Module IV

Part 2
 MODULE IV
• Noise: external, internal- signal-to-noise ratio – noise figure –noise
calculations – equivalent noise temperature,-noise figure in cascaded
networks. White noise, filtered white noise, noise equivalent bandwidth,
Narrow band noise.
• Effect of noise in Systems: Linear and angle modulation systems, threshold
effect and threshold extension, pre-emphasis and de-emphasis filtering.

References:
• Simon Haykin and Michael Moher, Communication Systems, John Wiley &
Sons, 5/e, (2017).
• B. P. Lathi and Zhi Ding, Modern Digital and Analog Communication
Systems, Oxford University Press, 4/e,(2017)
• K.N. Hari Bhat, D. Ganesh Rao, Analog Communication: simplified
Approach, Pearson

CE I Module IV Prof. (Dr.) Shahana T.K. 2

 NOISE IN ANALOG COMMUNICATIONS
• Modulated signals are perturbed by noise and by the imperfect characteristics of
the channel during transmission.
• There may be many sources of noise in a communication system
• The major sources are the communication devices themselves or interference
encountered during the course of transmission.
• The most common way that noise can affect the desired signal, is as an additive
distortion.
• That is, the received signal is modeled as
𝑟 𝑡 = 𝑠 𝑡 + 𝑤(𝑡)
where 𝑠 𝑡 : the transmitted signal and
𝑤(𝑡): the additive noise
• This module focusses on the comparison of performance of different modulation
schemes, for the detection of analog signals in the presence of additive noise.

CE I Module IV Prof. (Dr.) Shahana T.K. 3

• To compute noise power:
• We must measure the noise over a specified bandwidth.
• We have seen that the noise power at the output of a filter of equivalent-noise
bandwidth 𝐵𝑇 is:
𝑁 = 𝑁0 𝐵𝑇
• The smaller the bandwidth 𝐵𝑇 , the smaller the noise power N will be.
• When the received signal 𝑟(𝑡) is passed through a filter of bandwidth 𝐵𝑇 , where
𝐵𝑇 is the transmission bandwidth of the signal, the signal after initial filtering is

𝑥 𝑡 = 𝑠 𝑡 + 𝑛(𝑡)
Where 𝑛(𝑡) is narrowband noise, as contrasted to 𝑤(𝑡) which is assumed to be
white.
• Both 𝑠 𝑡 𝑎𝑛𝑑 𝑛(𝑡) are random processes with zero-mean.

CE I Module IV Prof. (Dr.) Shahana T.K. 4

• High-level block diagram of a communication receiver:

• The demodulator depends on the type of modulation used.

• Signal-to-noise ratio (SNR) is used as the measure of quality for analog systems
• For zero-mean processes, the signal-to-noise ratio is defined as the ratio of the
variances of the desired and undesired signals.
𝐸[𝑠 2 𝑡 ]
𝑆𝑁𝑅 =
𝐸[𝑛2 𝑡 ]
• Consequently, the signal-to-noise ratio is often considered as ratio of the average
signal power to the average noise power.
CE I Module IV Prof. (Dr.) Shahana T.K. 5
• The signal-to-noise ratio can be measured at several points in the receiver:
➢ If the SNR is measured at the front-end of the receiver, then it is usually a measure of the
quality of the transmission link and the receiver front-end.
➢ If the SNR is measured at the output of the receiver, it is a measure of the quality of the
recovered information-bearing signal.
• In the receiver block diagram, the SNR measured at the input to the demodulator
is referred to as the pre-detection signal-to-noise ratio.
• The SNR measured at the output of the demodulator is referred to as the post-
detection signal-to-noise ratio

❖A reference transmission model

• This reference model is equivalent to transmitting the message at baseband.
• Two assumptions are made:
• The message power is the same as the modulated signal power of the modulation scheme
under study.
• The baseband low-pass filter passes the message signal and rejects out-of-band noise.

CE I Module IV Prof. (Dr.) Shahana T.K. 6

• The reference signal-to-noise ratio is:

• The reference signal-to-noise ratio may be used to compare different

modulation–demodulation schemes by using it to normalize the post-detection
signal-to-noise ratios.

