UNIT - 4

NOISE IN ANALOG
COMMUNICATION
SYSTEMS
Noise in electrical terms may be defined as any unwanted
introduction of energy tending to interfere with the proper
reception and reproduction of transmitted signals.
Noise is mainly of concern in receiving system, where it sets a
lower limit on the size of signal that can be usefully received. Even
when precautions are taken to eliminate noise from faulty
connections or that arising from external sources, it is found that
certain fundamental sources of noise are present within electronic
equipment that limit the receivers sensitivity.
Classification of noise
NOISE

NOISE WHOSE SOURCES ARE NOISE CREATED WITHIN

EXTERNAL TO THE RECEIVER THE RECEIVER ITSELF
EXTERNAL NOISE

Noise created outside the receiver

External noise can be further classified as:
1. Atmospheric
2. Extraterrestrial
3. Industrial
ATMOSPHERIC NOISE

Atmospheric noise or static is generally caused by lightning

discharges in thunderstorms and other natural electrical disturbances
occurring in the atmosphere.

Since these processes are random in nature, it is spread over most of

the RF spectrum normally used for broadcasting.
Atmospheric Noise consists of spurious radio signals with
components distributed over a wide range of frequencies. It is
propagated over the earth in the same way as ordinary radio
waves of same frequencies, so that at any point on the
ground, static will be received from all thunderstorms, local and
distant.

Atmospheric Noise becomes less at frequencies above 30 MHz

Because of two factors:-

1. Higher frequencies are limited to line of sight propagation i.e.

less than 80 km or so.
2. Nature of mechanism generating this noise is such that very
little of it is created in VHF range and above.
EXTRATERRESTRIAL NOISE

COSMIC NOISE SOLAR NOISE

Solar Noise
Under normal conditions there is a constant noise radiation from
sun, simply because it is a large body at a very high temperature (
over 6000C on the surface, it therefore radiates over a very broad
frequency spectrum which includes frequencies we use for
communication.

Due to constant changing nature of the sun, it undergoes cycles of

peak activity from which electrical disturbances erupt, such as
corona flares and sunspots. This additional noise produced from a
limited portion of the sun, may be of higher magnitude than noise
received during periods of quite sun.
Cosmic Noise
Sources of cosmic noise are distant stars ( as they have high
temperatures)
The noise received is called Black Body noise and is distributed fairly
uniformly over the entire sky.

INDUSTRIAL NOISE

This noise ranges between 1 to 600 MHz ( in urban, suburban and

other industrial areas) and is most prominent.

Sources of such Noise : Automobiles and aircraft ignition, electric

motors, switching equipment, leakage from high voltage lines and a
multitude of other heavy electrical machines.
The Noise is produced by the arc discharge present in all these
operations. ( this noise is most intense industrial and densely
populated areas)
INTERNAL NOISE
Noise created by any of the active or passive devices found in
receivers.

Such noise is generally random, impossible to treat on individual

voltage basis, but easy to observe and describe statistically. Because the
noise is randomly distributed over the entire radio spectrum therefore it
is proportional to bandwidth over which it is measured.
Internal noise can be further classified as:
1. Thermal Noise
2. Shot Noise
3. Low frequency or flicker Noise
4. Burst Noise
Thermal Noise
The noise generated in a resistance or a resistive component is
random and is referred to as thermal, agitation, white or Johnson
noise.

CAUSE :
The free electrons within an electrical conductor possess kinetic
energy as a result of heat exchange between the conductor and its
surroundings.
Due to this kinetic energy the electrons are in motion, this motion is
randomized through collisions with imperfections in the structure of
the conductor. This process occurs in all real conductors and gives
rise to conductors resistance.
As a result, the electron density throughout the conductor varies
 randomly, giving rise to randomly varying voltage across the ends of
conductor. Such voltage can be observed as flickering on a very
sensitive voltmeter.

The average or mean noise voltage across the

conductor is zero, but the root-mean-square value
is finite and can be measured.
The mean square value of the noise voltage is
proportional to the resistance of the conductor, to
Ideal
its absolute temperature, to the frequency Filter
H(f)
bandwidth of the device measuring the noise.
The mean-square voltage measured on the
meter is found to be
2
En = 4RkTBn
 Where En = root-mean-square noise voltage, volts R =
resistance of the conductor, ohms
T = conductor temperature, kelvins Bn = noise
bandwidth, hertz
k = Boltzmanns constant ( )

And the rms noise voltage is given by :

En = (4RkTBn )

NOTE: Thermal Noise is not a free source of energy. To abstract the noise power, the
resistance R is to be connected to a resistive load, and in thermal equilibrium the load
will supply as much energy to R as it receives.
 R
V RL
En

In analogy with any electrical source, the available average power is

defined as the maximum average power the source can deliver. Consider
a generator of EMF En volts and internal resistance R .

Assuming that RL is noiseless and receiving the maximum noise

power generated by R; under these conditions of maximum power
transfer, RL must be equal to R. Then
Pn = V2/RL = V2/ R = (En/2)2 /R = En2 /4R
Using Equation ,
Pn = kTBn
 Example:
Calculate the thermal noise power available from any resistor at room
temperature (290 K) for a bandwidth of 1MHz. Calculate also the
corresponding noise voltage, given that R = 50
Solution For a 1MHz bandwidth, the noise power is
Pn = 1.38 10-23 290 106 R
= 4 10-15 W
G=1/R
En2 = 4 50 1.38 10-23 290
= 810-13 En2 = 4RkTBn
= 0.895 In2 = 4GkTBn

The thermal noise properties of a resister R may be a.) Equivalent CVuorltraegnet

represented be the equivalent voltage generator . SSoouurrccee

Equivalent current generator is found using the Nortons Theorem.

