Comm Theory

i i
“FM” — 2018/12/17 — 16:26 — page v — #1

i i
Preface
World has come very close due to rapid advancements in communication

technologies. High speed computers and microprocessors, broadband com-
munication channels, satellites, fiber optic cables have made communication
very effective in wide range of applications. This book mainly describes basic
principles of Communication Theory.
First chapter presents basic aspects of communication. It also presents
generation of amplitude modulation and detection of amplitude modulation.
Hilbert transform is also given in detail.
Second chapter is based on phase and frequency modulation. It presents
Narrow Band and Wide band FM. This chapter also presents FM to AM
conversion.
Third chapter presents Random variables, Random Process, Stationary
Processes, Mean, Correlation, Covariance functions. It also presents Power
Spectral Density, Ergodic Processes, Gaussian Process, and Transmission of
a Random Process through a LTI filter.
Fourth chapter presents Narrow band noise and Noise performance anal-
ysis in AM & FM systems. It also includes pre-emphasis and de-emphasis for
FM.
Fifth chapter describes Sampling and Quantization. It includes: PAM,
PPM, PWM, PCM-TDM, FDM.
The overview, short questions and answers for all units including univer-
sity solved questions are discussed in an eloquent manner.
Suggestions are also welcome for the improvement of this book in the
future revisions.
Dr.M.P.Chitra
chi_mp2003@yahoo.co.in
i i
i i
i i
“FM” — 2018/12/17 — 16:26 — page vi — #2

i i
i i
i i
i i
“FM” — 2018/12/17 — 16:26 — page vii — #3

i i
Acknowledgement
First and Foremost, We thank the God almighty for making us to complete
the book successfully “Communication Theory”. We owe a debt of gratitude
to a number of people who have assisted us in writing this book. It would
have been an impossible task to write it and endure the challenging times,
without the love and support of our family, friends and colleagues. They have
all played a major role in bringing this book to a completion. We know that
we owe each one of them a huge debt of gratitude.
Firstly, We take this opportunity to express our profound gratitude and
sincere regards to our Founder and Chairman Col. Dr. Jeppiaar M.A., B.L.,
Ph.D.
We would like to express our deepest appreciation and sense of gratitude
to our Secretary and Correspondent Dr. P. Chinnadurai M.A., Ph.D., for his
invaluable assistance and support which helped us in completing this work.
We also express our special thanks to our dynamic directors Mr. C.
Sakthikumar M.E., Mrs. C. Vijayarajeswari. They have been always
inspiring, motivating and encouraging us.
We owe our sincere thanks to Principal Dr. T. Jayanthy M.E., Ph.D., for
lending her suggestions and courage to bring out this book.
Our most sincere thanks as always, to our parents for their strong
encouragement, unconditional love. Thank you for supporting us every step
of the way and for constantly being there for us.
Last but not least, our gratitude goes to Anne Publications, for the effort
in bringing the book in time.
Dr. M.P. Chitra

Mrs. D. Kalaiarasi
Mrs. S. Saranya
i i
i i
i i
“FM” — 2018/12/17 — 16:26 — page viii — #4

i i
i i
i i
i i
“FM” — 2018/12/17 — 16:26 — page ix — #5

i i
Contents
Preface v
Acknowledgement vii
Introduction 1
UNIT 1
Amplitude Modulation 5
1.1 Amplitude Modulation/Double Side Band-Full
Carrier (DSB-FC) 5
1.1.1 Definition 5
1.1.2 Mathematical Expression 5
1.1.3 Waveform of AM 6
1.1.4 Modulation Index (ma ) 7
1.1.5 Percentage of Modulation 7
1.1.6 Calculation of Modulation Index from an AM Wave 7
1.1.7 Degree of Modulation 8
1.1.8 Frequency Spectrum and Bandwidth of AM 9
1.1.9 Bandwidth of AM 11
1.1.10 Power Distribution of AM 11
1.1.11 Power Spectrum of AM 13
1.1.12 Current Distribution of AM 13
1.1.13 Transmission Efficiency (η) of AM 14
1.1.14 Generation of AM Waves [AM Modulators] 15
1.1.15 Detection of AM Wave/AM Demodulating
Circuits/AM Detector Circuits 25
i i
i i
i i
“FM” — 2018/12/17 — 16:26 — page x — #6