CE I Module IV Prof. (Dr.) Shahana T.K. 7

• The figure of merit (FoM) for a particular modulation–demodulation scheme is
defined as:
postdetection SNR
Figure of merit =
reference SNR
• The higher the value of figure of merit, the better is the noise performance of the
receiver.
• To summarize our consideration of signal-to-noise ratios:
✓ The pre-detection SNR is measured before the signal is demodulated.
✓ The post-detection SNR is measured after the signal is demodulated.
✓ The reference SNR is defined on the basis of a baseband transmission model.
✓ The figure of merit is a dimensionless metric for comparing different analog modulation–
demodulation schemes

CE I Module IV Prof. (Dr.) Shahana T.K. 8

Noise in Linear Receivers Using Coherent Detection (DSB-SC)
• In doublesideband suppressed-carrier (DSB-SC) modulation, the modulated signal
is represented as

where 𝑓𝑐 is the carrier frequency, and 𝑚 𝑡 is the message signal; the carrier phase
𝜃 is a random variable, but not varying during the course of transmission.
• For suppressed-carrier signals, linear coherent detection was identified as the
proper demodulation strategy.
• A linear receiver for this signal could be implemented as shown in Fig.
• The received RF signal is the sum of the modulated signal and white Gaussian
noise w(t).
• The received signal 𝑓𝑅𝐹 is down-converted to an IF by mixing with a sinusoid of
frequency 𝑓𝑐 − 𝑓𝑅𝐹
CE I Module IV Prof. (Dr.) Shahana T.K. 9
• The down-conversion is performed by the product modulator shown as mixer 1 in
Fig.
Multiplier

A linear DSB-SC receiver using coherent demodulation

• After band-pass filtering, the resulting signal is: 𝑥 𝑡 = 𝑠 𝑡 + 𝑛(𝑡)
Where s(t) is the undistorted modulated signal and n(t) is the band-pass noise at
the output of the filter.

CE I Module IV Prof. (Dr.) Shahana T.K. 10

• The assumed power spectral density of the band-pass noise is illustrated in Fig.
below

• The band-pass filter has a sufficiently wide and flat passband that does not cause
any significant distortion to the modulated signal

CE I Module IV Prof. (Dr.) Shahana T.K. 11

PRE-DETECTION SNR
• For DSB-SC,
• The average power of the signal component is given by expected value of the
squared magnitude.

• Let the message power be, 𝑃 = 𝐸[𝑚2 𝑡 ]

• So, the average received signal power due to the modulated component is
𝐴2𝑐 𝑃
𝐸 𝑠2 𝑡 =
2
• If the band-pass filter has a noise bandwidth 𝐵𝑇 , then the noise power passed by
this filter is
𝑁 = 𝑁0 𝐵𝑇

CE I Module IV Prof. (Dr.) Shahana T.K. 12

• Hence, the pre-detection signal-to-noise ratio of the DSB-SC system is:

• POST-DETECTION SNR
• It is the ratio of the message signal power to the noise power after
demodulation/detection.
• depends on both the modulation and demodulation techniques used.
• The narrowband representation of the band-pass noise is:

• So, the signal at the input to the coherent detector is,

CE I Module IV Prof. (Dr.) Shahana T.K. 13

• The output of mixer 2 (in Fig. ) is given by
𝑣 𝑡 = 𝑥 𝑡 cos 2𝜋𝑓𝑐 𝑡
= [𝑠 𝑡 + 𝑛 𝑡 ] cos 2𝜋𝑓𝑐 𝑡
= s(t)cos 2𝜋𝑓𝑐 𝑡 + 𝑛(𝑡) cos 2𝜋𝑓𝑐 𝑡
= 𝐴𝑐 𝑚 𝑡 𝑐𝑜𝑠 2 2𝜋𝑓𝑐 𝑡 + 𝑛𝐼 𝑡 𝑐𝑜𝑠 2 2𝜋𝑓𝑐 𝑡 − 𝑛𝑄 𝑡 𝑠𝑖𝑛 2𝜋𝑓𝑐 𝑡 cos 2𝜋𝑓𝑐 𝑡
1 + cos 4𝜋𝑓𝑐 𝑡 1 + cos 4𝜋𝑓𝑐 𝑡 sin 4𝜋𝑓𝑐 𝑡
= 𝐴𝑐 𝑚 𝑡 + 𝑛𝐼 𝑡 − 𝑛𝑄 𝑡
2 2 2
1 1 1
= 𝐴𝑐 𝑚 𝑡 + 𝑛𝐼 𝑡 + 𝐴𝑐 𝑚 𝑡 + 𝑛𝐼 𝑡 cos 4𝜋𝑓𝑐 𝑡 − 𝑛𝑄 𝑡 sin 4𝜋𝑓𝑐 𝑡
2 2 2
{

}
• The above signal is passed through a LPF, which passes only the based band spectrum.

