Module 4

NOISE IN ANALOG
MODULATION
• Aim is to resume study of CW modulation systems by
evaluating the effects of noise on their Performance
• To undertake the analysis of the same we need number of
things like,
1) We must have a receiver model
2) We must customize and model receiver noise as
[additive, White, and Gaussian]
 Receiver Model
The idea of modeling is fundamental to the study of all physical
systems. It helps to understand the capabilities and limitations
of the system.
To formulate study of noise in CW modulation we need to keep
following key points in mind,
i) The model provides an adequate description of the form of
receiver noise that is of common concern
ii) The model accounts for the inherent filtering and
modulation characteristics of the system
 Receiver Model
iii) The model is simple enough for statistical analysis of the
system to be possible.

Modulated X(t) Output

Signal S(t) BANDPASS Signal
DEMODULATOR
FILTER

Noise W(t)

Fig- Noisy Receiver Model

 Receiver Model
• In the model S(t) denotes incoming modulated signal
and W(t) denotes front end receiver noise.
• Figure represents combined filtering action of the
tuned amplifiers used in the actual receiver for the
purpose of signal amplification prior to de-modualtion.
• The bandwidth of BPF is just wide enough to pass the
modulated signal S(t) without distortion.
• As the demodulator ckt used naturally depend upon
the type of modulation used.
 Receiver Model
• In the model we assume the noise w(t) to be additive,
White and Gaussian.
• We thus consider the power spectral density of noise is
No/2, defined for both positive and negative
frequencies.
• No is average noise power per unit bandwidth
measured at the front end of the receiver.
• Let the system have Bandwidth mentioned as BT
 Receiver Model
• Taking the mid-band frequency of the band pass filter
to be the same as the carrier frequency fc, we may
model power spectral density SN(f) of the noise n(t)
resulting from the passage of white noise W(t) through
the filter can be given as,

• fc>>BT
 Receiver Model
• The filtered noise component n(t) as narrowband noise
represented in canonical form as,
n(t)= nI(t) cos(2Πfct) – nQ(t) sin(2Πfct)
• Where nI(t) is In phase and nQ(t) is quadrature phase
component both measured wrt carrier wave.
• From the noisy receiver model,
X(t)= s(t) + n(t)
where s(t) details depend upon type of modulation used. In
any event average noise power at the demodulator input is
equal to total area under curve of power spectral density SN(f).
 Receiver Model
• For figure 6.2 the average noise power is equal to N0BT.
• Looking into demodulator perspective we can have 2 kinds
of measures,
Input Signal to Noise ratio
= Average power of Modulated signal S(t)
Average power of filtered noise n(t)

Output Signal to Noise ratio

= Average power of demodulated signal
Average power of noise
where both measured at receiver output.
 Receiver Model
• However on comparison among different models to be
meaningful value, it must be made on equal basis as shown,
• The S(t) transmitted by each system has the same
average power
• The front end receiver noise w(t) has the same average
power measured in the message bandwidth W.
Accordingly as frame of reference Channel-SNR can be
given by,
SNRC = Average power of Modulated signal S(t)
Average power of noise in message bandwidth
 Receiver Model
• Thus for purpose of comparing different CW modulation
systems we define a Figure of Merit as,
Figure of Merit= (SNR)o
(SNR)c

The higher the value of Figure of Merit the better will

be the performance of receiver.

It may be equal to 1 , be less than 1 or greater than 1

depend upon type of modulation used.
 Noise in DSB-SC Receiver
Coherent Detector
Output
V(t) Signal
Modulated X(t)
BANDPASS Product Y(t)
Signal S(t) LPF
FILTER Modulator

Noise W(t)

• The DSBSC system employing Coherent Detector is proposed,

• The incoming signal is multiplied with local oscillator frequency
Cos 2Πfct
• However it is necessary that Local oscillator should be exactly in
Synchronous with Transmitter
 Noise in DSB-SC Receiver
• The DSBSC component of the filtered signal X(t) is expressed as,