Using conductance G = (1/R), the rms noise current is given by :
In2 = 4GkTBn
 Resisters in Series
let Rser represent the total resistance of the series chain, where
Rser = R1 + R2 + R3 + .; then the noise voltage of equivalent series
resistance is
En2 = 4Rser kTBn
= 4( R1 + R2 + R3 + )kTBn
= En12 + En22 + En32 + .....
Hence the noise voltage of the series chain is given by:
2 2 2
En = (En1 + En2 + En3 + .....)

Resisters in Parallel
With resistors in parallel it is best to work in terms of conductance.
Let Gpar represent the parallel combination where Gpar = G1 + G2 + G3 +
; then
In2 = 4Gpar kTBn
= 4( G1 + G2 + G3 + )kTBn
= I 2 + I 2 + I 3 + .
n1 n2 n3
REACTANCE
Reactances do not generate thermal noise. This
follows from the fact that reactances cannot
Dissipate power.
Consider an inductive or capacitive reactance
connected in parallel with a resistor R.
In thermal equilibrium, equal amounts of power must be exchanged;
that is, P1 = P2 . But since the reactance cannot dissipate power, the
power P2 must be zero, and hence P1 must also bezero.

Shot Noise
Shot noise is random fluctuation that accompanies any direct current
crossing potential barrier. The effect occurs because the carriers
(electrons and holes in semiconductors) do not cross the barrier
simultaneously but rather with random distribution in the timing of
each carrier, which gives rise to random component of current
superimpose on the steady current.
In case of bipolar junction transistors , the bias current crossing the
forward biased emitter base junction carries the shot noise.
When amplified, this noise sounds as though a shower of lead shots
were falling on a metal sheet. Hence the name shot noise.

Although it is always present, shot noise is not normally observed

during measurement of direct current because it is small compared to
the DC value; however it does contribute significantly to the noise in
amplifier circuits.

The mean square noise component is proportion

to the DC flowing, and for most devices the mean- I
DC
Square, shot-noise is given by:
In2 = 2Idc qe Bn ampere2
Time

Where Idc is the direct current in amperes, qe is the magnitude of

electronic charge and Bn is the equivalent noise bandwidth in hertz.
 Example
Calculate the shot noise component of the current present on the direct
current of 1mA flowing across a semiconductor junction, given that the
effective noise bandwidth is 1 MHz.
SOLUTION
In2 = 2 10-3 1.6 10-19 106
= 3.2 10-16 A2
= 18 nA

Flicker Noise ( or 1/f noise )

This noise is observed below frequencies of few kilohertz and its
spectral density increases with decrease in frequency. For this reason it is
sometimes referred to as 1/f noise.

The cause of flicker noise are not well understood and is recognizable
by its frequency dependence. Flicker noise becomes significant at
frequency lower than about 100 Hz. Flicker noise can be reduced
significantly by using wire-wound or metallic film resistors rather than
the more common carbon composition type.
In semiconductors, flicker noise arises from fluctuations in the carrier
densities (holes and electrons), which in turn give rise to fluctuations in
the conductivity of the material. I.e the noise voltage will be developed
whenever direct current flows through the semiconductor, and the
mean square voltage will be proportional to the square of the direct
current.

In Electronic devices, it shows up as a low frequency phenomenon, as

the higher frequencies overshadowed by white noise from other
sources.
Burst Noise
It consists of sudden step-like transitions between two or more
discrete voltage or current levels, as high as several
hundred microvolts, at random and unpredictable times. Each shift in
offset voltage or
current often lasts from several milliseconds to seconds, and sounds
like popcorn popping if hooked up to an audio speaker.

The most commonly invoked cause is the random trapping and

release of charge carriers at thin film interfaces or at defect sites in
bulk semiconductor crystal. In cases where these charges have a
significant impact on transistor performance (such as under an MOS
gate or in a bipolar base region), the output signal can be substantial.
These defects can be caused by manufacturing processes, such as
heavy ion-implantation, or by unintentional side-effects such as surface
contamination.

Typical popcorn noise, showing discrete levels of channel current modulation due to the trapping and
release of a single carrier, for three different bias conditions
 SPECTRAL DENSITY

spectral density is power distribution in frequency spectrum

it use to distinguish type of noise. plote of light intensity/power as a
function of frequency or wavelength.

Thermal noise lies in category of power signals has spectral

densities. Bn is property of external receiving system and Bn
assume flat.

from eq. Pn=E2 n/4R=4RkTBn/4R =kTBn , since E2n=4RkTBn where

En=rms noise voltage(volt)
R=resistance of conductor()
Bn=noise bandwidth(hertz)
K=Boltzmaan constant=1.38x10-23(J/k)
T=conductor temperature(kelvin)
available power spectral density in Watts/Hz or Joule is
Ga(f)=Pn/Bn=kT
 spectral density for mean square voltage is given by
Gv(f)=E2n/Bn=4RkT(v2/hz) since E2n=4RkTBn. Spectral densities
are flat that is independent of frequency in fig. below

Spectral
density
4RkT Gv 2

kT Gp

Thermal noise spectral

densities
 So thermal noise is also called white noise because of
analogy to white light that has flat spectrum means signal
contain equal power within a fixed bandwidth at any center
frequency.

when white noise is passed through a network then spectral

density is changed by shape of network frequency response.

total noise power at output is sum of the noise contribution

over complete frequency range. taking into account of
frequency response shape.
 Consider a power spectral density response in figure below

Sp (f)
f
Sp(f1)

f1 f2 f3 f
 at frequency f1 the noise power for small band width f about
f1 is Pn1=Sp(f1).f Here f assume flat about f1

Pn1= spectral density (watts/hertz). Bandwidth (Hz) or

Pn1=Sp(f1).f

So noise power is equal to area of shaded strip about f1 similarly

for f2,f3,.so that power is sum of all these small areas equal to
total area under the curve. Area of curves gives total mean
square voltage

area of curve= Ga(f) ;for f=0 to f=

 for mean square voltage Sp(f1)xf=Sp(f1)f(V2)=area of curves
gives total mean square voltage.