CE I Module IV Prof. (Dr.) Shahana T.K. 14

• Hence, the signal at the demodulator output is:
1 1
𝑦 𝑡 = 𝐴 𝑐 𝑚 𝑡 + 𝑛𝐼 𝑡
2 2
= 𝑚 𝑑 𝑡 + 𝑛𝑑 𝑡
• Two observations can be made:
• The message signal 𝑚 𝑡 and the in-phase component of the filtered noise
𝑛𝐼 𝑡 appear additively in the output.
• The quadrature component of the noise 𝑛𝑄 𝑡 is completely rejected by the
demodulator
1 2
• The post-detection signal power : 𝐸 𝑚𝑑2 𝑡 = 𝐴 𝑃
4 𝑐

CE I Module IV Prof. (Dr.) Shahana T.K. 15

• PSD of filtered white noise n(t):
✓ Bandwidth of the bandpass filter is
𝐵𝑇 = 2𝐵, where B is the message bandwidth

• PSD of in-phase and quadrature components

CE I Module IV Prof. (Dr.) Shahana T.K. 16

• The average noise power of 𝑛𝑑 𝑡 :
1
𝐸 𝑛𝑑2
𝑡 = 𝐸 𝑛𝐼2 𝑡
4
1
= × 𝐴𝑟𝑒𝑎 𝑢𝑛𝑑𝑒𝑟 𝑡ℎ𝑒 𝑃𝑆𝐷 𝑐𝑢𝑟𝑣𝑒 𝑜𝑓𝑛𝐼 𝑡
4
1
= × 2𝑁0 𝑊, where W is the message bandwidth
4
𝑁0 𝑊
∴ The post-detection noise power =
2

• Hence, Post-detection SNR is:

1 2
𝐴𝑐 𝑃 𝐴2𝑃
𝐷𝑆𝐵
𝑆𝑁𝑅𝑝𝑜𝑠𝑡 = 4 =
𝑐
𝑁0 𝑊 2𝑁0 𝑊
2
CE I Module IV Prof. (Dr.) Shahana T.K. 17
• FIGURE OF MERIT
𝐴2𝑐 𝑃
• For DSB-SC modulation, the average modulated message power is
2
• The average noise power for a message of bandwidth W is 𝑁0 𝑊
𝐴2𝑐 𝑃
• So, the reference SNR is 𝑆𝑁𝑅𝑟𝑒𝑓 =
2𝑁0 𝑊
• Hence, for DSB-SC modulation
𝐴2𝑐 𝑃
𝐷𝑆𝐵
𝑆𝑁𝑅𝑝𝑜𝑠𝑡 2𝑁0 𝑊
Figure of Merit = = 2 =1
𝑆𝑁𝑅𝑟𝑒𝑓 𝐴𝑐 𝑃
2𝑁0 𝑊

CE I Module IV Prof. (Dr.) Shahana T.K. 18

 Noise In AM Receivers Using Envelope Detection
• Envelope detection of amplitude modulation is used in AM receivers with non-
suppressed carrier.
• It does not require the circuitry to produce a synchronized carrier for
demodulation.
• This results in a simpler receiver than the coherent approach.
• The AM signal is given by
𝑠 𝑡 = 𝐴𝑐 (1 + 𝑘𝑎 𝑚 𝑡 )cos(2𝜋𝑓𝑐 𝑡)
where 𝐴𝑐 cos(2𝜋𝑓𝑐 𝑡) is the carrier wave, 𝑚 𝑡 is the message signal, and 𝑘𝑎 is
the amplitude sensitivity of the modulator
• The front end of this receiver is identical to that of the coherent receiver.

CE I Module IV Prof. (Dr.) Shahana T.K. 19

• Receiver model:

• PRE-DETECTION SNR
• 𝑠 𝑡 = 𝐴𝑐 (1 + 𝑘𝑎 𝑚 𝑡 )cos(2𝜋𝑓𝑐 𝑡)
𝐴2𝑐
• The average power of the carrier component is due to the sinusoidal nature of
2
the carrier
• The power in the modulated part of the signal is
We assume the message signal has zero
mean, and the message power is P

CE I Module IV Prof. (Dr.) Shahana T.K. 20

 2 𝐴2𝑐
• So, the received signal power 𝐸 𝑠 𝑡 = 1 + 𝑘𝑎2 𝑃
2
• The noise power passed by the bandpass filter is
𝑁 = 𝑁0 𝐵𝑇
• Hence, Pre-detection SNR is:
𝐴2𝑐
1 + 𝑘𝑎2 𝑃 𝐴 2 1 + 𝑘2𝑃
𝐴𝑀
𝑆𝑁𝑅𝑝𝑟𝑒 = 2 =
𝑐 𝑎
𝑁0 𝐵𝑇 2𝑁0 𝐵𝑇
POST-DETECTION SNR
• The envelope detector can be modeled as shown in Fig.

