O OPSNBM QBJST BOE PQFSBUF JO GVMM EVQMFY NPEF UIF OFUXPSL JT TXJUDIFE
NPEF UIF GPSXBSE FSSPS DPSSFDUJPO '&$
JT QFSGPSNFE UIVT UIF DPOOFDUJPOT BSF IBOEMFE QPJOUUPQPJOU BOE DBOOPU
UISPVHI B 3FFE 4PMPNPO 34
FODPEFS BOE B DPOWPMVUJPOBM CF TIBSFE CZ NVMUJQMF EFWJDFT FSFGPSF UIF GVMM EVQMFY
FODPEFS XIFSFBT UIF FSSPS DIFDL JT EPOF XJUI B GSBNF DIFDL NPEF FMJNJOBUFT DBSSJFS TFOTF NVMUJQMF BDDFTT XJUI DPMMJTJPO
TFRVFODF '$4
*O SPCVTU NPEF CFTJEF 34 BOE DPOWPMVUJPOBM EFUFDUJPO $4."$%
NFDIBOJTNT CFDBVTF UIFSF JT OP OFFE
FODPEJOH UIFSF JT B SFQFUJUJPO DPEF 3$
UIBU SFQFBUT FBDI CJU UP EFUFSNJOF XIFUIFS UIF DPOOFDUJPO JT BMSFBEZ CFJOH VTFE
GPMMPXJOH UIF QSFBNCMF UJNFT "U UIF SFDFJWFS TJEF UIF IBSE *O UIJT QFSTQFDUJWF XF EFDJEFE UP DPOTJEFS UIF #"4&
EFDJTJPO 7JUFSCJ EFDPEFS BOE UIF 34 EFDPEFS BSF VTFE 59 UIBU JT UIF QSFEPNJOBOU GPSN PG 'BTU &UIFSOFU *&&&
8IFO TIPXJOH TJNVMBUJPO SFTVMUT XF DPOTJEFS (1-$ V
PWFS UXJTUFE QBJS DBCMFT DBU
PFSJOH CJUSBUFT VQ
XPSLJOH JO UIF $&/&-&$" CBOE EFTQJUF UIF GBDU UIBU UP .CQT .PSFPWFS TJODF #"4&59 JT GVMMEVQMFY JU
$&/&-&$" JT HFOFSBMMZ OPU VTFE GPS IPNF OFUXPSLJOH SFRVJSFT UIF VTF PG TXJUDIFT
QFSGPSNBODF JT OPU TJHOJDBOUMZ BFDUFE CZ UIF PQFSBUJOH
CBOE
*O UIJT DBTF JU VTFT 140'%. XJUI B SBJTFE DPTJOF
XJOEPX BOE ! TVCDIBOOFMT PVU PG XIJDI ! $POWFSHFOU /FUXPSL *NQMFNFOUBUJPO
BSF VTFE JO UIF o L)[ GSFRVFODZ CBOE 'VSUIFSNPSF *O UIJT QBQFS XF GPDVT PO UIF JOUFHSBUJPO PG (1-$
XF BTTVNF UIF 1): GSBNF UP MBTU ! 0'%. TZNCPMT UFDIOPMPHZ XJUIJO BO &UIFSOFU OFUXPSL 4JODF UIF TXJUDIFE
UIBU DBSSZ EBUB NPEVMBUFE XJUI %214, BOE SPCVTU %#14, &UIFSOFU OFUXPSL FYIJCJUT B TUBS UPQPMPHZ BOE UIF QPXFS
JT BTTVNQUJPO SFTQFDUJWFMZ MFBET UP ! PS ! MJOF DIBOOFM DBO CF DPOTJEFSFE BT B CVT XF BEPQU B USFFMJLF
CZUFT PG EBUB BOE PS 34 QBSJUZ DIFDL CZUFT $POTFRVFOUMZ UPQPMPHZBT B DPNCJOBUJPO PG CVT BOE TUBS UPQPMPHJFTUP
UIF NBYJNVN BDIJFWBCMF CJUSBUFT BSF LCQT BOE LCQT JNQMFNFOU UIF DPOWFSHFOU OFUXPSL TFF 'JHVSF
F NBKPS
5BCMF SFQPSUT UIF TFU PG UIF 1): MBZFS QBSBNFUFST CFOFU EFSJWJOH GSPN UIF VTF PG TVDI B UPQPMPHZ JT JUT BCJMJUZ UP
CF TDBMBCMF FYUFOTJCMF BOE SFMJBCMF F DPOWFSHFODF CFUXFFO
."$ 4VCMBZFS "DDPSEJOH UP <> UIF ."$ TVCMBZFS &UIFSOFU BOE (1-$ DBO CF PCUBJOFE UISPVHI UIF EFOJUJPO
JT CBTFE PO UIF *&&& TQFDJDBUJPO GPS MPX PG B TIBSFE DPNNPO MBZFS UIBU QSPWJEFT JOUFSDPOOFDUJWJUZ
SBUF XJSFMFTT QFSTPOBM BSFB OFUXPSLT 81"/T
<> #BTJDBMMZ BNPOH IFUFSPHFOFPVT MPXFS MBZFST
UIF DIBOOFM BDDFTT NFUIPE JT CBTFE PO UIF $4." XJUI F OFUXPSL DPOWFSHFODF JT BDIJFWFE CZ EFOJOH EJFS
DPMMJTJPO BWPJEBODF $4."$"
NFDIBOJTN BOE B SBOEPN FOU OFUXPSL EFWJDFT UIBU JT FOE OPEFT SPVUFST BOE TXJUDIFT
JT BCMF UP SFDPHOJ[F BOE EJTDBSE QBDLFUT EJSFDUMZ BSSJWFE GSPN

PUIFS SPVUFST
Ethernet cat5 4XJUDI F TXJUDI JT B XFMMLOPXO OFUXPSL EFWJDF

"T EFQJDUFE JO 'JHVSF UIF TXJUDI IBT CFFO NPEJFE JO
PSEFS UP XPSL TFBNMFTTMZ XJUI UIF SPVUFST *O QBSUJDVMBS JU
End node
Router JT BCMF UP CVJME BOE VQEBUF B GPSXBSEJOH UBCMF FYQMPJUJOH
OPEFT JOGPSNBUJPO IBSWFTUFE CZ FBDI SPVUFS " GPSXBSEJOH
Power line grid UBCMF FOUSZ JT DPNQPTFE PG TPVSDF BEESFTT JOTFSUJPO UJNF
Switch Ethernet cat5
MJOL RVBMJUZ BOE BSSJWBM QPSU OVNCFS FSFGPSF UIF TXJUDI
DPNQBSFT UIF JOGPSNBUJPO DBSSJFE CZ B QBDLFU XJUI UIF
Router End node DPSSFTQPOEFOU UBCMF FOUSJFT BOE GPSXBSET UIF QBDLFU UP UIF
DPSSFDU QPSU PS CSPBEDBTU JG UIF EFTUJOBUJPO BEESFTT IBT OP
Power line grid DPSSFTQPOEFODF JO UIF UBCMF
*O UIJT DBTF UIF JOTFSUJPO UJNF
QBSBNFUFS JT FYQMPJUFE UP SFNPWF BHFE FOUSJFT GSPN UIF UBCMF
' /FUXPSL UPQPMPHZ BOE UIVT JODSFBTJOH UIF TZTUFN GBVMU UPMFSBODF "HBJO UIF
TXJUDI JT BCMF UP QSFWFOU QBDLFU MPPQT 8F BMTP QPJOU PVU UIBU
TJODF MJOL RVBMJUZ JT VQEBUFE QFSJPEJDBMMZ UIF TXJUDI JT BCMF
&OE /PEFT FZ SFQSFTFOU UIF EFWJDFT PG UIF OFUXPSL
UP EZOBNJDBMMZ IBOEMF UIF OFUXPSL DIBOHFT *U JT OPX DMFBS
UIBU EJSFDUMZ JOUFSBDU XJUI UIF TVSSPVOEJOH FOWJSPONFOU GPS
UIBU UIF DPNCJOBUJPO PG UIF SPVUFS BOE TXJUDI QSPDFEVSFT
FYBNQMF TFOTPST BDUVBUPST TXJUDIFT NFUFST BOE TP PO
FOBCMFT UIF JOUFHSBUJPO PG IFUFSPHFOFPVT DPNNVOJDBUJPO
FTF OPEFT BSF HSPVQFE JOUP TVCOFUXPSLT BDDPSEJOH UP UIF
UFDIOPMPHJFT MFBEJOH UP B DPOWFSHFOU OFUXPSL .PSFPWFS
TBNF DPNNVOJDBUJPO UFDIOPMPHZ UIBU JT (1-$ 'SPN B
UIJT DPNCJOBUJPO QSPWJEFT UIF CBTJT GPS TBUJTGZJOH RVBMJUZ PG
MPHJDBM QPJOU PG WJFX UIF FOE OPEFT DBO CF BMM DIBSBDUFSJ[FE
TFSWJDF 2P4
DPOTUSBJOUT
CZ UIF TBNF CVJMEJOH CMPDLT UIBU JT B USBD HFOFSBUPS XIJDI
*U TIPVME CF OBMMZ OPUFE UIBU UIF JOUFHSBUJPO PG (
JT SFTQPOTJCMF PG EBUB QBDLFUT HFOFSBUJPO
BOE UIF OFUXPSL
1-$ XJUI &UIFSOFU FBTJMZ BMMPXT GPS JOUFHSBUJOH UIF (1-$
BEBQUFS F OFUXPSL BEBQUFS DPNQSJTFT B 1): B ."$ BOE
UFDIOPMPHZ JO *1 OFUXPSLT
B CVFS PG EBUB QBDLFUT DPNJOH GSPN UIF USBD HFOFSBUPS
3PVUFS 4JODF (1-$ EPFT OPU QSPWJEF BOZ TQFDJDBUJPO 4JNVMBUJPO 4FUVQ BOE 3FTVMUT
GPS UIF JOUFHSBUJPO JO B TXJUDIFE &UIFSOFU OFUXPSL XF OFFE B F DPOWFSHFOU OFUXPSL IBT CFFO JNQMFNFOUFE VTJOH UIF
OFUXPSL EFWJDF UIBU HSPVQT (1-$ OPEFT JOUP B TVCOFUXPSL 0./F5 TJNVMBUPS BOE JUT FYUFOTJPO */&5'SBNFXPSL
BOE JOUFHSBUFT UIF TVCOFUXPSL XJUI UIF SFTU PG UIF &UIFSOFU <> 0./FU JT BO PQFO TPVSDF DPNQPOFOUCBTFE TJN
OFUXPSL 5P EP UIBU XF EFOF B SPVUFS UIBU PFST OFUXPSL VMBUJPO MJCSBSZ BOE GSBNFXPSL UIBU JT QSJNBSJMZ UIPVHIU
BEBQUFST UPXBSET CPUI &UIFSOFU BOE (1-$ #FTJEF UIF GPS CVJMEJOH OFUXPSL TJNVMBUPST #BTJD DPNQPOFOUT UIBU JT
OFUXPSL BEBQUFST UIF 3PVUFS IBT B SPVUJOH NPEVMF UIBU JT TJNQMF NPEVMFT BSF QSPHSBNNFE JO $ BOE UIFO DPNCJOFE
SFTQPOTJCMF PG USBOTMBUJOH BOE GPSXBSEJOH QBDLFUT GSPN POF JOUP MBSHFS DPNQPOFOUT UIBU JT DPNQPVOE NPEVMFT VTJOH B
OFUXPSL BEBQUFS UP UIF PUIFS BOE WJDF WFSTB UIJT NPEVMF JT OFUXPSL EFTDSJQUJPO /&%
MBOHVBHF /&% MBOHVBHF JT BMTP
SFTQPOTJCMF PG JOUFSDPOOFDUJWJUZ CFUXFFO &UIFSOFU BOE ( VTFE UP DPOOFDU DPNQPVOE NPEVMFT BOE BTTFNCMF UIF XIPMF
1-$ *U JT XPSUI OPUJOH UIBU TJODF UIF NBYJNVN BMMPXFE OFUXPSL " TDSFFOTIPU PG UIF TJNVMBUFE OFUXPSL JT TIPXO JO
(1-$ 1): GSBNF TJ[F JT CZUFT DPSSFTQPOEJOH UP B 'JHVSF XIFSF XF BTTVNF B DBU DBCMF QSPQBHBUJPO EFMBZ PG
QBZMPBE PG EBUB CZUFT CPUI JO %214, BOE %#14, OPSNBM OT
NPEF
UIF SPVUFS FODBQTVMBUFT FBDI (1-$ GSBNF JOUP POF *O PSEFS UP RVBOUJGZ UIF DPOWFSHFOU OFUXPSL QFSGPS
&UIFSOFU GSBNF XIPTF NBYJNVN QBZMPBE EJNFOTJPO JT NBODF XF EFOF B SFQSFTFOUBUJWF NFUSJD OBNFMZ UIF BHHSF
CZUFT 0O UIF PUIFS TJEF &UIFSOFU GSBNFT FYDFFEJOH HBUF OFUXPSL UISPVHIQVU 5)3
*U JT FWBMVBUFE BT
CZUFT BSF GSBHNFOUFE CZ UIF SPVUFS JO PSEFS UP GVMMM (
1-$ DPOTUSBJOUT )PXFWFS UIJT BTTVNQUJPO JT OPU OFDFTTBSJMZ
PQUJNBM 5)3 ! 5)3 CQT
"T EFQJDUFE JO 'JHVSF UIF SPVUJOH NPEVMF LFFQT USBDF PG !