Equivalent noise bandwidth

it is the frequency span of a noise power curve with amplitude

equal to actual peak value and with same integrated area

If R is connected to input of LC filter as in figure (a) this

represent an input generator of mean square voltage spectral
density 4RkT feeding a network of R and LC filter
Let transfer function of network including R be H(f), so spectral
density for mean square voltage is=4RkT|H(f)|2 ; H(f) is ratio of
output to input voltage for mean square voltage.
 H(f)

4RkT Vn 2

|H(f)|

Frequency
(Linear scale)
 Total mean square output voltage is given by
Vn2=4RkT.|H(f)|2f for f=0 to
=4RkTx(area under |H(f)|2) curve

or total mean square voltage at the output can be stated as

Vn2=4RkTBn

By above two equation gets equivalent noise bandwidth of

network

Bn=(|H(f)|2f) =area under curve |H(f)|2 for f=0 to

Example considered in fig. below input capacitance of the
voltmeter used measure the noise voltage across
Circuit diagram
H(f)

R
C C V2
R n

4RkT
 transfer function of RC network |H(f)|=1/[1+(CR)]
Equivalence noise bandwidth of the RC network
is Bn=|H(f)|2df=1/4RC; f=0 to

Vn2=4RkTx1/4RC=kT/C ,so mean square voltage orginates

from R even though it is independent of R and inversly
proportional to C even though C does not generate noise.

transfer function |H(f)|=|Xc/Zs| ; Zs=r(1+jyQ) is impedance of

the series tuned circuit and,Xc=1/jc i.e. reactance of c.
 let circuit is resonant at fo and noise restricted to bandwidth
f fo about resonant frequency fo so transfer function given by
|H(f)|=1/o Cr=Q ; area under|H(f)|2 curve of small bandwidth
f is Q2f so

mean square noise voltage is

Vn2=4rkTBn =4rQ2kTf=4RDkTf use the relation
Q2r=RD ; f=0 to and Bn =1/4RDC

RD is dynamic resistance of the tuned circuit . Noise bandwidth

expressed as a function of 3-dB bandwidth of circuit

From equ. B3dB= fo /Q and RD=Q/o C combine these equ. Gets

Bn=/2B3dB
 the mean square voltage at the output as
V n2=4RDkTx1/4RDC=kT/CSo noise from resister RD is limited by
bandwidth Bn is V2n=4RDkTx1/4RDC=kT/C

here assume that Q factor is constant and independent of

frequency for end result gives good indication of noise
expected in practice but not true for range zero to infinity

In radio receivers noise is generated at antenna receiver input

and output noise bandwidth is determined by the audio part of
the receiver

Equivalent noise bandwidth=area of normalized frequency

curve for low frequency section. normalized means curve max.
value is unity
 Response curve show output in decibels relative to maximum as
in next slide fig.

Equivalent noise bandwidth is area of curve for a single

sideband type, then noise bandwidth appears on both side of
the carrier and gets doubled.
Response curve
Relative Relative response
Response Power ratio
dB
I
Area under curve
Equal to effective
Noise bandwidth

Frequency(log scale) Frequency linear scale

(a) (b)

Noise bandwidth

RF bandwidth
(c)
(a)Amplifier frequency response curve (b) curve of (a) using linear scales (c)
noise bandwidth Of a double sideband receiver
 Signal to noise ratio

in communication it is the signal to noise ratio rather than

absolute value of noise.
It is defined as a power ratio S/N=Ps/Pn=V2 /V2
s n
Repeater /amplifier insert to make up for the loss in analog
telephone cable
If power loss of a line section is L then repeater amplifier power gain G
is chosen so LG=1,long line divided into identical section

If input signal power=Ps to first section as signal passes along the link
power output at
 each repeater is Ps since LG=1 for each link but noise power
are additive and the total noise at the output of mth link is
Pn=Pn1+Pn2++Pnm

If each links are identical and contribute Pn then total noise

power is Pnm=MPn then output signal to noise ratio is
(N/S)odB=10logPs/MPn=(S/N)1dB-(M)dB

Where (S/N)1 is ratio for one link and (M)dB is no of links

expressed as power ratio in decibels
Ques: the is equivalent noise resistance for an amplifier is 300 and
equivalent shot noise current 5A.the amplifier is fed from 150 10V
rms sinusoidal signal source. Calculate the individual noise voltage at the
input signal to noise ratio in decibels. the noise bandwidth is 10MHz.

Solution: let room temperature so that kT=4x10-21J

And qe=1.6x10-19C shot noise current is Ina=[2qeIEQBn]1/2=4nA
So noise voltage across source resistance is
InaRs=0.6V shot noise current does not develop a voltage across
Rn . The noise voltage generated by Rn is Vna=[4Rn kToBn]1/2=6.93V
Thermal noise voltage from source is Vns=[4RskToBn]1/2=4.9V
total noise voltage at input to the amplifier is Vn=[4.92+6.932+.62]1/2=8.51V
so signal to noise ratio in decibels is
S/N=20logVs/Vn=1.4dB
 NOISE FACTOR

Noise factor is the ratio of available S/N ratio at the input to the
available S/N ratio at the output .