S(t)= C.Ac.Cos(2Πfct) m(t) -------1

Where Ac.Cos(2Πfct) is carrier component
C is system dependent scaling factor
The average power P of the message signal is the total area under
the curve of power spectral density, as shown by
-------2
 Noise in DSB-SC Receiver
• Using the example 12 of chapter 5 on modulated random
process, we may express average power DSBSC modulated
signal component S(t) as, Ac2C2P/2.
• With noise spectral density No/2, the average noise power in
the message bandwidth W is equal to Wno.
• The channel SNR of the DSBSC modulation is there fore,
(SNR)c,DSB = Ac2C2P ------------3
2WNo
Using the previous relation of receiver model WKT,
X(t)= s(t) + n(t)
 Noise in DSB-SC Receiver
X(t)= s(t) + n(t)
= C.Ac.Cos(2Πfct) m(t)+ nI(t).cos(2Πfct) – nQ(t)sin(2Πfct)
The output of Product modulator component of the coherent
detector is therefore,
V(t)= X(t). Cos(2Πfct)
= ½ C.Ac.m(t)+ ½ nI(t) + ½ [C.Ac.m(t)+nI(t) cos(4Πfct)-
1/2 Ac.nQ(t) sin(2Πfct)]
The lowpass filter in the coherent detector removes the high
frequency components of V(t) yielding receiver output
Y(t)= ½ C.Ac.m(t) + ½ nI(t)
 Noise in DSB-SC Receiver
X(t)= s(t) + n(t)
= C.Ac.Cos(2Πfct) m(t)+ nI(t).cos(2Πfct) – nQ(t)sin(2Πfct)
The output of Product modulator component of the coherent
detector is therefore,
V(t)= X(t). Cos(2Πfct)
= ½ C.Ac.m(t)+ ½ nI(t) + ½ [C.Ac.m(t)+nI(t) cos(4Πfct)-
1/2 Ac.nQ(t) sin(2Πfct)]
The lowpass filter in the coherent detector removes the high
frequency components of V(t) yielding receiver output
Y(t)= ½ C.Ac.m(t) + ½ nI(t) ------------4
 Noise in DSB-SC Receiver
Equation 3 indicates following,
• The m(t) and In-phase component of noise appear
additively at the output of receiver
• The quadrature component is completely rejected by
the coherent detector
• The message signal component at the receiver output is
given by C.Ac.m(t)/2
• On taking average power of it we get
C2Ac2P/4 where P is Avg power of signal m(t).
 Noise in DSB-SC Receiver
The Band pass filter used in this case is assumed to have
Bandwidth BT equal to 2W. Hence average power of filtered
noise can become 2WNo.

The average power at the receiver output is (1/2)2 2WNo

The output SN for DSBSC receiver using Coherent receiver
is therefore
( SNR)o= C2Ac2P/4 = C2Ac2P ---------------5
WNo/2 2WNo
 Noise in DSB-SC Receiver
Using equation 3 and 5, the figure of Merit can be given as,
Figure of Merit= (SNR)o == 1
(SNR)c
 Noise in AM Receiver
Output
Signal
AM X(t)
BANDPASS Y(t)
Signal S(t)
Envelope Detector
FILTER

Noise W(t)

• The AM system employing Envelope Detector is proposed,

• The incoming signal S(t)= Ac [1+ka.m(t)]Cos(2Πfct)
• The average power of carrier component in AM signal S(t) is
Ac2/2
 Noise in AM Receiver
• The average power of full AM signal s(t) is given by
Ac2(1+ka2P)/2
Considering message bandwidth as Wno. The Channel Signal to
Noise ratio for AM is therefore,
(SNR)c,AM = Ac2(1+ka2P) ------------1
2WNo
Now similar to earlier case lets apply analysis to find Y(t) using X(t)
as,
X(t)= s(t) + n(t)
= Ac+ Ac Ka.m(t)+ nI(t).cos(2Πfct) – nQ(t)sin(2Πfct) -------------2
It is informative to represent this signal x(t) by means of phasar as
Noise in AM Receiver
 Noise in AM Receiver
From the phasor diagram the receiver output is readily obtained as,
Y(t)= envelope of x(t)
= {[Ac+ Ac Ka.m(t)+ nI(t)2– nQ(t) 2}]

The expression obtained is somewhat complex and needs to be

observed in simplified way to derive meaningful results.
Here the term Ac+ Ac Ka.m(t) is usually large compared with the
terms like nI(t) and nq(t).
Also the phase term is of NO INTEREST to us as envelope detector
is insensitive to phase changes. Thus nq(t) can be ignored.
 Noise in AM Receiver
• Then y(t)= {[Ac+ Ac Ka.m(t)+ nI(t)2 ]
• As Ac is constant term which represents just a carrier amplitude
and can be considered as part with NO INFORMATION. Thus,
• y(t)= { Ac Ka.m(t)} indicates almost repetition of DSBSC signal, so
( SNR)o= Ka2Ac2P ---------------3
2WNo

Using equation 1 and 3, the figure of Merit can be given as,

Figure of Merit= (SNR)o = Ka2P

(SNR)c 1+ Ka2P
 Noise in AM Receiver
To conclude,
THE NOISE PERFORMANCE OF AN AM RECEIVER IS ALWAYS LESS
THAN THAT OF A DSBSC RECEIVER AS ITS FIGURE OF MERIT IS LESS
THAN UNITY.
 NOISE IN FM RECEIVERS
Output
Signal
FM X(t) Y(t)
Signal S(t) BANDPASS Discriminator BASEBAND
Limiter
FILTER LOWPASS
FILTER

Noise W(t)

Key points,
In FM systems message information is transmitted by variations of
instantaneous frequency of sinusoidal carrier wave and its
amplitude maintained constant. Thus any variations in carrier
amplitude must result from noise or interference.
 NOISE IN FM RECEIVERS
• Amplitude limiter used to remove these variations by clipping
modulated wave at the filter output to almost zero axis.
• The resulting wave is rounded off by another bandpass filter
that is an integral part of limiter thus suppressing harmonics.
• Thus filter (limiter) output is again sinusoidal with amplitude
that is practically independent of the carrier amplitude at
receiver input.
• The discriminator has 2 components,
 NOISE IN FM RECEIVERS
• The discriminator has 2 components,
• A slope network or differentiator
• An envelope detector that recovers amplitude variations
and thus produces message signal