QBDLFUT SFDFJWFE GSPN JUT TVCOFUXPSL OPEFT BOE JU HFOFSBUFT
B GPSXBSEJOH UBCMF XJUI TPVSDF BEESFTT JOTFSUJPO UJNF BOE XIFSF JT UIF OVNCFS PG OFUXPSL OPEFT BOE 5)3 JT
MJOL RVBMJUZ *O UIJT QFSTQFDUJWF UIF SPVUFS EZOBNJDBMMZ MFBSOT UIF BWFSBHF UISPVHIQVU BDIJFWFE CZ UIF UI OPEF XIJDI JT
BCPVU UIF FYJTUFODF PG OPEFT EVSJOH SFDFQUJPO PG QBDLFUT BOE PCUBJOFE BT
NPEJFT JUT UBCMF VQEBUJOH UIF MJOL RVBMJUZ PS SFNPWJOH BHFE
FOUSJFT BDDPSEJOH UP UIF JOTFSUJPO UJNF *U JT XPSUI OPUJOH 5)3 !
CQT
UIBU B HJWFO TVCOFUXPSL DBO CF NBOBHFE CZ UXP PS NPSF
SPVUFST JO PSEFS UP FOTVSF B NPSF SFMJBCMF DPNNVOJDBUJPO PO XIFSF BSF UIF EBUB CZUFT FODBQTVMBUFE JO UIF 1): GSBNF
IBSTI QPXFS MJOF DIBOOFMT PS FRVBMMZ UP JODSFBTF UIF OFUXPSL
JT UIF OVNCFS PG DPSSFDU SFDFJWFE 1): GSBNFT QFS
DPWFSBHF 'VSUIFSNPSF JO PSEFS UP QSFWFOU MPPQT UIF SPVUFS TFDPOE CZ UIF UI OPEF XIJDI BSF EFUFDUFE FYQMPJUJOH FJUIFS
INIT
Frame No
received?
Yes
Process G3-PLC G3-PLC Ethemet Process ethemet

Arrival network
frame interface frame
Frame from Yes No destAddr in table or

another router? broadcast frame?
No Yes
No Yes Forward packet to

srcAddr in table?
G3-PLC interface
Add table entry Update table

(srcAddr (insTime, link
insTime, Quality)
linkQuality)
destAddr not in table No

or broadcast Discard frame
frame?
Yes
Forward packet to
ethemet interface
' 3PVUFS PX DIBSU
UIF LOPXMFEHF PG UIF USBOTNJUUFE GSBNF PS UIF '$4 FME XIFSF

JT UIF UPUBM OVNCFS PG USBOTNJUUFE GSBNFT EFTUJOFE
'VSUIFSNPSF XF EFOF UIF GSBNF FSSPS SBUF '&3
BT GPMMPXT UP UI VTFS *U JT XPSUI OPUJOH UIBU UIF '&3 UBLFT JOUP BDDPVOU
DPSSVQUFE GSBNFT BT XFMM BT NJTTFE GSBNFT

*O PSEFS UP NPEFM UIF QPXFS MJOF DIBOOFM JO UIF
'&3 !
0./F5 JNQMFNFOUBUJPO XF NBLF VTF PG UIF IBSEXBSF

INIT
No
Frame received?
Yes
Process frame
No Yes
srcAddr in table?
Add table entry

Yes No
(scrAddr, portNo, portNo equal to
insTime, linkQuality) table entry?
Update table Yes linkQuality No

(insTime, linkQuality) better than
table entry?
Update table (portNo,

Discard frame
insTime, linkQuality)
No
portNo equal to
arrival port?
Yes
Find correct No destAddr not in Yes Forward frame to all

destination port and table or broadcast ports except arrival
forward frame frame? port
' 4XJUDI PX DIBSU
QMBUGPSN <> UIBU DPOTJTUT PG B QBJS PG /#1-$ 0'%. JO B UISFFPPS IPVTF XIPTF FMFDUSJDBM QPXFS JT EJTUSJCVUFE
CBTFE NPEFNT XIPTF QBSBNFUFST BSF WFSZ TJNJMBS UP UIF POFT GSPN UIF NBJO QBOFM .1
UP FBDI PPS UISPVHI B PPS DJSDVJU
TQFDJFE CZ (1-$ TFF 4FDUJPO
8F IBWF QFSGPSNFE UXP CSFBLFS $#
MPDBUFE BU UIF .1 *O UIF MBUUFS DBTF XF IBWF
USJBM DBNQBJHOT DPOOFDUJOH BU FBDI UJNF UIJT QBJS PG NPEFNT DPOTJEFSFE FJUIFS UIF USBOTNJTTJPO CFUXFFO PVUMFUT CFMPOHJOH
UP UXP QPXFS TPDLFUT XJUIJO B IPVTF F STU DBNQBJHO UPPL UP UIF TBNF PPS PS CFUXFFO PVUMFUT CFMPOHJOH UP EJFSFOU
QMBDF JO B TJOHMF PPS IPVTF 8IFSFBT UIF TFDPOE UPPL QMBDF PPST "QQMJBODFT GPS FYBNQMF UFMFWJTJPO XBTIJOH NBDIJOF
g3node (3)
g3node (5)
g3node (2) plcgrid (1)
plcgrid (2)
g3plcrouter (1) g3node (4)
g3plcrouter (2)
Switch
Ethernode (0)
g3plcrouter (0)
Ethernode (1)
plcgrid (0)
g3node (1)
g3node (0)
' 4DSFFOTIPU PG BO FYBNQMF PG DPOWFSHFOU OFUXPSL TJNVMBUJPO JO 0./&5
Single floor house Multifloor house

1 1
0.9 0.9
0.8 0.8
0.7 0.7
0.6 0.6
)
0.5 0.5
(THR
(THR
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
0 0
0 10 20 30 0 10 20 30
(kbits/s) (kbits/s)
Normal DQPSK Normal DQPSK