Consider a signal source at room temperature To = 290K

providing an input to an amplifier . The available noise power from
this would be
Pni = kToBn .
where , k = boltzmann constant = 1.38X10-23 J/K
Bn = equivalent noise bandwidth in Hz
 Let the available signal power be Psi , then available s
ignal
to noise ratio from the source is

(S/N)ni = Psi /kTo Bn

The source connected to the amplifier represents
available signal to noise ratio.
If amplifier has the available power gain denotedby G
, the available output signal power Pso = GPsi and if t he
amplifier was entirely noiseless , the available noise power
would be Pso = GkTo Bn .
However , it is known that all real amplifiers contribute noise
and the available output signal to noise ratio will be less than
that at the input.
 The noise factor F us defined as
F = (available S/N power ratio at the input) /
(available S/N power ratio at the output)

F = ( Psi /kTo Bn ) X (Pno /GPsi )

F = Pno /GkTo Bn
It follows from this that the available output noise power is given by
Pno =FGkTo Bn
F can be interpreted as the factor by which the amplifier increases
the output noise , for ,if amplifier were noiseles the output noise
would be GkTn Bn .

The available output power depends on the actual input power

delivered to the amplifier .
Noise factor is a measured quantity and will be specified for
given amplifier or network. It is usually specified in decibels , when
it is referred to as the noise figure. Thus

noise figure = (F) dB = 10logF

Example
The noise figure of an amplifier is 7dB. Calculate the output signal
to noise ratio when the input signal to noise ratio is 35 dB.
Sol . From the definition of noise factor ,
(S/N)o = (S/N)in (F) dB
= (35 7) dB
= 28 db
 Amplifier Input Noise in terms of F

Amplifier noise is generated in many

components throughout the amplifier , but it
proves convenient to imagine it to originate from
some equivalent power source at the input of the
amplifier .
Then the total available power input noise is

Noiseless
amplifier
Pni = Pno / G
(F-1)kTo Bn
Gain, G Pno = FGkTo Bn
kT
o Bn
Noise Factor = FkTo Bn
F
 Example
An amplifier has a noise figure of 13dB. Calculate
equivalent amplifier input noise for a bandwidth of 1
MHz.
Sol. 13 dB is a power ratio of approximately 20 : 1. hence
Pna = (20 1)X 4 X 10-21 X 106
= 1.44pW.
Noise figure must be converted to a power ratio F to be
used in the calculation.
 Noise factor of amplifiers in cascade
consider first two amplifiers in cascade . The problem is to
determine the overall noise factor F in terms of individual noise
factors and available power gains .
the available noise power at the output of the amplifier 1 is
Pno1 = F1 G1 kTo Bn and this available to amplifier 2.
Amplifier 2 has noise (F2 1)kTo Bn of its own at its input, hence
total available noise power at the input of amplifier 2 is
Pni2 = F1 G1 kTo Bn + (F2 -1)kTo Bn
Now since the noise of amplifier 2 is represented by its equivalent
input source , the amplifier itself can be regarded as being noiseless
and of available power gain G2 , so the available noise output of
amplifier 2 is
Pno2 = G2 Pni2
= G2 ( F1 G1 kTo Bn + (F2 1)kTo Bn ) (1)
The overall available power of the two amplifiers in cascade is
G = G1 G2 and let overall noise factor be F ; then output noise power
can also be expressed as
Pno = FGkToBn (2)
equating the two equations for output noise (1) and (2)
 F1 G1 kTo Bn + (F2 -1)kTo Bn = FGkToBn

F1 G1 + (F2 -1) = FG

F = F1 G1 / G + (F2 1)/G

where G = G1 G2

F = F1 + ( F2 1)/ G1
This equation shows the importance of high gain , low
noise amplifier as the first stage of a cascaded system. By
making G1 large, the noise contribution of the second
stage can be made negligible, and F1 must also be small so
that the noise contribution of the first amplifier is low.
The argument is easily extended for additional amplifiers
to give

F = F1 + (F2 -1)/G1 + (F3 -1)/ G1 G2

This is known as FRISS FORMULA.

There are two particular situations where a low noise
, front end amplifier is employed to reduce the noise. One
of these is in satellite receiving systems.

The other is in radio receivers used to pick up weak

signals such as short wave receivers.

In most receivers , a stage known as the mixer stage is

employed to change the frequency of the incoming signal
, and it is known that the mixer stages have notoriously
high noise factors. By inserting an RF amplifier ahead of
the mixer , the effect of the mixer noise can be reduced to
negligible levels. This is illustrated in following example.
 Example
A mixer stage has a noise figure of 20dB and this is preceded by an
amplifier that has a noise figure of 9 dB and an available power gain
of 15dB. Calculate overall noise figure referred to the input .
Sol. It is necessary to convert all decibel values to the equivalent
power ratios :
F2 = 20dB = 100:1 power ratio
F1 = 9dB = 7.94:1 power ratio
G1 =15dB = 31.62:1 power ratio
F = F1 + (F2 1)/ G1
= 7.94 + (100-1)/31.62
= 11.07
This is overall noise factor. The overall noise figure is
(F)dB= 10 log 11.07
= 10.44dB
 NOISE TEMPERATURE