CE I Module IV Prof. (Dr.) Shahana T.K. 21

• The input to the envelope detector is

• This can be represented in a phasor diagram as shown in Fig.

• The signal component of the phasor is 𝐴𝑐 (1 + 𝑘𝑎 𝑚 𝑡 ), and the noise has two
orthogonal phasor components, 𝑛𝐼 𝑡 and 𝑛𝑄 𝑡 .

CE I Module IV Prof. (Dr.) Shahana T.K. 22

• The output of the envelope detector is the amplitude of the phasor representing x(t)
• It is given by

• If we assume that the signal is much larger than the noise, then using the approximation

• We may write , under high SNR conditions

• This new expression for the demodulated signal has three components:
• dc component
• signal component
• noise component.
• The dc term can be removed with a capacitor, as shown in Fig.
• So, finally 𝑦 𝑡 = 𝐴𝑐 𝑘𝑎 𝑚 𝑡 + 𝑛𝐼 𝑡

CE I Module IV Prof. (Dr.) Shahana T.K. 23

• Accordingly, the post-detection SNR for the envelope detection of AM, using a
message bandwidth W, is given by
𝐴 2𝑘2𝑃
𝐴𝑀 𝑐 𝑎
𝑆𝑁𝑅𝑝𝑜𝑠𝑡 =
2𝑁0 𝑊
• where 𝐴2𝑐 𝑘𝑎2 𝑃 represents the average power of the message and 2𝑁0 𝑊
represents the average power of noise 𝑛𝐼 𝑡
FIGURE OF MERIT of AM
𝐴2𝑐
• For AM modulation, the average modulated message power is 1 + 𝑘𝑎2 𝑃
2
• The average noise power for a message of bandwidth W is 𝑁0 𝑊
• So, the reference SNR is
𝐴2𝑐 1+𝑘𝑎2𝑃
𝑆𝑁𝑅𝑟𝑒𝑓 =
2𝑁0 𝑊

CE I Module IV Prof. (Dr.) Shahana T.K. 24

• Hence, for DSB-SC modulation
𝐴2𝑐 𝑘𝑎2 𝑃
𝐴𝑀
𝑆𝑁𝑅𝑝𝑜𝑠𝑡 2𝑁0 𝑊 𝑘𝑎2 𝑃
Figure of Merit = = 2 2 =
𝑆𝑁𝑅𝑟𝑒𝑓 𝐴𝑐 1 + 𝑘𝑎 𝑃 1 + 𝑘𝑎2 𝑃
2𝑁0 𝑊
• Since 𝑘𝑎2 𝑃 is always less than unity (to avoid over modulation), the figure of merit for
this system is always less than 0.5
𝐴2
𝐴2𝑚 𝑘𝑎 2𝑚
2
𝜇2
• As the message power 𝑃 = , FOM = 𝐴2
=
2 1+𝑘𝑎 2𝑚
2 2+𝜇2

where 𝜇 = 𝑘𝑎 𝐴𝑚 , is the modulation index

1
• When 𝜇 = 1, FOM =
3
• This means that, this AM system must transmit at 3 times the average power as that
of suppressed carrier system to achieve the same quality of performance.
• Hence, the noise performance of an envelope-detector receiver is always inferior to a
DSB-SC receiver, the reason is wastage of the power in the carrier component of AM.
CE I Module IV Prof. (Dr.) Shahana T.K. 25
 Noise in SSB Receivers
• Assume that the lower sideband is transmitted, so that we may express the SSB
modulated wave as:

Where 𝑚(𝑡)
ෝ is the Hilbert transform of the message signal m(t).
• We may make the following observations:
• m(t) and 𝑚(𝑡)
ෝ are uncorrelated with each other.
• Their power spectral densities are additive.
• The Hilbert transform 𝑚(𝑡)
ෝ is obtained by passing m(t) through a linear filter with transfer
function −𝑗 sgn(𝑓).
𝜋
• So 𝑚(𝑡)
ෝ is nothing but m(t) in which all the frequency components are shifted in phase by
2
radians.
• The average powers of m(t) and 𝑚(𝑡)
ෝ are identical and is P.