Robust DBPSK Robust DBPSK
' $$%' PG UIF NFBTVSFE 5)3 GPS (1-$ UFDIOPMPHZ
CBUUFSZ DIBSHFS VPSFTDFOU MBNQT GSJEHF BOE TP PO IBWF EFHSBEFE GPS DIBOOFMT CFMPOHJOH UP UIF NVMUJPPS IPVTF JT
CFFO QMVHHFE BOE VOQMVHHFE JT TJNQMZ FYQMBJOBCMF PCTFSWJOH UIBU DIBOOFMT BTTPDJBUFE UP
'JHVSFT BOE SFTQFDUJWFMZ TIPX UIF DPNQMFNFOUBSZ NVMUJPPS IPVTFT DPWFS JO BWFSBHF MBSHFS EJTUBODFT UIBO UIPTF
DVNVMBUJWF EJTUSJCVUJPO GVODUJPO $$%'
PG UIF 5)3 BOE JO TJOHMF PPS IPVTFT BOE UIFZ FYQFSJFODF IJHIFS BUUFOVB
UIF DVNVMBUJWF EJTUSJCVUJPO GVODUJPO $%'
PG UIF '&3 GPS UJPO 'SPN UIF PCUBJOFE SFTVMUT XF DPNQVUFE UIF EJTUSJCVUJPO
OPSNBM BOE SPCVTU NPEF GPS CPUI UIF DPOTJEFSFE TDFOBSJPT PG UIF '&3 *O QBSUJDVMBS GPS UIF TJOHMF PPS IPVTF UIF '&3
"T XF DBO TFF CPUI UIF 5)3 BOE UIF '&3 BSF NVDI NPSF DBO CF BTTVNFE UP CF FYQPOFOUJBMMZ EJTUSJCVUFE XJUI NFBO
Single floor house Multifloor house

1 1
0.9 0.9
0.8 0.8
0.7 0.7
)
)
0.6 0.6
(FER
(FER
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
0 0
0 0.02 0.04 0.06 0 0.5 1
Normal DQPSK Normal DQPSK

Fitting normal DQPSK Fitting normal DQPSK
Robust DBPSK Robust DBPSK
Fitting robust DBPSK Fitting robust DBPSK
' $%' PG UIF NFBTVSFE '&3 GPS (1-$ UFDIOPMPHZ F EJTUSJCVUJPO UUJOH JT BMTP TIPXO
FRVBM UP BOE SFTQFDUJWFMZ GPS OPSNBM %214,

BOE SPCVTU %#14, NPEF 3FHBSEJOH UIF NVMUJPPS IPVTF 20
XF NPEFM UIF '&3 BT VOJGPSNMZ EJTUSJCVUFE JO UIF SBOHF
PS SFTQFDUJWFMZ GPS OPSNBM %214,
BOE SPCVTU %#14, NPEF F PCUBJOFE TUBUJTUJDT BSF EFQJDUFE 15
JO 'JHVSF SFE EPUUFE MJOFT
.PSF SFTVMUT SFHBSEJOH UIF UFTU
STH (kbps)
DBNQBJHOT BSF SFQPSUFE JO <>

*U JT JNQPSUBOU UP OPUF UIBU JO UIF DBTF PG B NVMUJPPS 10
IPVTF UIF (1-$ OFUXPSL DBO CF OBUVSBMMZ TQMJU JOUP TFWFSBM
TVCOFUXPSLT POF GPS FBDI PPS F TVCOFUXPSLT DBO UIFO
CF DPOOFDUFE UISPVHI &UIFSOFU F TBNF BSDIJUFDUVSF DBO CF 5
VTFE JO MBSHF CVJMEJOHT UIBU NBZ BMSFBEZ IBWF B XJSFE &UIFSOFU
EFQMPZNFOU
8F OPX UVSO PVS BUUFOUJPO UP UIF DPOWFSHFOU OFUXPSL 0
CFIBWJPS 5P EP UIBU XF DPOTJEFS UIF TBUVSBUJPO UISPVHIQVU 6 12 18 24 30 36 42 48 54 60
45)
XIJDI JT EFOFE BT UIF MJNJU SFBDIFE CZ UIF 5)3 XIFO Number of nodes
UIF PFSFE MPBE JODSFBTFT BOE JU SFQSFTFOUT UIF NBYJNVN
CSMA/CA (no routers)
MPBE UIBU UIF TZTUFN DBO DBSSZ <> *O 45) DPOEJUJPOT FBDI CSMA/CA (2 routers)
OPEF IBT JNNFEJBUFMZ B QBDLFU BWBJMBCMF GPS USBOTNJTTJPO CSMA/CA (3 routers)
BFS UIF DPNQMFUJPO PG FBDI TVDDFTTGVM USBOTNJTTJPO
' 4JNVMBUFE 45) GPS EJFSFOU OFUXPSL DPOHVSBUJPOT
/PX XF CVJME UIF TJNVMBUJPO TDFOBSJP VTJOH GSPN VQ
UP (1-$ OPEFT BOE XF FWBMVBUF UIF 45) XIFO OP
3PVUFST BSF JOUSPEVDFE BOE XIFO UXP BOE UISFF 3PVUFSTGPS 'JHVSF TIPXT UIF 45) "T XF DBO TFF UIF JOUSPEVDUJPO PG
FYBNQMF POF QFS FBDI PPSBSF DPOTJEFSFE F USBD BOE SPVUFST TVCTUBOUJBMMZ JNQSPWFT UIF QFSGPSNBODF *U JT
JT HFOFSBUFE BNPOH (1-$ OPEFT CFMPOHJOH UP UIF TBNF XPSUI OPUJOH UIBU UIF 45) JNQSPWFNFOUT EJSFDUMZ USBOTMBUF JO
TVCOFUXPSL PS UP EJFSFOU TVCOFUXPSLT JO %214, NPEF B DPWFSBHF JODSFBTF
Beacon Beacon
CAP CFP
GTS GTS Inactive
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
SD = aBaseSuperframeDuration 2BO symbols

(Active)
BI = aBaseSuperframeDuration 2BO symbols
' "O FYBNQMF PG TVQFSGSBNF TUSVDUVSF <>
5 4VQFSGSBNF QBSBNFUFST BOE WBMVFT F WBMVFT PG UIF QBSBNFUFST VTFE JO CFBDPOFOBCMFE

NPEF TJNVMBUJPOT BSF SFQPSUFE JO 5BCMF
1BSBNFUFST 7BMVF F (54 BTTJHONFOU JT EPOF CZ FBDI SPVUFS TPMWJOH UIF
B#BTF4VQFSGSBNF%VSBUJPO B#BTF4MPU%VSBUJPO GPMMPXJOH PQUJNJ[BUJPO QSPCMFN
B/VN4VQFSGSBNF4MPUT
B#BTF4MPU%VSBUJPO

B/VN4VQFSGSBNF4MPUT QE\
5)3

! 4UPU
40 ! #0

TU !
! 4UPU
'VSUIFS *NQSPWFNFOUT 6TJOH 5JNF

%JWJTJPO .VMUJQMF "DDFTT

5)3 5)3 !
4UPU
%FTQJUF UIF 45) JNQSPWFNFOUT SFMBUFE UP UIF JOUSPEVDUJPO PG
SPVUFST UIF CPUUMFOFDL PG UIF OFUXPSL JT SFQSFTFOUFE CZ UIF XIFSF JT UIF OVNCFS PG OPEFT JT UIF OVNCFS PG UJNF
EFHSBEBUJPO PG UIF QFSGPSNBODF XJUI UIF JODSFBTJOH OVNCFS TMPUT BTTJHOFE UP OPEF BOE ! ?
A

PG (1-$ OPEFT XJUIJO FBDI TVCOFUXPSL TFF 'JHVSF
'VSUIFSNPSF 5)3 JT UIF UISPVHIQVU PG OPEF JO CQT *U
5P QSFWFOU UIJT PDDVSSFODF XF FOWJTJPO UIF (1-$ DPN DBO CF PCUBJOFE BT 5)3 ! '&3 XJUI
NVOJDBUJPO UFDIOPMPHZ QSPQPTJOH B EJFSFOU DIBOOFM BDDFTT
EFOPUJOH UIF UJNF TMPU EVSBUJPO JO TFDPOET ? A
NFUIPE UIBU QSPWJEFT B IJHIFS 2P4 OBNFMZ UJNF EJWJTJPO
BSF 2P4 DPFDJFOUT FBDI JOEJDBUFT UIF QFSDFOUBHF PG UIF
NVMUJQMF BDDFTT 5%."
5P EP UIJT XF JNQMFNFOU UIF
UISPVHIQVU UIBU UIF UI OPEF IBT UP BDIJFWF XJUI SFTQFDU UP
CFBDPOFOBCMFE NPEF PG UIF *&&& TQFDJDB
UIF POF UIBU JU XPVME BDIJFWF JO UIF DPSSFTQPOEJOH TJOHMF VTFS
UJPOT <> *U EFOFT B TVQFSGSBNF TUSVDUVSF BT EFQJDUFE JO
TDFOBSJP 'JOBMMZ UIF DPOEJUJPO JO UIF TFDPOE MJOF PG
GPSDFT
'JHVSF &BDI TVQFSGSBNF JT DIBSBDUFSJ[FE CZ BO BDUJWF BOE
BMM UIF UJNF TMPUT JO B $'1 UP CF VTFE
BO JOBDUJWF QPSUJPO F BDUJWF QPSUJPO XIJDI JT EFOFE CZ
1SPCMFN
JT BO JOUFHFS MJOFBS QSPHSBNNJOH QSPCMFN
UIF TVQFSGSBNF EVSBUJPO 4%
DPOTJTUT PG B DPOUFOUJPO BDDFTT
FSFGPSF JU JT JO HFOFSBM /1 IBSE 5P TJNQMJGZ UIF QSPCMFN
QFSJPE $"1
BOE B DPOUFOUJPOGSFF QFSJPE $'1
*O UIF
XF TPMWF
VTJOH MJOFBS QSPHSBNNJOH -1
BOE XF SPVOE
$"1 QFSJPE EBUB QBDLFU USBOTNJTTJPO GPMMPXT B $4."$"
UIF PCUBJOFE DPFDJFOUT UP UIF MPXFS DMPTFTU JOUFHFS WBMVF
BMHPSJUIN XIJMF EVSJOH UIF $'1 UIF DIBOOFM BDDFTT JT CBTFE
$MFBSMZ UIFSF DPVME CF DBTFT XIFSF UIF OVNCFS PG TMPUT
PO 5%." F OFUXPSL DPPSEJOBUPS SFQSFTFOUFE CZ UIF
BTTJHOFE UP POF PS NPSF OPEFT JT [FSP *O UIFTF DBTFT UIF
SPVUFS BTTJHOT B HVBSBOUFFE UJNF TMPU (54
UP FBDI OPEF
DPSSFTQPOEFOU OPEFT BSF EFGFSSFE UP USBOTNJU JO UIF $"1
8F OPUJDF UIBU UIF DPPSEJOBUPS NBZ BTTJHO NPSF UIBO POF
'VSUIFSNPSF XIFO TPNF UJNF TMPUT BSF OPU PDDVQJFE BT SFTVMU
(54 UP FBDI TUBUJPO JO PSEFS UP TBUJTGZ HJWFO 2P4 DPOTUSBJOUT
PG UIF SPVOEJOH PG UIF DPFDJFOUT UIFTF XJMM CF BTTJHOFE UP
FSFGPSF UIF $'1 HSPXT PS TISJOLT EZOBNJDBMMZ GVMMMJOH UIF
UIF OPEFT UIBU IBWF UIF IJHIFTU UISPVHIQVU MFBWJOH UIF $"1
NJOJNVN $"1 MFOHUI TZNCPMT
8F GVSUIFS EFOPUF XJUI
4UPU UIF OVNCFS PG UJNF TMPUT QSFTFOU JO UIF $'1 GSFF PG USBOTNJTTJPOT 'JOBMMZ XF BTTVNF !