The concept of noise temperature is based on available noise power

equation
Pn = kTa Bn
Here the subscript has been included to indicate the noise temperature is
associated only with available noise power.
In general, Ta will not be same as that physical temperature of the noise
source. As an example, an antenna pointed at deep space will pick up a small
amount of cosmic noise. The equivalent noise temperature of antenna that
represents this noise power may be a few tens of kelvins, well below the
physical ambient temperature of the antenna. If the antenna is directly
pointed at the sun , the received noise power increases enormously and the
corresponding equivalent noise temperature is well above the ambient
temperature.
When the concept is applied to an amplifier, it relates to equivalent noise
of the amplifier referred to the input. If the amplifier noise referred to the
input is denoted by Pna , the equivalent noise temperature of the amplifier
referred to the input is
Te = Pna / kBn (3)
We know equivalent input power for an amplifier is given in terms
of its noise factor by
Pna = (F-1)kTo Bn
putting this in equation (3)
we get equivalent noise temperature of the amplifier as
Te = (F-1)/Ta
This shows the proportionality between Te and F.
In practice it is found that noise temperature is the better measure
for low noise devices , such as low noise amplifiers used in satellite
receiving systems while noise factor is a better measure for the main
receiving system.
Frisss formula can be expressed in terms of equivalent noise
temperatures. Denoting by Te the overall noise of the cascaded
system referred to the input , and by Te1 , Te2 , and so on , the noise
temperatures of the individual stages , the in Frisss formula is easily
rearranged to give
Te = Te1 + Te2 /G1 + Te3 / G1 G2 + ..
Q. A receiver has a noise figure of 12dB and it is fed by a low noise
amplifier that has gain of 50dB and a noise temperature of 90 K.
calculate the noise temperature of the receiver and the overall noise
temperature of the receiving system.
SOL. 12dB represents a power ratio of 15.85 : 1. Hence
Tem = (15.85-1) X 290 = 4306 K
The 50dB gain represents a power ratio of 105 : 1 . Hence
Te = 90 + 4306/ 105
= 90 K
This example shows the relatively high noise temperature of the
receiver , which clearly cannot be its physical temperature. It also
shows how the low noise amplifier controls the noise temperature of
the overall receiving system.
Receiver Model
Noise in DSB-SC Receivers
Noise in SSB-SC Receivers
Noise in AM Receivers
Noise in FM Receivers

2
Introduction:

To undertake an analysis of noise in continuous-

wave (CM) modulation systems, we need a
receiver model.

The customary practice is to model the receiver

noise (channel noise) as additive, white, and
Gaussian. These simplifying assumptions enable us
to obtain a basic understanding of the way in which
noise affects the performance of the receiver.

4
 Receiver Model

s(t) denotes the incoming modulated signal.

w(t) denotes front-end receiver noise. The power spectral density
of the noise w(t) is denoted by N0/2, defined for both positive and
negative frequencies. N0 is the average noise power per unit
bandwidth measured at the front end of the receiver.
The bandwidth of this band-pass filter is just wide enough to pass
the modulated signal without distortion.
Assume the band-pass filter is ideal, having a bandwidth equal to
the transmission bandwidth BT of the modulated signal s(t) , and a
mid- band frequency equal to the carrier frequency fc , fc >> BT .
6
Receiver Model

Idealized characteristic of band-pass filtered noise.

The filtered noise n(t) may be treated as a
narrow band noise represented in the canonical form:
n t nI t cos 2f ct nQ t sin 2f ct 6.1
where nI(t) is the in-phase noise component and nQ(t) is the
quadrature noise component, both measured with respect to
the
carrier wave 7
Accos(2fct).
 Receiver Model

The filtered signal x(t) available for demodulation is defined

by
x t s t n t 6.2
The average noise power at the demodulator input is equal to
the total area under the curve of the Npower spectral density SN( f ):
Pavg-noise 2 B 0 B N
T T 0
2
Input signal-to-noise ratio (SNR)I is defined
as: average power of the modulated signal s t
SNR
average power of the filtered noise n
I

t ratio (SNR) is defined as:
Output signal-to-noise O

average power of the demodulated message signal

SNR) O

average power of the noise

55
 Receiver Model
This ratio may be viewed as the signal-to-noise ratio that
results from baseband (direct) transmission of the message
signal m(t) without modulation, as demonstrated in the
following figure:

The message power at the low-pass filter input is adjusted to be the same as
the average power of the modulated signal
The low-pass filter passes the message signal and rejects out-of-band noise.

56
 Receiver Model

Figure of merit
For the purpose of comparing different continuous-wave (CW) modulation
systems, we normalize the receiver performance by dividing the output
signal-to-noise ratio by the channel signal-to-noise ratio.
The higher the value of the figure of merit, the better will the noise
performance of the receiver be.
The figure of merit may equal one, be less than one, or be greater than
one, depending on the type of modulation used.

SNR O
Figure of merit=
SNR C
 Noise in DSB--SC Receiver
The model of a DSB-SC receiver using a coherent
detector

Figure 6.4

The amplitude of the locally generated sinusoidal wave is

assumed to be unity.
For the demodulation scheme to operate satisfactorily, it is
necessary that the local oscillator be synchronized both in phase
and in frequency with the oscillator generating the carrier wave in
the transmitter. We assume that this synchronization has been
achieved.
 Noise in DSB--SC Receiver
The DSB-SC component of the modulated signal s(t) is
expressed as
s t CA cos 2 fc t m t
c
where C is the system dependent scaling factor. The purpose of
which is to ensure that the signal component s(t) is measured in
the same units as the additive noise component n(t).

m(t) is the sample function of a stationary process of zero mean,

whose power spectral density SM( f ) is limited to a maximum
frequency W, i.e. W is the message bandwidth.
The average power P of the message signal is the total area under
the curve of power spectral density
P W S M f df
W
 Noise in DSB--SC Receiver
In Example 5.12: Mixing of a Random Process with a
Sinusoidal Process
Y f X t cos 2 f t c
RY RX cos 2 fc
1
2

S f
1
X f fc f fc
4

S
Therefore, the average powerX of the DSB-SC modulated signal
component s(t) is given by: C 2 Ac 2 P
2
The average power of the noise in the message BW is WN0
The channel signal-to-noise ratio of the DSB-SC
modulation C 2 Ac2
system SNR C P 2WN
,DSB 0
is:
 Noise in DSB--SC Receiver
Next, we wish to determine the output signal-to-noise ratio.
Using the narrowband representation of the filtered noise n(t), the
total signal at the coherent detector input may be expressed as:
x t s t n t
CAc cos 2 fct m t nI t cos 2 fct nQ t sin 2 fct
The output of the product-modulator component of the coherent
detector is:
t x t cos 2 f c t
cos(+) + cos() = 2cos()cos ()
1 CA m t 1 n t sin(+) + sin() = 2sin()cos()
c
2 2
CA m t n t cos 4 f t 1 A n
1
t sin 4 f t
I

c
c
2
c Q c
2 I

y t CA m t n t
1 1
c
2 2
I
 Noise in DSB--SC Receiver
Equation indicates the following:

The message signal m(t) and in-phase noise component nI(t) of the filtered noise
n(t) appear additively at the receiver output.