CE I Module IV Prof. (Dr.) Shahana T.K. 26

• The average signal power is:
2 2
2
𝐴 𝑐 𝐴𝑐
𝐸 𝑠 𝑡 =𝐸 𝑚 𝑡 cos 2𝜋𝑓𝑐 𝑡 +𝐸 𝑚ෝ 𝑡 sin 2𝜋𝑓𝑐 𝑡
2 2
𝐴2𝑐 1 2 𝐴2𝑐 1 2 𝐴2𝑐
= 𝑃( ) + 𝑃( ) = 𝑃
4 2 4 2 4
PRE-DETECTION SNR
• For SSB signal, the transmission bandwidth 𝐵𝑇 is same as the message bandwidth
W.
• So, noise power at the BPF output is 𝑁0 𝑊
• Hence, Pre-detection SNR is:
𝐴2𝑐
𝑃 𝐴 2
𝑐𝑃
𝑆𝑆𝐵
𝑆𝑁𝑅𝑝𝑟𝑒 = 4 =
𝑁0 𝑊 4𝑁0 𝑊

CE I Module IV Prof. (Dr.) Shahana T.K. 27

POST-DETECTION SNR
• The same receiver model as that of DSB is considered.

• x(t) after multiplication with the synchronous oscillator cos 2𝜋𝑓𝑐 𝑡 is:

CE I Module IV Prof. (Dr.) Shahana T.K. 28

• With a band-pass filter tailored for the SSB signal, the band-pass noise n(t) will
also be of single sideband nature.
• The spectrum of the in-phase component of the noise 𝑛𝐼 𝑡 is given by
𝑆 𝑓 − 𝑓𝑐 + 𝑆𝑁 𝑓 + 𝑓𝑐 , −𝑊 ≤ 𝑓 ≤ 𝑊
𝑆𝑁𝐼 𝑓 = ቊ 𝑁
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

Figures (a) and (b) show the pre-detection (BPF)

and post-detection (LPF) output noise spectra of
SSB(LSB)

𝑁
For SSB, 𝑆𝑁 (𝑓) is 0 for 𝑓𝑐 − 𝑊 < 𝑓 < 𝑓𝑐
2
and for −𝑓𝑐 < 𝑓 < −𝑓𝑐 + 𝑊. So

CE I Module IV Prof. (Dr.) Shahana T.K. 29

• The receiver output :

𝐴2𝑐
• The message signal power at the receiver output is 𝑃
16
1
• The corresponding noise power is 𝑁0 𝑊
4
• Accordingly, the post-detection SNR for SSB receiver:
𝐴2𝑃
𝑆𝑆𝐵 𝑐
𝑆𝑁𝑅𝑝𝑜𝑠𝑡 =
4𝑁0 𝑊
FIGURE OF MERIT
𝐴2𝑐
• The average signal power for the SSB system is 𝑃
4
• The average noise power for a message of bandwidth W is 𝑁0 𝑊
𝐴2𝑐 𝑃
• So, the reference SNR is: 𝑆𝑁𝑅𝑟𝑒𝑓 =
4𝑁0 𝑊

CE I Module IV Prof. (Dr.) Shahana T.K. 30

• Hence, for SSB modulation
𝐴2𝑐 𝑃
𝑆𝑆𝐵
𝑆𝑁𝑅𝑝𝑜𝑠𝑡 4𝑁0 𝑊
Figure of Merit = = 2 =1
𝑆𝑁𝑅𝑟𝑒𝑓 𝐴𝑐 𝑃
4𝑁0 𝑊
• SSB transmission has the same figure of merit as DSB-SC
• Comparing the results for the different amplitude modulation schemes:
• DSB modulation provides the same SNR performance as the baseband reference model, but
requires synchronization circuitry to perform coherent detection
• Non-suppressed-carrier AM simplifies the receiver design as it is implemented with an
envelope detector. However, it requires significantly more transmitter power to obtain the
same SNR performance as the baseband reference model
• SSB achieves the same SNR performance as the baseband reference model but only requires
half the transmission bandwidth of the DSC-SC system. On the other hand, SSB requires more
transmitter processing.
• Communication system design involves a tradeoff between power, bandwidth,
and processing complexity.
CE I Module IV Prof. (Dr.) Shahana T.K. 31