F
*U JT XPSUI OPUJOH UIBU UIF CFBDPO GSBNF XIJDI JT MBUUFS BTTVSFT UIBU UIF -1 BMXBZT HJWF B GFBTJCMF TPMVUJPO JG
QFSJPEJDBMMZ TFOU CZ UIF 81"/ DPPSEJOBUPS GPS TZODISPOJ[B 4UPU
UJPO QVSQPTFT DBO CF SFQMBDFE CZ TZODISPOJ[BUJPO WJB UIF *U JT FBTZ UP QSPWF UIBU XIFO !

! GPS FYBNQMF

FYQMPJUBUJPO PG UIF NBJOT DZDMF ! UIF PQUJNBM TPMVUJPO UP
DBO CF GPVOE
40 "DLOPXMFEHNFOUT
35 F XPSL PG UIJT QBQFS IBT CFFO QBSUJBMMZ TVQQPSUFE CZ UIF
'SJVMJ 7FOF[JB (JVMJB SFHJPO VOEFS UIF 103'&43 o
30 QSPHSBNNF QSPKFDU -",-JWJOH GPS "MM ,JUDIFO
25
STH (kbps)
3FGFSFODFT
20
<> &VSPQFBO $PNNJTTJPO &OFSHZ " 4USBUFHZ GPS $PNQFUJ
15 UJWF 4VTUBJOBCMF BOE 4FDVSF &OFSHZ
<> - -BNQF " 5POFMMP BOE % 4IBWFS i1PXFS MJOF DPNNV
10
OJDBUJPOT GPS BVUPNBUJPO OFUXPSLT BOE TNBSU HSJE w *&&&
$PNNVOJDBUJPOT .BHB[JOF WPM OP QQ o
5
<> " ;BCBMMPT " 7BMMFKP BOE + 4FMHB i)FUFSPHFOFPVT DPNNV
0 OJDBUJPO BSDIJUFDUVSF GPS UIF TNBSU HSJE w *&&& /FUXPSL WPM
6 12 18 24 30 36 42 48 54 60 OP QQ o
Number of nodes <> ) $ 'FSSFJSB - -BNQF + /FXCVSZ BOE 5 ( 4XBSU 1PXFS
CSMA/CA (no routers) TDMA (no routers) -JOF $PNNVOJDBUJPOT FPSZ BOE "QQMJDBUJPOT GPS /BSSPXCBOE
CSMA/CA (2 routers) TDMA (2 routers) BOE #SPBECBOE $PNNVOJDBUJPOT 0WFS 1PXFS -JOFT 8JMFZ
CSMA/CA (3 routers) TDMA (3 routers) 4POT /FX :PSL /: 64"
<> 7 $ (VOHPS % 4BIJO 5 ,PDBL FU BM i4NBSU HSJE UFDIOPMPHJFT
' 45) DPNQBSJTPO CFUXFFO $4."$" BOE 5%."
DPNNVOJDBUJPO UFDIOPMPHJFT BOE TUBOEBSET w *&&& 5SBOTBDUJPOT
PO *OEVTUSJBM *OGPSNBUJDT WPM OP QQ o
<> . . 3BINBO $ 4 )POH 4 -FF + -FF . " 3B[[BRVF BOE
JNQPTJOH UIF ,BSVTI ,VIO 5VDLFS ,,5
DPOEJUJPOT <> + ) ,JN i.FEJVN BDDFTT DPOUSPM GPS QPXFS MJOF DPNNVOJDB
BOE JU JT HJWFO CZ
! 4UPU GPS BMM ! *O
UJPOT BO PWFSWJFX PG UIF *&&& BOE *565 (IO TUBOEBSET w
UIF GPMMPXJOH XF BTTVNF UIF MBTU DPOEJUJPO IPMET USVF *&&& $PNNVOJDBUJPOT .BHB[JOF WPM OP QQ o
/PX JO 'JHVSF XF DPOTJEFS UIF DPNQBSJTPO CFUXFFO
UIF $4."$" BOE UIF PQUJNJ[FE 5%." JODSFBTJOH UIF <> )PNF(SJE 'PSVN (IO$PNQBUJCJMJUZ XJUI &YJTUJOH )PNF
OVNCFS PG OFUXPSL OPEFT 'SPN 'JHVSF XF OPUJDF UIBU /FUXPSLJOH 5FDIOPMPHJFT
UIF 5%." TDIFNF BMMPXT GPS B TVCTUBOUJBM JODSFBTF PG <> 7 4VSBDJ " $JNNJOP 3 $PMFMMB ( 0EEJ BOE . $BTUSVDDJ
UIF BHHSFHBUF OFUXPSL UISPVHIQVU XJUI SFTQFDU UP $4." i$POWFSHFODF JO IPNF HJHBCJU OFUXPSLT JNQMFNFOUBUJPO PG UIF
'VSUIFSNPSF JU TPMWFT UIF CPUUMFOFDL QSPCMFN PG $4." SFQ JOUFS."$ MBZFS BT B QMVHHBCMF LFSOFM NPEVMF w JO 1SPDFFEJOHT PG
SFTFOUFE CZ B DPOTJEFSBCMF EFHSBEBUJPO PG UIF QFSGPSNBODF UIF *&&& TU *OUFSOBUJPOBM 4ZNQPTJVN PO 1FSTPOBM *OEPPS BOE
.PCJMF 3BEJP $PNNVOJDBUJPOT 1*.3$
QQ o
XJUI UIF JODSFBTJOH OVNCFS PG OPEFT XJUIJO FBDI TVCOFUXPSL
*TUBOCVM 5VSLFZ 4FQUFNCFS
8F OBMMZ OPUJDF UIBU UIF NJOJNVN $"1 MFOHUI DPOTUSBJOU
TQFDJFE CZ UIF TUBOEBSE BFDUT UIF CFIBWJPS PG UIF <> 7 .JPSJ - 5BSSJOJ . .BODB BOE ( 5PMPNFJ i"O PQFO TUBO
EBSE TPMVUJPO GPS EPNPUJD JOUFSPQFSBCJMJUZ w *&&& 5SBOTBDUJPOT
BHHSFHBUF UISPVHIQVU *O GBDU UIF UISPVHIQVU FYIJCJUT B GBTUFS
PO $POTVNFS &MFDUSPOJDT WPM OP QQ o
EFDBZ XIFO DPOTJEFSJOH BOE SPVUFST XJUI SFTQFDU UP UIF
OP SPVUFS DBTF JT JT CFDBVTF UIF OFHBUJWF FFDU PG UIF $"1 <> 4 8PM 1 ( -BSTFO , -BVTEBIM " 3JCFJSP BOE 5 4
5PFHBBSE i'BDJMJUBUJOH IPNF BVUPNBUJPO UISPVHI XJSFMFTT
JT QSFTFOU JO FBDI TVCOFUXPSL OBNFMZ UXJDF PS UISFF UJNFT
QSPUPDPM JOUFSPQFSBCJMJUZ w JO 1SPDFFEJOHT PG UIF *OUFSOBUJPOBM
SFTQFDUJWFMZ 4ZNQPTJVN PO 8JSFMFTT 1FSTPOBM .VMUJNFEJB $PNNVOJDBUJPOT
4FQUFNCFS
$PODMVTJPOT <> % &MTIBOJ BOE 1 'SBODR iF BOBUPNZ PG B VOJWFSTBM EPNPUJDT
JOUFHSBUPS GPS HMPCBMMZ JOUFSDPOOFDUFE EFWJDFT w JO 1SPDFFEJOHT PG
(1-$ EFWJDFT TIPX QPPS QFSGPSNBODF XIFO XPSLJOH PWFS UIF *OUFSOBUJPOBM $POGFSFODF PO 1FSWBTJWF 4ZTUFNT BOE $PNQVU
MBSHF JOIPNFCVJMEJOH 1-$ OFUXPSLT XIPTF MJOLT DBO CF JOH QQ o +VOF
TFWFSFMZ BUUFOVBUFE " DPOWFSHFOU OFUXPSL DPOTJTUJOH PG <> 4 5PNQSPT / .PVSBUJEJT . %SBBJKFS " 'PHMBS BOE )
&UIFSOFU BOE (1-$ EFWJDFT DBO CF CVJMU VTJOH SPVUFST )SBTOJDB i&OBCMJOH BQQMJDBCJMJUZ PG FOFSHZ TBWJOH BQQMJDBUJPOT
BOE TXJUDIFT OPEFT UP DPQF XJUI UIJT QSPCMFN SPVHI PO UIF BQQMJBODFT PG UIF IPNF FOWJSPONFOU w *&&& /FUXPSL WPM
TJNVMBUJPO SFTVMUT PCUBJOFE JNQMFNFOUJOH UIF DPOWFSHFOU OP QQ o
OFUXPSL JO UIF 0./&5 TJNVMBUPS XF IBWF TIPXO UIF <> *&&& 1 % *&&& %SB 4UBOEBSE GPS " $POWFSHFOU
GFBTJCJMJUZ PG TVDI B TPMVUJPO "FS UIBU XF IBWF BOBMZTFE %JHJUBM )PNF /FUXPSL GPS )FUFSPHFOFPVT 5FDIOPMPHJFT *&&&
UIF CFIBWJPS PG UIF $4."$" QSPUPDPM JNQMFNFOUFE JO UIF
(1-$ EFWJDFT BOE XF IBWF GPVOE UIBU JUT QFSGPSNBODF <> *&&& 4UBOEBSET "TTPDJBDJPO *&&& 4UE 1BSU
EFDSFBTFT BT UIF OVNCFS PG OFUXPSL OPEFT JODSFBTFT 5P 8JSFMFTT -"/ .FEJVN "DDFTT $POUSPM ."$
BOE 1IZTJDBM
DPQF XJUI UIJT QSPCMFN XF IBWF QSPQPTFE UP JNQMFNFOU B -BZFS 1):
4QFDJDBUJPOT *&&&
5%." TDIFNF BOE XF IBWF TFFO UIBU JU TFOTJCMZ JNQSPWFT <> - $IBOH 8IJUF 1BQFS.BLJOH PG *&&& $PNQMJBOU
UIF OFUXPSL QFSGPSNBODF &RVJQNFOU /BUJPOBM 4FNJDPOEVDUPS
<> .VMUJNFEJB PWFS $PBY "MMJOBDF .P$" 4QFDJDBUJPO GPS