The quadrature component nQ(t) of the noise n(t) is completely rejected by the

coherent detector.

We note that these two results are independent of the input signal- to-noise ratio.

Thus, coherent detection distinguishes itself from other demodulation techniques in

the important property: the output message component is unmutilated and the noise
component always appears additively with the message, irrespective of the input
signal-to-noise ratio.
 Noise in DSB--SC Receiver
The receiver output signal : y t 1 CA m t 1 n
c
t 2
I
2
The average power of message component may be expressed as
C 2 Ac2 P
Pavg
4

In the case of DSB-SC modulation, the band-pass filter in Figure 6.4 has a band-
width BT equal to 2W in order to accommodate the upper and lower sidebands of
the modulated signal s(t). Therefore, the average power of the filtered noise n(t) is
2WN0.
From Property 5 of Sect 5.11, the average power of the (low-pass) nI(t) is the same
as that of the (band-pass) filtered noise n(t).
 Noise in DSB--SC Receiver
The output signal-noise ratio for DSB-
2 2
SC C Ac P
SNR O,DSB-SC 4
WN0 2
2 2
C Ac P

2WN
0
We obtain the figure of
merit SNR O 1
SNR C DSB-
SC
 Effect of Noise on SSB AM
SSB modulated signal :
u(t ) Ac m(t ) cos(2 f ct ) Ac m
(t ) sin( 2 f ct )
Input to the demodulator
r (t ) Ac m(t ) cos( 2 f ct ) Ac m (t ) sin( 2 f ct ) n(t )
Ac m(t ) cos( 2 f ct ) Ac m (t ) sin( 2 f ct ) nc (t ) cos2 f ct ns (t ) sin 2 f ct
Ac m(t ) nc (t )cos( 2 f ct ) Ac m (t ) ns (t )sin 2 f ct

Assumption : Demodulation with an ideal phase reference.

Hence, the output of the lowpass filter is the in-phase
component (with a coefficient of ) of the preceding signal.
y(t ) 12 Ac m(t ) nc (t )

65
 Effect of Noise on SSB AM
Parallel to our discussion of2 DSB, we have
Ac
Po PM
4
1 1 S P0 Ac2 PM
Pn0 Pnc Pn
4 4 N 0 Pn0 WN 0
N
Pn S n( f )df 0 2W WN 0
2

PR PU Ac2 PM

S PR S

0 SS B
N N 0W N b
The signal-to-noise ratio in an SSB system is equivalent to that of a
DSB system.

66
 Noise in AM Receivers
A full AM signal is given
by
s t Ac 1 kam t
cos2fct
where Accos(2fct) is the carrier wave, m(t) is the message
signal and bandwidth is W, ka is a constant that determines the
percentage modulation.
We would like to perform noise analysis for an AM
system using an envelope detector.

Yu-sing Lin () 67
 Noise in AM Receivers
We perform the noise analysis of the AM receiver by first
determining the channel signal-to-noise ratio, and then the
output signal-to-noise ratio.
We can easily obtain average power of the AM signal

s t Ac cos 2fct Ac kam t cos 2fct

1 2
Ps Ac Ac2 ka P
2

1
The average power of noise in the message bandwdith is WN0
2 DSB-SC system)
(the same as the
The channel signal-to-noise ratio for AM is therefore:

A a P
2
2 1 k
SNR C,AM
c

2WN 0

68
 Noise in AM Receivers
The filtered signal x(t) applied to the envelope detector in the
receiver is given by:
x t s t n t
Ac Ac kam t n1 t cos 2fct nQ t sin 2fct
6.13
y t envelope of x t

12
Ac Ac ka m t nI t n
2 2

Assume average carrier power

t >> average noise power
y t Ac Ac ka m t nI t

The dc term or constant term Ac may be removed simply by means

of a blocking capacitor.
If we ignore the dc term Ac, we find that the remainder has a form similar to the
output of a DSB-SC receiver using coherent detection.
2

c
2
 Noise in AM Receivers
The output signal-to-noise ratio of an AM using an
envelope detector is approximately
Ac2 k a2 P
SNR O,AM
2WN0

Eq is valid only if the following two conditions are satisfied

The average noise power is small compared to the average carrier power at the
envelope detector input.
The amplitude sensitive ka is adjusted for a percentage modulation less than or
equal to 100 percent. (| ka m(t)|<=1)
 Noise in AM Receivers
Comparison of figure of merit ( AM, DSB-SC, SSB )

SNR O 2
ka SNR SNRO
1 1
SNR C AM using
Envelope
1Oka2
SNR C DSB- SNR SSB
Detector PP SC

C
The figure of merit of a DSB-SC receiver or that of an SSB
receiver using coherent detection is always unity.
The corresponding figure of merit of an AM receiver using
envelope detection is always less than unity.
In other words, the noise performance of an AM receiver is
always inferior to that a DSB-SC receiver. This is due to
the wastage of transmitter power, which results from
transmitting the carrier as a component of AM wave.
 Noise in AM Receivers
Single-Tone Modulation
Consider a sinusoidal wave of frequency fm and amplitude Am
as the modulating wave, as shown by
m(t) = Amcos(2fmt)
The corresponding AM wave is

s(t) = Ac [1+ cos(2fmt)] cos(2fct)

modulation factor : = kaAm

The average power of the modulation wave m(t) is (assuming a

load resistor of 1ohm) 1 2
P Am
2
 Noise in AM Receivers
We obtain the figure of
merit
1 2 2
SNR O 2
ka Am
2