• The post detection filter [Base band LPF] has bandwidth large
enough to accommodate high frequency component of
message signal. It removes out of band components of noise at
discriminator output and there by keeps effect of output noise
to minimum.
NOISE IN FM RECEIVERS
 NOISE IN FM RECEIVERS
• The incoming FM Signal is given by,
S(t)= Ac Cos [(2Πfct)+ 2ΠKf
Considering Φ(t)= 2ΠKf
Then equation reduces to,
S(t)= Ac Cos [(2Πfct)+Φ(t)]
Thus the noisy signal at the output of bandpass filter is,
X(t)= s(t) + n(t)
= Ac Cos [(2Πfct)+Φ(t)]+nI(t).cos(2Πfct) – nQ(t)sin(2Πfct)
 NOISE IN FM RECEIVERS
• Thus by applying various analogies and statistics, the average
power of output noise is given by,
Average power of output noise = 2NoW3
3Ac2

• The resulting (SNR)o FM = 3Ac2KfP/ 2NoW3

• The average power in the modulated signal s(t) is Ac2/2.
• (SNR)c FM= Ac2/2WNo.

Figure of Merit= (SNR)o = 3Kf2P

(SNR)c W2
 NOISE IN FM RECEIVERS
• We know that Δf is proportional to Kf of modulator and
Deviation ratio D is ratio of Δf/W.
• Thus we may state that When the carrier to noise ratio is high,
an increase in transmission bandwidth Bt provides
corresponding quadratic increase in output signal to noise ratio
or Figure of merit of FM System.

• Practice Example 6.2

 Capture Effect
• The inherent ability of FM systems is to minimize the effects of
unwanted signals.
• The same is applicable for interference signal from another
FM signal also which is having closer Carrier frequency that of
desired signal.
• Usually FM receiver suppress the weaker FM signal if it is
having less strength than the desired Signal.
• If interference signal is Strong than FM receiver locks onto
strong signal there by suppresses Desired signal.
• If both are of nearly equal strength , the FM receiver
fluctuates back and forth between them.
• This phenomenon is known as CAPTURE EFFECT.
 FM Threshold Effect
• It is found experimentally that as input noise power is
increased in FM, so that carrier to noise ratio is decreased,
then the FM receiver Breaks.
• At first individual clicks are heard and as carrier to noise ratio
decreases still further, the clicks rapidly merged into Crackling
& Sputtering sound.
• Near the braking point (SNR)o FM = 3Ac 2KfP/ 2NoW3 starts
failing by predicting values of output signal to noise ratio
larger than the actual ones.
• This phenomenon is called FM THRESHOLD EFFECT.
 FM Threshold Effect
• The threshold is defined as minimum carrier to noise ratio
yielding an FM improvement that is not significantly
deteriorated from the value predicted by the usual signal to
noise formula assuming a small noise power.
 FM Threshold Reduction
• In specific applications such as space communication we
require lesser threshold in FM receiver.
• FM threshold can be reduced by FM demod with negative
feedback or PLL. Such devices are referred as EXTENDED
THRESHOLD DEMODULATORS OR FMFB DEMODULATOR.
Output
Signal
FM BASEBAND Y(t)
Wave BANDPASS Limiter
MIXER LOWPASS
FILTER Discriminator
FILTER

VCO
 FM Threshold Reduction
• An FMFB demodulator is essentially a Tracking filter that can
track only the slowly varying frequency of a WBFM and
consequently it responds only to a narrow band of noise
centered about instantaneous carrier frequency.
• FMFB receiver is capable of realizing threshold extension on
the order of 5-7db, which represents a significant
improvement in design of minimum power FM systems.
Pre-Emphasis & De-Emphasis in FM
 Pre-Emphasis & De-Emphasis in FM
• It is a process widely used in commercial FM radio
transmission and reception.
• It is seen from above graphs that power spectral density of
message signal usually falls of appreciably at higher
frequencies and power spectral density of noise signal
increases rapidly with frequency.
• In pre-emphasis we artificially emphasize the high frequency
components of message signal prior to modulation in
transmitter & there fore before the noise is introduced in Rx.
 Pre-Emphasis & De-Emphasis in FM
• In effect of this the low frequency and high frequency
components portions of power spectral density of messages
are equalized in such way that message fully occupies the
band allotted to it.
• At receiver we de-emphasize high frequency components so
as to restore the original signal power distribution of the
message.
• Thus high frequency components of noise at discriminator
output are also reduced, thus effectively increasing SNR.
 Pre-Emphasis & De-Emphasis in FM
Message
Plus
noise
m(t)
PRE FM De-emphasis
FM
EMPHASIS Receiver filter
Transmitter
FILTER
Noise W(t)

• Practice Example 6.3

COMPLETE