%FWJDF 3' $IBSBDUFSJTUJDT
<> ;JH#FF "MMJBODF ;JH#FF BOE 8JSFMFTT 3BEJP 'SFRVFODZ $PFYJT
UFODF
<> )PNF1MVH 1PXFSMJOF "MMJBODF )PNF1MVH "7 8IJUF 1BQFS
<> ;JH#FF "MMJBODF BOE )PNF1MVH 1PXFSMJOF "MMJBODF MJBJTPO
4NBSU &OFSHZ 1SPMF 7FSTJPO 5FDIOJDBM 3FRVJSFNFOUT
<> &3%' 1-$ ( 1IZTJDBM -BZFS 4QFDJDBUJPO
<> &3%' 1-$ ( ."$ -BZFS 4QFDJDBUJPO
<> 7 0LTNBO BOE + ;IBOH i()/&. UIF OFX *565 TUBOEBSE
PO OBSSPXCBOE 1-$ UFDIOPMPHZ w *&&& $PNNVOJDBUJPOT .BH
B[JOF WPM OP QQ o
<> - % #FSU 4 %"MFTTBOESP BOE " . 5POFMMP i"O *OUFS
DPOOFDUJPO BQQSPBDI BOE QFSGPSNBODF UFTUT GPS JOIPNF 1-$
OFUXPSLT w JO 1SPDFFEJOHT PG UIF *&&& *OUFSOBUJPOBM 4ZNQPTJVN
PO 1PXFS -JOF $PNNVOJDBUJPOT BOE *UT "QQMJDBUJPOT #FKJOH
$IJOB .BSDI
<> *&&& 4UE 1BSU 8JSFMFTT .FEJVN "DDFTT
$POUSPM ."$
BOE 1IZTJDBM -BZFS 1):
4QFDJDBUJPOT GPS -PX
3BUF 8JSFMFTT 1FSTPOBM "SFB /FUXPSLT 81"/T
*&&&
<> " 7BSHB 0./F5 6TFS .BOVBM
<> */&5 'SBNFXPSL 'PS 0./F5 .BOVBM
<> ."9 *OUFHSBUFE 1PXFS-JOF %JHJUBM 5SBOTDFJWFS 1SPHSBN
NJOH .BOVBM .BYJN *OUFHSBUFE 1SPEVDUT
<> ( #JBODIJ i*&&& TBUVSBUJPO UISPVHIQVU BOBMZTJT w *&&&
$PNNVOJDBUJPOT -FUUFST WPM OP QQ o
<> 4 #PZE BOE - 7BOEFOCFSHIF $POWFY 0QUJNJ[BUJPO $BNCSJEHF
6OJWFSTJUZ 1SFTT
doi:10.1155/2012/452402
Research Article
Improved Maximum Likelihood S-FSK Receiver for
PLC Modem in AMR
Mohamed Chaker Bali and Chiheb Rebai

GRESCOM Research Laboratory, SUPCOM, University of Carthage, City of Communications Technologies,
2083 El Ghazala Ariana Tunis, Tunisia
Correspondence should be addressed to Chiheb Rebai, chiheb.rebai@supcom.rnu.tn
Received 3 August 2012; Accepted 26 October 2012
Academic Editor: Justinian Anatory
Copyright 2012 M. C. Bali and C. Rebai. This is an open access article distributed under the Creative Commons Attribution
cited.
This paper deals with an optimized software implementation of a narrowband power line modem. The modem is a node in
automatic meter reading (AMR) system compliant to IEC 61334-5-1 prole and operates in the CENELEC-A band. Because of the
hostile communication environments of power line channel, a new design approach is carried out for an S-FSK demodulator
capable of providing lower bit error rate (BER) than standard specications. The best compromise between eciency and
architecture complexity is investigated in this paper. Some implementation results are presented to show that a communication
throughput of 9.6 kbps is reachable with the designed S-FSK modem.
1. Introduction Despite its obvious advantages, AMR have not been yet
rolled out signicantly. Actually, a major broad deployment
International concerns about natural environment preserva- inconvenient of smart meters was the lack of reliability on
tion have been increasingly serious during the last decades. hostile communication environments of power line channel.
In fact, one of the most ecologically inuencing factors is In fact, early implementations of PLC modems were basic
energy. Besides, energy consumption rise was unexpectedly on ordinary amplitude shift keying (ASK) or frequency shift
important and quick, neglecting eciency and ecological
keying (FSK) techniques.
considerations. These facts have pushed several countries to
try to change their energy consumption policies. In this paper, we investigate the importance of spread
The widest idea behind operating energy eciently frequency shift keying (S-FSK) modulation scheme to make
is called SmartGrid. This concept, as its name suggests, transmissions robust against narrowband noise and attenua-
involves integrating intelligence into the whole power grid; tion in such hostile channel. Hence, an intelligent power line
generation, transmission, distribution, and management communication (PLC) modem solution for automatic meter
are concerned. The goal is to increase power generation, reading using International Electrotechnical Commission
transmission, distribution, and usage eciency by reducing (IEC) S-FSK prole is simulated and implemented using
power waste, favoring renewable energies, and sensitizing digital signal processor (DSP) [2].
consumers about their actual consumption [1]. The paper is organized as follows. In Section 2 we will
This big concept was only expressed lately after arise start by presenting PLC-based automatic meter reading
of more specic and actually applicable ideas. The rst solution. The chosen S-FSK prole is briey introduced.
is automatic meter reading (AMR), enabling automated A description of the proposed S-FSK receiver is presented
remote meter reading. Later were introduced automatic in Section 3. In Section 4 we focus on implementation of
meter infrastructure (AMI) and automatic meter Manage- S-FSK modulation scheme using DSP architecture. The
ment (AMM), which are two expansions providing more eciency of the proposed design was illustrated by some
consumer- and management-oriented services. implementation results that show the performances of the
Manual
Internet
Consumer
PSTN
Wireless
Analysis and Communication PLC

Database
management front end
Figure 1: Dierent automatic meter reading techniques.
realized PLC Modem. Finally some conclusions are outlined customers better control of their usage of electric energy,
in Section 5. gas, or water. From the other side, predicting energy usage
remains a key advantage for energy distributors. With AMR,
distributors can get accurate information of consumption
2. PLC Modems for AMR prole of each consumer and monitor the network in order
The evolution of meter reading has been outstanding during to prevent or capture defects.
the last decades. Several power suppliers, distributers jointly The advantages of AMR are several and obvious:
with their technological partners, have tried several novel (i) increasing meter reading and billing accuracy and
approaches in order to automate meter reading. security;
The evolution from traditional manual meter reading to
actual and future intelligent infrastructures passing through (ii) permitting a exible tari changing;
e-meters, semiautomatic meter reading, and fully automated (iii) giving user the control over its consumption;
meter reading gave these actors a great experience in this
ever-evolving eld. (iv) enabling a better grid monitoring and load manage-
Despite the abundance of the available technologies, ment;
power line communication has been agreed to be the best (v) remote power disconnection and reconnection.
t for last-mile meter reading and meter management
communication. In fact, this technology has one of the lowest Automatic meter reading system is summarized by
costs and is easily set up. Moreover, the technology is now Figure 1. Meters data are collected using one of the available
considered as suciently ripe to be widely deployed. ways of communication into a database. This database is
PLC, as a technology, is very wide. A myriad of tech- then accessible for analysis and management purposes in the
niques are available using dierent modulation techniques information system center. A subset of these data can also be
and dierent protocols. From another side, the regulation accessed by customers using dedicated services.
is still under work. Nevertheless, some proles have already Several automatic meter reading technologies can be
been standardized and are being adopted by the market. The used depending on grid topology. Most important ones are
IEC S-FSK prole, for example, is actually one of the most as follows
used for AMR because it proved its simplicity and maturity.
In this section, we will briey introduce automatic meter (i) handheld, walk-by, and drive-by AMR;
reading concepts, then present PLC from both technical and (ii) public switched telephone network-based AMR;
technological sides, and nally give a short survey on S-FSK
PLC modems. (iii) wireless communication-based AMR;
(iv) power line communication-based AMR.
2.1. Automatic Meter Reading. Automatic meter reading is a
technique used to collect data from electricity, gas, water, or 2.2. Power Line Communication-Based AMR. Power line
other utility meters. Unlike manual meter reading, automatic communication consists of the use of the power lines as
meter reading relies on communication technologies to a physical communication medium. PLC has been used
collect users consumption. Meters send data automatically for data transfer for both indoor and outdoor networks.
through a communication network to the management Anyhow, the prole of these applications is dierent.
system. Collected data can be then transferred to a central Concerning PLC use for AMR and outdoor communi-
database to be analyzed and used for billing. This means that cation, PLC is the most approved technology by electricity
billing can be based on actual consumption rather than on distributors. In fact, electric network is already well expanded
an estimate based on previous consumption statistics, giving and oers a great coverage. Thus, no additional wireless
or wired communication medium needs to be used and Mains 50 Hz/220 V