SNR C AM
1 2 2 2 2
1 ka Am
2
When = 1 (100% modulation using envelope detection), we
get a figure of merit = 1/3.
This means that, other factors being equal, an AM system (using
envelope detection) must transmit three times as much average
power as a suppressed-carrier system (using coherent detection)
in order to achieve the same quality of noise performance.
 Effect of Noise on Angle Modulation
In this section, we study the performance of angle-modulated signals
when contaminated by additive white Gaussian noise (AWGN)
We will also compare this with the performance of AM signals.
Recall that in AM, the message is contained in the amplitude of the
modulated signal
Since noise is additive, the noise is directly added to the signal.
However, in a frequency-modulated signal, the noise is added to the
amplitude and the message is contained in the frequency of the
modulated signal.
Therefore, the message is contaminated by the noise to the extent that
the added noise changes the frequency of the modulated signal.
The frequency of a signal can be described by its zero crossings.
So the effect of additive noise on the demodulated FM signal can be
described by the changes that it produces in the zero crossings of the
modulated FM signal.
74
 Effect of Noise on Angle Modulation
A figure shown in below is the effect of additive noise on zero crossings of
two FM signals, one with high power and the other with low power.
From the previous discussion and also from the figure it should be clear
that the effect of noise in an FM system is different from that for an AM
system.
We also observe that the effect of noise in a low-power FM system is
more severe than in a high-power FM system.
In a low power signal, noise causes more changes in the zero crossings.
The analysis that we present in this chapter verifies our intuition based on
these observations.

Fig. 6.1 Effect of noise in FM

75
 Effect of Noise on Angle Modulation
The receiver for a general angle-modulated signal is shown in below
The angle-modulated signal is represented as
A cos 2f t 2k t m( )d FM
c f
u (t ) Ac cos2f ct (t ) c
Ac cos2f ct k p m(t ) PM

The AWGN nw(t) is added to u(t), and the result is passed through a
noise-limiting filter whose role is to remove the out-of-band noise.
The bandwidth of this filter is equal to that of the modulated signal
Therefore, it passes the modulated signal without distortion.
However, it eliminates the out-of-band noise.
Hence, the noise output of the filter is a filtered noise denoted by n(t).

76
 Effect of Noise on Angle Modulation
The output of this filter is
r (t ) u(t ) n(t ) u(t ) nc (t ) cos2 f ct ns (t ) sin 2 f ct

A precise analysis is complicate due to the nonlinearity of

demodulation .
Let us assume that the signal power is much higher than the noise
power.
Then, the bandpass noise is represented as

n (t )
n(t ) nc2 (t ) ns2 (t ) cos 2f ct arctan s Vn (t ) cos2f ct n (t )
where Vn(t) and n(t) represent nc (t )
the envelope and the phase of the
bandpass noise process, respectively.

77
 Effect of Noise on Angle Modulation
Assume that the signal is much larger than the noise, that is,
PVn (t ) Ac 1
The phasor diagram of signal and noise are shown in below.
From this figure, it is obvious that we can write
Vn (t ) sin n (t ) (t )
r (t ) Ac Vn (t ) cos n (t ) (t ) cos 2f c t (t ) arctan
Ac Vn (t ) cos n (t ) (t )

Ac Vn (t ) cos n (t ) (t ) cos 2f c t (t ) sin n (t ) (t )
Vn (t )

Ac

78
 Effect of Noise on Angle Modulation

k p m(t ), PM
(t )
Noting that

2k
t
f m( ) d ,

FM

We see that the output of the demodulator is given

by

sin n (t ) (t )
Vn (t )
(t ) PM
Ac
y (t )

1 d (t ) Vn (t ) sin n (t ) (t ) FM
2 dt Ac

sin n (t ) (t )
Vn (t )
k m ( t ) PM k p m(t ) Yn (t ) PM

p
Ac
1 d

sin n (t ) (t )
1 d Vn (t ) k m (t ) Yn (t ) FM
k f m(t ) FM f 2 dt
2 dt Ac

where we define
sin n (t ) (t )
Vn (t )
Yn (t ) def
Ac
79
Effect of Noise on Angle Modulation

sin n (t ) (t )
Vn (t )
k m (t ) PM k p m(t ) Yn (t ) PM

p
Ac
y (t ) 1 d

sin n (t ) (t ) FM
1 d Vn (t ) k m(t ) Yn (t ) FM
k f m(t )
f
2 dt

2 dt Ac
(Eq. 6.2.7)
The first term in above equation is the desired signal component.
The second term is the noise component.
The noise component is inversely proportional to the signal
amplitude Ac.
Hence, the higher the signal level, the lower the noise level.
This is in agreement with the intuitive reasoning presented at the
beginning of this section and based on Fig. 6.1.
This is not the case with amplitude modulation.
In AM systems, the noise component is independent of the signal
component, and a scaling of the signal power does not affect the
received noise power.
80
 Effect of Noise on Angle Modulation
The properties of the noise component
sin n (t ) (t ) Vn (t ) sin n (t ) cos (t ) Vn (t ) cos n (t ) sin (t )
Vn (t ) 1
Yn (t )
Ac Ac

1
ns (t ) cos (t ) nc (t ) sin (t )
Ac
when we compare variations in nc(t) and ns(t), we can assume that
(t) is almost constant, i.e., (t) .