deployment costs are then cut.
The main idea behind PLC is the use of the power
line to carry radio frequency signals. Actually, a low power
modulated signal containing information is added to the Slot indicator Slot indicator
electric signal. Data then propagates over the electric network Preamble Delimiter P SDU Pause Time
and is detected by remote stations.
2 bytes 2 bytes 38 bytes 3 bytes
Several modulation techniques can be used to transport
data over power lines. But most of them are based on Physical frame: 45 bytes = 150 ms at 2400 bps
frequency modulation. Actually, data is converted to a higher
Figure 2: Physical frame structure.
frequency signal which is superimposed to the 50 Hz/60 Hz
electrical signal. The signal is then repeatedly transmitted
over the network until it reaches the destination node. 2.3. PLC Modem Based on S-FSK Prole. The communica-
Several PLC communication proles have been proposed tion prole described by the IEC 61334-5-1 standard is based
and each prole is essentially based on the modulation on the S-FSK modulation technique.
scheme chosen. S-FSK is a modulation and demodulation technique
Two dierent modulation scheme classes can be distin- which combines some of the advantages of a classical
guished: spread spectrum system (e.g., immunity against narrowband
interferers) with the advantages of a classical FSK system
(i) single-carrier modulations; (low-complexity, well-investigated implementations).
(ii) multicarrier modulations. As the classical FSK, S-FSK uses two frequencies to
transmit binary information at each bit time. By spreading
The rst technique is the simpler one. It uses a nar-
the two used frequencies, S-FSK makes these two channels
row frequency band for data transfer. Examples of these
independent. This characteristic is then used by the demod-
modulation schemes are FSK, S-FSK, and continuous phase ulator and ensures a better reception quality than FSK. In
frequency shift keying (CPFSK). These modulation schemes fact, if signal qualities of the two channels are close, the
are often chosen for their maturity and implementation demodulator makes the decision by comparing the signal
simplicity. Though, they do not oer great data transfer on both channels. Otherwise, demodulation is based only
rates. Actually, data rates generally range from 300 bps to on the channel having a better reception quality. Channel
2.4 kbps [3]. Narrowband PLC has been receiving widespread quality estimation is calculated using a predened preamble
attention due to its applications in the SmartGrid. preceding transmitted data.
The second uses multiple adjacent carriers in order Synchronization in this prole is based on zero crossing
to transfer data. Usually orthogonal frequency division detection of electric signal. Therefore, the transmission and
multiplexing (OFDM) or a derived modulation scheme is reception must start on the main zero cross. Due to the phase
used. These modulation schemes are applied in order to shift between 50 Hz and the carrier, the zero cross signal may
increase raw data throughput and/or to cope with harsh provide incorrect timing of the bit-wise synchronization.
channel conditions. Broadband PLC is seen as an exciting To recover this delay the bit synchronization adjustment
and eective technology for multimedia distribution within method is implemented in the modem software. This
homes. algorithm based on correlation method can move bit border
during reception.
In either case, AMR PLC technique must handle poor
Time is divided into system wide synchronized time slots,
channel quality. In fact, outdoor power lines are exposed
and physical frames are only transmitted with the beginning
to several noise sources. Furthermore, power lines present
of timeslots.
highly varying impedance due to topology changes and high
Timeslot synchronization is achieved using detection of
attenuation. Hence, power line channel quality is considered any frames preamble and delimiter as described in Figure 2.
as time, space, and frequency dependent [4]. After physical synchronization, each station must keep track
In order to overcome these impairments, high perfor- of slot indicator using an internal clock.
mance processing is unavoidable. This includes channel As described earlier, communication between meters
estimation and equalization, strong forward error correction and management system is done through a special node
algorithms, and signal repetition. called access node usually placed at the medium/low voltage
In addition to noise and channel quality diculties, PLC- (MV/LV) transformer stations. Access nodes are specic
based AMR has two other main challenges. The rst is nodes that manage the communication on a specic meter
that unlike usual communication techniques where transfer network.
speed is the most signicant criteria, cost and reliability This prole uses a master/slave communication para-
are the most important factors in AMR. The second is the digm based on polling mechanism. In fact, meters can only
existence of a lot and very dierent protocols and standards, respond to queries made by master station. This method
their specic underlying problems, and interoperability combined with slotted time simplies considerably medium
issues. access control.
Power line
DSP processor Mixed front end Coupling stage

DAC BPF Line driver
TX
ADC VGA BPF

RX
IO
Zero crossing
Synchronisation stage
Figure 3: PLC modem functional block diagram.
The modem that we propose in this paper is an AMR PLC A digital signal waveform with binary signaling consists
modem using IEC61334-5-1 compliant prole and operates of two kinds of signals s0 (t) and s1 (t) for nT t (n + 1)T,
in the CENELECA band [5]. It is based on three main n is a positive integer:
stages as described in Figure 3; DSP processor, mixed front
end, and a coupling interface: H0 : s0 (t) = A sin 2 f0 t ,
(1)
(i) digital stage including DSP processor and external H1 : s1 (t) = A sin 2 f1 t ,
memories. DSP processor provides exible software
implementation and easily upgrade to new software where A is a real constant.
version or merging standards; A frequency selective channel with an additive nonwhite
Gaussian noise is considered; however, the channel gain Gi
(ii) mixed front end based on digital to analog converter
and the noise power spectral density Ni are assumed to be
(DAC) and line driver for transmitter section, analog
at around the frequency fi . Therefore, at the receiver input,
to digital converter (ADC) and variable gain ampli-
the signal-to-noise ratio (SNR) for the channel i is
er (VGA) for receiver section, and external band-
pass lter (BPF); A2 G2i /2
SNRi = , with i {0, 1}. (2)
(iii) coupling interface makes connection between the Ni /T
mixed front end and the power lines. It provides pro-
tection from high voltage and peak voltage/current, The SNRi completely characterize the quality of the received
attenuation of 50/60 Hz signal, impedance matching signal. Moreover, another characterization of the quality of
to the mains for both transmitter and receiver paths, the received S-FSK signal may be made through the unbal-
and nonisolated power supply. ancing factor x and the average signal-to-noise ratio SNRav .
This last term is dened as the ratio of the signal energy
The use of a DSP permits a greater control over the signal and the average noise power densities. These parameters are
processing stage and a greater exibility of the implemented related to (2) as follows:
S-FSK modem.
SNR0 SNR1 SNR1
SNRav = 2 , x= . (3)
SNR0 + SNR1 SNR0
3. S-FSK Modulation Technique
This section details S-FSK modulation principle and gives 3.2. The Maximum Likelihood S-FSK Receiver. In practical
theory and simulations of suboptimum receiver. channels, the received signal phase is very dicult or even
impossible to track. Thus, the detection process may have to
3.1. The S-FSK Principle. S-FSK modulation consists of a disregard the phase information to avoid complex circuits,
binary FSK modulation in which the frequency deviation f at some expense of performance degradation. This is called
is large enough to generate a spectrum with two separate noncoherent detection [8, 9].
lobes. For this reason, the concept of dual channel is Using the channel model early presented the received
introduced: channel 0 refers to the signal placed around a signal under hypotheses H0 and H1 is
frequency f0 and channel 1 refers to the signal placed around
G0 s0 (t, ) + n0 (t) under H0 ,
a frequency f1 , with f = | f1 f0 |/2 [6, 7]. r(t) = (4)
The symbols to be transmitted are generated with a G1 s1 (t, ) + n1 (t) under H1 ,
rate 1/T, where T is the symbol period, and belong to
the alphabet {1, +1}. Therefore, binary hypotheses H0 where si (t, ) is the signal with an unknown phase and ni (t)
and H1 can be associated with 0 and 1 being transmitted, is the white Gaussian noise with zero mean and a noise power
respectively. spectral density Ni /2, with i {0, 1}.
The unknown phase is random with a power density 3.3. Improved ML S-FSK Receiver. Implementation of the
function p (). We assume that is uniformly distributed ML receiver is dicult due to the complexity of formulae
on [0, 2], that is, from (8) to (9). An improved method of estimating log-
1 likelihood metric is proposed for a practical realization.
p () = , 0 2. (5) In order to describe the receiver, the log-likelihood ratio
2
li (ri ) of the signal ri is introduced:
The correlation receiver correlates the input signal r(t)

with a stored replica of the signal si (t). The outputs ri p(ri | H1 )
are necessary to discriminate whether +1 or 1 has been li (ri ) = log . (11)
p(ri | H0 )
transmitted.
The modulus of the envelop detectors outputs may be Using the distributions (8)(9), (11) can be simplied
modeled as follows for two orthogonal S-FSK signals: into the following equation:

| n0 | under H1
r0 = 2ri i 2
|s0 + n0 | under H0 , li (ri ) = (2i 1) log I0 i2 . (12)
i2 i
(6)
| n1 | under H0 Logarithm and Bessel function are approached using
r1 = approximating function. Let g() be a piecewise linear
|s1 + n1 | under H1 ,
approximation of the composed function log(I0 ()) dened
where ni is an additive circularly Gaussian noise with zero as
mean and variance i2 , with i {0, 1}.
Under the assumption that the noise is Gaussian, the g(X) = A j X + B j . (13)
sampled outputs of the envelope detectors r0 and r1 are
Rician or Rayleigh distributed depending on which of the The approximation is dened over M intervals I1 , I2 , . . .,
two signals s0 (t) and s1 (t) is transmitted. IM . A j and B j are calculated by imposing g() to be equal to
Under hypothesis H0 , the probability density function log(I0 ()) on the boundary of each interval that denes the
p(ri | H0 ) of the amplitude of the signal ri with i {0, 1} piecewise approximation:
is