Yn (t ) ns (t ) cos nc (t ) sin
1
Ac
cos sin
ns (t ) nc (t )
Ac Ac
ans (t ) bnc (t ), where a = cos/Ac and b = -sin /Ac
81
 Effect of Noise on Angle Modulation
By using the result of Exercise 5.3.3, we have

SYn ( f ) a b S nc ( f )
2 2
S nc ( f )
Ac2

Snc(f) is the power spectral density (psd) of the in-phase

component of the filtered noise given in (Eq. 5.3.10).
N0 | f | Bc
S nc ( f ) 2

0 otherwise
Note that the bandwidth of the filtered noise extends from fc Bc/2 to
fc + Bc/2 . Hence, the spectrum of nc(t) extends from Bc/2 to +Bc/2.
Therefore
N20 | f | Bc
(Eq. 6.2.13)
SYn ( f ) Ac 2

0 otherwise
82
 Effect of Noise on Angle Modulation
This equation provides an expression for the power spectral
density of the filtered noise at the front end of the receiver.
After demodulation, another filtering is applied; this reduces
the noise bandwidth to W, which is the bandwidth of the
message signal.
Note that in the case of FM modulation, as seen in (Eq. 6.2.7),
the process Yn(t) is differentiated and scaled by 1/2.
The PSD of the process (1/2) (dYn(t)/dt) is given by
(see Eq. 5.2.17)
4 2 f 2 N0
f 2
| f | (Eq.
Bc
6.2.14)
S ( f ) f 2
SYn ( f ) Ac
2 2
4 2 Y
0
n
otherwise
83
 Effect of Noise on Angle Modulation
In PM, the demodulated-noise PSD is given by (Eq. 6.2.13)
In FM, it is given by (Eq. 6.2.14).
In both cases, Bc/2 must be replaced by W after Lowpass filter.
Hence, for |f|<W
Ac2
N0
PM
S n0 ( f ) N 0 2
Ac2 f FM

Fig. 6.4 shows the power spectrum of the noise component at

the output of the demodulator for PM and FM.

Fig. 6.4 Noise power spectrum at demodulator

84 output for |f|<W in (a) PM (b) and (b) FM.
 Effect of Noise on Angle Modulation
It is interesting to note that PM has a flat noise spectrum and FM
has a parabolic noise spectrum.
Therefore, the effect of noise in FM for higher frequency
components is much higher than the effect of noise on lower
frequency components.
The noise power at the output of the lowpass filter is the noise
power in the frequency range [W, +W].
Therefore, it is given by

W N 0 df
W Ac2
2W N0
W Ac2 PM
Pn0 (Eq. 6.2.16)
S n0 ( f )df W 2 N 0W 3
W Ac2 df 3 Ac2
W N0 2 FM
f
85
 Effect of Noise on Angle Modulation
(Eq. 6.2.7) is used to determine the output SNR2in angle modulation.
k p PM PM
First, we have the output signal power PS 2
O
k f PM FM
Then the SNR, which is defined as k p2 Ac2 PM
S PSO PM
S 2 NW
def 2 2 0
N O PnO N O 3k f Ac PM
FM
2
2W N 0W

Noting that Ac2/2 is the received signal power, denoted by PR , and

p
2

PR PM
p k p max m(t ) PM
max m(t ) N 0W
PM
S
k f max m(t )
f
2
FM N O f PM
3PR max m(t ) N W

W FM
0
86
 Effect of Noise on Angle Modulation
S PM
Denote
N b N 0W

The SNR of a baseband system with the same received power,

we obtain PM p2 S
2 PM
S max m(t ) N b

O
N P 2
S

M f
3 FM
max m(t ) 2 N b

PM / max m(t )
2

: The average-to-peak-power-ratio of the

message signal (or equivalently, the power content of the
2 S
normalized message, PMn ).
p M
P PM
S n
N b

N O 3 2 P S FM
87 f M n N b
 Effect of Noise on Angle Modulation
Now using Carson's rule Bc = 2( + 1)W, we can express the
output SNR in terms of the bandwidth expansion factor, which
is defined as the ratio of the channel bandwidth to the
message bandwidth and is denoted by :
2 1
Bc
def
W

From this relationship, we have =/2 1.

Therefore, 2
2 1 S

PM PM
S max m(t ) N b

2
N O 2 1 S

3PM max m(t ) N FM

b
88
 Effect of Noise on Angle Modulation
p PM
2
1 2 S
PR PM 2
max m(t ) N 0W
PM PM
S S max m(t ) N b
2
N O f PM N O
2
2 1 S
3PR max m(t ) N W FM 3PM max m(t ) N FM
0 b

Observations
In both PM and FM, the output SNR is proportional to 2.
Therefore, increasing increases the output SNR.
Increasing increase the bandwidth (from Carsons rule).
So angle modulation provides a way to trade off
bandwidth for transmitted power.

89
Effect of Noise on Angle Modulation
Although we can increase the output SNR by increasing ,
having a large means having a large Bc (by Carson's rule).
Having a large Bc means having a large noise power at the
input of the demodulator. This means that the approximation
P(Vn(t) << Ac) 1 will no longer apply and that the preceding
analysis will not hold.
In fact, if we increase such that the preceding
approximation does not hold, a phenomenon known as the
threshold effect will occur and the signal will be lost in the
noise.
This means that although increasing the modulation index, ,
up to a certain value improves the performance of the
system, this cannot continue indefinitely.
After a certain point, increasing will be harmful and
deteriorates the performance of the system.
90
Effect of Noise on Angle Modulation
A comparison of the preceding result with the SNR in AM
shows that, in both cases (AM and angle modulation),
increasing the transmitter power (and consequently the
received power) will increase the output SNR
But the mechanisms are totally different. In AM, any
increase in the received power directly increases the signal
power at the output of the demodulator.
This is basically because the message is in the amplitude of
the transmitted signal and an increase in the transmitted
power directly affects the demodulated signal power.
However, in angle modulation, the message is in the phase
of the modulated signal and increasing the transmitter
power does not increase the demodulated message power.
In angle modulation, the output SNR is increased by a
decrease in the received noise power, as seen from
Equation (6.2.16) and Fig. 6.1.
91
Effect of Noise on Angle Modulation
In FM, the effect of noise is higher at higher frequencies.
This means that signal components at higher frequencies
will suffer more from noise than signal components at
lower frequencies.
To compensate for this effect, preemphasis and
deemphasis filtering are used.

92