g(X) = log(I0 (X)) with X 0, 21 , 22 , . . . . (14)
2r 2r0 a0 r 2 + 2
p(r0 | H0 ) = 20 I0 exp 0 2 0 ,
0 02 0 Using (14) in (12), an approximated estimation of the
(7) loglikelihood li (ri ) ratio is obtained with the equation:
2r1 r12
p(r1 | H0 ) = 2 exp 2 .
1 1 2ri i 2
li (ri ) = (2i 1) g i2 . (15)
Under hypothesis H1 , the probability density function i2 i
p(ri | H1 ) of the amplitude of the signal ri with i {0, 1}
The proposed receiver decides accordingly to (10) on the
is
following decision values:
2r0 r02
p(r0 | H1 ) = 2 exp 2 , p(r | H1 ) = l1 (r1 ), p(r | H0 ) = l0 (r0 ).
0 0 (16)
(8)
2r1 2r1 a1 r2 + 2 Assuming to have knowledge of the rst P symbols
p(r0 | H1 ) = 2 I0 2 exp 1 2 1 , creating the Preamble (alternative 1 and 0 symbols), the
1 1 1
T T channel and noise parameters may be estimated using the
where 0 = 0 s0 (t)2 dt and 1 = 0 s1 (t)2 dt. I0 () is the signals (6) as follows:
modied Bessel function of the rst kind of order 0.
Assuming the symbols {+1, 1} to be transmitted with 2 2
02 = r0 (k)2 , 12 = r1 (k)2 ,
the same probability and to deal with independent noises P kH P kH
1 0
n0 and n1 (typical in the S-FSK modulation), the maximum

likelihood (ML) decision turns out to be the optimum 2
decision rule [10]. 20 = 02 +
r0 (k)
2
, (17)
P kH
In particular, the decision rule uses the following decision 0

values:
2 2
= .
2
p(r | H0 ) = p(r0 | H0 )p(r1 | H0 ), 21

1 P+ r1 (k)
(9) kH1

p(r | H1 ) = p(r0 | H1 )p(r1 | H1 ).
The decision rule is to compare likelihood functions and 3.4. Simulations Results. The performance of dierent
choose the largest: receiver is compared through communication schema imple-
mentation using Matlab. A packet-based transmission has
detect 0 p(r | H0 ) p(r | H1 ) detect 1. (10) been adopted, with preamble length P equal to 32 and
100 100
102 102
BER
BER
104 104
106 106
0 5 10 15 20 25 10 5 0 5 10 15 20 25
SNRav (dB) SNRav (dB)
IEC 61334-5-1 specication ML S-FSK receiver IEC 61334-5-1 specication ML S-FSK receiver
FSK receiver Improved ML FSK receiver Improved ML
S-FSK receiver S-FSK receiver
Figure 4: BER versus SNRav with x = 5 dB. Figure 6: BER versus SNRav with x = 20 dB.
100 Frequency
path Accumulator Sine ROM DAC
(step)
102 Figure 7: Modulator block diagram.

BER
implementation of the S-FSK receiver. The DSP-based digital

104 part communicates, through serial port in full duplex, with
the host device. At the other end, DSP communicates, in half-
duplex, through power line via a mixed front end coupling
106
interface.
0 5 10 15 20 25 The DSP programming structure was dened to handle
SNRav (dB) in real-time transmitting or receiving S-FSK signal.
IEC 61334-5-1 specication
The S-FSK base-band modem is obtained by the imple-
ML S-FSK receiver
Improved ML
mentation of an S-FSK modulator at the transmitter side and
FSK receiver
S-FSK receiver an improved ML receiver at the receiver one.
Figure 5: BER versus SNRav with x = 10 dB.
4.1. Modulator Implementation. The transmitter is com-
posed by three stages:
a payload of 304 random bits. The following curves are (i) a numeric stage involving a DSP that performs fre-
averaged over 1000 packets. quency synthesizing with a direct digital synthesizer
Figures 4, 5, and 6 show the bit error rate (BER) versus (DDS);
the average signal-to-noise ratio SNRav for three unbalancing
factors x {+5, +10, +20} dB. (ii) a digital to analog convertor (DAC) capable to
From the previous gures, the FSK receiver loses in generate a linear signal up to its full scale output;
performance with the increasing of the unbalancing factor; (iii) line driver delivering amplied signal.
however, the ML S-FSK receiver presents relevant improve-
ment on balanced channels. For bit error rate equal to 104 , As described in Figure 7, DDS is based on storing the
more than 6 dB gain at x = 10 dB. samples of a sinusoidal signal in a look-up table (LUT) and to
For X [0, 256] this approximation guarantees a mean read it by a specied integer step index which determines the
square error lower than 103 , which is adequate to obtain phase increment, in order to generate the desired frequency
negligible loss of performance between the ideal ML S-FSK fi which is related to the step index k, the sampling frequency
receiver and the improved ML S-FSK receiver. fs and the LUT length N by the following relation:
k fs
4. DSP Implementation Methodology fi = . (18)
N
Priority in design was given to modularity, simplicity, low It is important to minimize the LUT size since the
cost, and reliability. A 32-bit-xed point general purpose implementation will be done in an embedded processor
DSP architecture is considered to optimize the software where the resources especially the memory size are limited.

1
sin(2 f1 t)
T

dt | |2
r12 l1
( )
g 4r12 21 /04

dt | |2
2

02
1
1
cos(2 f1 t)
T
20 , 21 , 02 and 12
r(t)

1 estimation
sin(2 f0 t)
T 2

12

0
dt | |2
l0
g ( 4r02 20 /14 )

dt | |2 r02

1
cos(2 f0 t)
T
Figure 8: Implemented improved ML S-FSK receiver architecture.
The sampling frequency fs is chosen as multiple of the data Table 1: Orthogonality frequency choice for baud rate 9.6 kbps.
rate 1/T, thus the number of samples in a bit period is an
Frequency (kHz)
integer. CENELEC band Carrier frequency (kHz)
Once the appropriate sine samples are read they serve as f0 f1
input for the DAC. The generated signal by the DAC pin is 91.2 72 81.6
amplied by the line driver. 86.4 67.2 76.8
The S-FSK modulator generates signal in the CENELEC A band 81.6 62.4 72
band from 3 kHz to 95 kHz responding to the following 76.8 57.6 67.2
specications:
72 52.8 62.4
(i) frequency bandwidth 2 f > 10 kHz and multiple of
bit rate D;
(ii) programmable bit rate D;
4.2. Improved ML Receiver Implementation. Coherent FSK
(iii) frequencies f0 and f1 are multiple of D/2. signals can be noncoherently demodulated to avoid the car-
The sampling frequency fs is xed at 3.125 MHz and the rier recovery. The improved ML demodulator is a quadrature
samples number is set to 320 samples to optimize the error receiver capable of detecting signals with unknown phases.
performance at the demodulator side. Therefore, the data It can be implemented with four correlators as shown
rate is equal to 9.6 kbps. in Figure 8, where the four reference signals are sin(2 f0 t),
The step index k is an integer; therefore, the resolution cos(2 f0 t), sin(2 f1 t), and cos(2 f1 t). We will use the
frequency is found by setting k = 1, same DDS module as the modulator one to generate those
reference signals.
fs The signal consists of an in-phase component and a
fres = . (19)
N quadrature component. Thus, the signal is partially corre-
Resolution frequency is set to 4.8 kHz to respect orthogonal- lated with sin(2 fi t) and partially correlated with cos(2 fi t).
ity constraint between two frequencies f0 and f1 . Therefore, we use two correlators to collect the signal energy
The minimum LUT lengths that satisfy the conditions in these two parts.
already cited and the generation of the frequencies f0 and f1 The rst P outputs are used to estimate channel param-
with zero error are 656. eters. Then, we apply probability function (15) to correlator
In Table 1, we present the possible choice of orthogonal output using estimated channel parameters and g function.
frequency f0 and f1 in the case of S-FSK demodulator at baud The g function is a piecewise linear approximated and stored
rate 9.6 kbps. in data memory.
All samples of received bits are processed according to Table 2: Memory and processing time analysis results for the PLC-
Figure 8. The main constraint in the receiver is to tune the modem DSP implementation.
sampling frequency of ADC fADC so as to have
PM space DM space Number of cycles
Module
(16-bit Word) (16-bit Word) machine during T
fs Modulator
= M. (20) module
1620 656 3840
fADC
ADC reception
1066 160 1280
module
Dierent congurations are possible; we have to choose Correlation
the one that maximizes fADC . In this case M is equal to 2 246 656 2560
module
and the ADC sampling frequency fADC becomes equal to S-FSK decision
1.565 MHz. 87 32 156
module
Thus, samples count during bit time T is 160 samples. Initialization
The number of samples per symbol period T must be 513 10 2323
PHY module
multiple of 8 for direct memory access (DMA) use that oer PHY layer FSM
transfer facility and rapidity. 3869 822 5236
module
4.3. Implementation Results. The DSP processor BF506F,

sited to an evaluation board [11, 12], operates with frequency The physical layer is designed and implemented. The
up to 400 MIPS with 32 Kbytes of L1 memory associated to remaining available cycles and memory will be used to build
instructions (L1 code), 32 Kbytes for data (L1 data) accessed upper layers: MAC layer and Application layer.
at full processor speed, and 32 Mbytes of external ash
memory. 5. Conclusions
To evaluate the complexity of the S-FSK modem soft-
In this paper, we have described the design and optimized
ware, it is important to determine the consumed cycles and
DSP implementation of an S-FSK prole for a PLC node in
the consumed data memory space [13].
an AMR system. To overcome power line channel condition,
We have used the data memory to store the LUT table
an improved ML S-FSK receiver is used. Improved receiver
that contains 656 samples encoded on 16 bits.
presents close error performance to the ideal ML S-FSK
g function is stored also on data memory space. g func-
receiver but has simplier architecture.
tion is deened over 8 intervals and the anity coecients
Analysis of new receiver reveals excellent results in terms
are encoded on 16 and 32 bits.
of memory occupations, required cycles, and BER perform-
The cycles consumption is limited by the available
ances.
number of cycles per sample that is governed by the DSP
Data rate of 9.6 kbps is easily provided with exibility and
speed which is 400 MIPS.
programmability to change receiver parameters.
The DDS algorithm consumes only 2 cycles per sample,
one cycle for memory access to read the sample from the
LUT, and one cycle for incrementing the reading index. The Acknowledgments
transfer of DDS samples to DAC convertor requires 10 cycles This work was supported by the Embedded Systems Technol-
per sample. ogy (EBSYS) SmartGrid Division and GRESCOM Research
At the receiving site, the demodulator invokes 4 correla- Laboratory of Higher School of Communication of Tunis.
tors. At each correlator, one sample is treated on 4 cycles to
read, multiply, accumulate, and update index.
Finally, we apply g function on correlators output at the
References
end of symbol reception. [1] H. C. Ferreira, L. Lampe, J. Newbury, and T. G. Swart, Power
The cycles consumptions per sample of these dierent Line CommunicationsTheory and Applications for Narrow-
modules are presented in Table 2. It is important to report band and Broadband Communications over Power Lines, Wiley,
that additional modules are implemented to ensure synchro- 2010.
nization, build packets, and handling dierent events. [2] IEC 61334-5-1, 2.0 b: Distribution automation using distribu-
The cycles consumption of the S-FSK modem software tion line carrier systemspart 5-1: Lower layer prolesthe
composed of the modulator, demodulator, and PHY layer spread frequency shift keying (S-FSK) prole, 2001.
functionalities according to IEC 61334-5-1 is lower than the [3] A. Lotito, R. Fiorelli, D. Arrigo, and R. Cappelletti, A complete
narrow-band power line communication node for AMR, in
available cycles per symbol period T.
Proceedings of the IEEE International Symposium on Power Line
By considering the DSP implementation, we measured Communications and Its Applications (ISPLC 07), pp. 161
an average cycles consumption of 9076 cycles during trans- 166, March 2007.
mission (21.78% of available cycles) and 9232 cycles during [4] D. Cooper and T. Jeans, Narrowband, low data rate com-
reception (22.15% of available cycles). munications on the low-voltage mains in the CENELEC
Memory consumption is 10.25% for data memory and frequenciespart I: noise and attenuation, IEEE Transactions
45.17% for code memory. on Power Delivery, vol. 17, no. 3, pp. 718723, 2002.
[5] CENELEC EN 50065-1:2001.Signaling on low-voltage elec-

trical installations in the frequency range from 3 kHz to
148.5 kHzpart I: General requirements, frequency bands
and electromagnetic disturbances.
[6] T. Schaub, Spread frequency shift keying, IEEE Transactions
on Communications, vol. 42, no. 2, pp. 10561064, 1994.
[7] T. Schaub and B. Steinle, Minimax noncoherent detection of
BFSK signals in nonwhite gaussian noise, IEEE Transactions
on Communications, vol. 34, no. 9, pp. 955958, 1986.
[8] B. Sklar, Digital Communications: Fundamentals and Applica-
tions, Prentice Hall, 2001.
[9] F. Xiong, Digital Modulation Techniques, Artech House, Lon-
don, UK, 2000.
[10] D. Veronesi, L. Guerrieri, and P. Bisaglia, Improved spread
frequency shift keying receiver, in Proceedings of the 14th
Annual International Symposium on Power Line Communica-
tions and its Applications (ISPLC 10), pp. 166171, March
2010.
[11] ADSP-BF50x Blackn Processor Hardware Reference. Rev 1.0,
Analog Devices, 2010.
[12] [12] ADSP-BF506F EZ-KIT Lite Evaluation System Manual.
Rev 1.0, Analog Devices, 2009.
[13] H. Hajji and A. Ghazel, Design and implementation of a
correlation-based DSP software for narrowband power line
communication receiver, in Proceedings of the 13th IEEE
International Conference on Electronics, Circuits and Systems
(ICECS 06), pp. 114118, December 2006.

Power-Line Communications: Smart Grid, Transmission, and Propagation

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Power-Line Communications: Smart Grid, Transmission, and Propagation

Uploaded by

Copyright:

Available Formats

Journal of Electrical and Computer Engineering

Power-Line Communications: Smart Grid,

Circuits and Systems

Enhancements of G3-PLC Technology for Smart-Home/Building Applications, Luca Di Bert,

Improved Maximum Likelihood S-FSK Receiver for PLC Modem in AMR,

Justinian Anatory,1 Moises V. Ribeiro,2 Andrea M. Tonello,3 and Ahmed Zeddam4

Correspondence should be addressed to Justinian Anatory; janatory@udom.ac.tz

Received 28 January 2013; Accepted 28 January 2013

Lars Torsten Berger,1 Andreas Schwager,2 and J. Joaqun Escudero-Garzs3

Correspondence should be addressed to Lars Torsten Berger; lars.berger@kenus.es

Received 3 August 2012; Accepted 29 December 2012

Academic Editor: Ahmed Zeddam

1. Introduction deployments [3]. Industry adoption and large-scale customer

Service Distribution, medium voltage (MV)

Customer, low voltage (LV) M3

Communication link EVSE: electrical vehicle service equipment

and distribution automation field. However, in the light of Load

For the LV-In-Home situation the mean gain is given L T1

Mean SNR (dB)

where the last term normalizes out the 30 kHz bandwidth 15

Non-SG G. SGH SGH SGH Non-SG

SGH AMI domain

7. Application Layer Interoperability 8. Conclusions

Amilcar Mescco,1 Pascal Pagani,2 Michel Ney,2 and Ahmed Zeddam1

Correspondence should be addressed to Amilcar Mescco; amescco@gmail.com

Received 4 August 2012; Revised 16 January 2013; Accepted 21 January 2013

Academic Editor: Moises V. Ribeiro

1. Introduction at frequencies below 500 kHz, in both indoor and out-

50 at the transmitter, hence generating less Electro Magnetic

energy dissipated at any location except the intended receiver.

( ,r0) ( ,r0) ( ,r0)

Figure 5: Experimental setup calibration.

Preamble Symbol 1 Symbol 2 Symbol 3 Symbol 4 Symbol 5 Symbol 6 Preamble () () () () () ()

Figure 7: HomePlug frame for calibration and CTF measurement

5 transmission with some degree of channel attenuation. On

1 electrical network topologies, where the rich multipath envi-

0.5 Figure 12 shows that the application of TR provides better

The cumulative distribution function (CDF) of TR is TR = TR + TR . (13)

MHz-part II: time-frequency analysis, IEEE Transactions on

Larry Yonge,1 Jose Abad,2 Kaywan Afkhamie,1

Correspondence should be addressed to Pascal Pagani; pascal.pagani@ieee.org

Received 3 August 2012; Accepted 21 November 2012

Academic Editor: Andrea M. Tonello

1. Introduction wires to deploy wide band services is the main source of

AV2 Frame Turbo Frame Control MIMO IFFT Cyclic Prefix,

Time Frame Control Turbo AV2 Frame

Figure 1: HomePlug AV2 transmitter and receiver PHY layer.

1.0.1 Frame Control data passes through a separate HomePlug L

The numbers of used transmit ports and used receive

Spatial Multiplexing Spatial Multiplexing

(i) The FM band region 87.5108 MHz shall be avoided.

Copy and window

Figure 11: OFDM symbol with GI and window.

beginning and at the end by copying the bits as done for

an increase of the radiated EMI, the design of the IMC factor 1

(i) 6 different locations in Germany and 3 different 0.4

1 (ii) For 10% of the records, the Tx power could be

() , 0) , IMCmax ) , We then focused on the effect of the EMC Friendly Power

1 (i) The Transition Interval defines the part of the Roll-

Long MPDU using a AV1.1 delimiter AV1.1

AV1.1 AV1.1 Payload AV1.1

RIFS RIFS CIFS

Long MPDU using a AV1.1 delimiter Short Delimiter based

-VDB %J #FSU 4BMWBUPSF %"MFTTBOESP BOE "OESFB . 5POFMMP

$PSSFTQPOEFODF TIPVME CF BEESFTTFE UP "OESFB . 5POFMMP UPOFMMP!VOJVEJU

3FDFJWFE 4FQUFNCFS "DDFQUFE %FDFNCFS

"DBEFNJD &EJUPS "INFE ;FEEBN

*OUSPEVDUJPO B TNBMM BNPVOU PG EBUB %FTQJUF UIF TNBMM BNPVOU PG EBUB

5 (1-$ TZTUFN TQFDJDBUJPOT

Ethernet cat5 4XJUDI F TXJUDI JT B XFMMLOPXO OFUXPSL EFWJDF

"T EFQJDUFE JO 'JHVSF UIF SPVUJOH NPEVMF LFFQT USBDF PG !

' 3PVUFS PX DIBSU

UIF LOPXMFEHF PG UIF USBOTNJUUFE GSBNF PS UIF '$4 FME XIFSF

' 4XJUDI PX DIBSU

' 4DSFFOTIPU PG BO FYBNQMF PG DPOWFSHFOU OFUXPSL TJNVMBUJPO JO 0./&5

' $$%' PG UIF NFBTVSFE 5)3 GPS (1-$ UFDIOPMPHZ

FRVBM UP BOE SFTQFDUJWFMZ GPS OPSNBM %214,

DBNQBJHOT BSF SFQPSUFE JO <>

' "O FYBNQMF PG TVQFSGSBNF TUSVDUVSF <>

5 4VQFSGSBNF QBSBNFUFST BOE WBMVFT F WBMVFT PG UIF QBSBNFUFST VTFE JO CFBDPOFOBCMFE

'VSUIFS *NQSPWFNFOUT 6TJOH 5JNF

<> .VMUJNFEJB PWFS $PBY "MMJOBDF .P$" 4QFDJDBUJPO GPS