i i
x . Contents
1.2 Double Sideband-Suppressed Carrier System 30

1.2.1 Power Relation for DSB-SC Signal 32
1.2.2 Transmission Efficiency/Power Saving from
DSB-SC System 32
1.2.3 Generation of DSB-SC 33
1.2.4 Ring Modulate/Diode Bridge Modulator 35
1.2.5 Demodulation of DSB-SC/DSB-SC Detection 37
1.3 Hilbert Transform 39
1.3.1 Characteristics of Phase Shifting System 39
1.3.2 Properties of Hilbert Transform 40
1.3.3 Pre Envelope (or) Analytic Signal 41
1.3.4 Complex Envelope 41
1.4 Single-Side Band-Suppressed Carrier System 41
1.4.1 Power Distribution 42
1.4.2 Power Saving of SSB-SC with Respect
to AM Signal 42
1.4.3 Power Saving with DSB-SC 43
1.4.4 Time-Domain of SSB-SC 43
1.4.5 Single Side Band Suppressed Carrier (SSB-SC) 46
1.4.6 Demodulation of SSB-SC 49
1.5 Vestigial Sideband Suppressed Carrier (VSB-SC) 49
1.5.1 Frequency Spectra of (VSB-SC) 49
1.5.2 Generation of VSB-SC 50
1.5.3 Demodulation of VSB-SC (Coherent Detector) 51
1.6 AM Super Heterodyne Receiver 53
1.6.1 Heterodyning 53
UNIT 2
Angle Modulation 61
2.1 Introduction 61
2.1.1 Frequency Modulation (FM) 61
2.1.2 Phase Modulation (PM) 61
2.1.3 Time Domain Representation for
Phase Modulation 61
i i
i i
i i
“FM” — 2018/12/17 — 16:26 — page xi — #7

i i
Contents . xi
2.1.4 Time-Domain Representation of FM 62

2.1.5 Deviation Ratio (DR) 63
2.1.6 Percentage of Modulation of FM 63
2.2 Generation of FM from PM 64
2.3 Generation of PM from FM 64
2.4 Types of FM 65
2.4.1 Narrowband FM 65
2.4.2 Wideband FM 67
2.4.3 Generation of NBFM 68
2.4.4 Comparison of WBFM and NBFM 68
2.5 Generation of FM Waves 69
2.5.1 FET Reactance Modulator 69
2.5.2 Frequency Modulation Using Varactor Diode 70
2.5.3 Indirect Method of FM Generation
(Armstrong Method) 72
2.6 FM Demodulators/Detectors 73
2.6.1 Slope Detectors 73
2.6.2 Balanced Slope Detector 75
2.6.3 Foster-Seeley Discrimination
(Phase Discrimination) 77
2.6.4 Ratio Detector 79
2.6.5 Block Diagram of PLL 81
2.7 Frequency Spectrum Analysis of Single Tone
Sinusoidal FM 81
2.7.1 Introduction 81
2.7.2 Power Calculation of FM 84
2.8 Comparison of FM and PM 85
UNIT 3
Random Process 91
3.1 Random Process 91
3.1.1 Random Variables 91
3.1.2 Discrete Random Variable 91
3.1.3 Continuous Random Variable 93
i i
i i
i i
“FM” — 2018/12/17 — 16:26 — page xii — #8

i i
xii . Contents
3.2 Stationary and Non-stationary Process 95

3.2.1 Stationary Process 95
3.2.2 Wide Sence Stationary 95
3.3 Mean 99
3.4 Covariance 99
3.4.1 Properties of Co-variance 100
3.5 Correlation 100
3.6 Ergodic Process 100
3.7 Auto Correlation 101
3.7.1 Properties 101
3.8 Spectral Densities 102
3.8.1 Power Spectral Density (PSD) 102
3.8.2 Cross Power Spectral Density (CPSD) 102
3.8.3 Properties of Power Spectral Density 103
3.8.4 Energy Spectral Density (ESD) 103
3.9 Ergodic Process 104
3.10 Gaussian Process 104
3.10.1 Properties 104
3.11 Stationary Process 107
3.11.1 Types of Stationary Process 107
3.12 Transmission of Random Process Through LTI a Filter 108
3.12.1 Gaussian (or) Normal Distribution 109
UNIT 4
Noise Characterisation 111
4.1 Noise 111
4.1.1 Classification of Noise 111
4.1.2 Basic Definitions for Noise Characteristic 112
4.1.3 Noise in CW Modulation System 112
4.2 Noise Sources 113
4.2.1 External Noise 114
4.2.2 Internal Noise (Fluctuation Noise) 114
4.2.3 Thermal Noise 114
4.2.4 To Reduce Thermal Noise 118
i i
i i
i i
“FM” — 2018/12/17 — 16:26 — page xiii — #9

i i
Contents . xiii
4.2.5 Shot Noise 119

4.2.6 White Noise 121
4.3 Narrow Band Noise 122
4.3.1 Representation of Narrowband Noise in
Terms of Inphase and Quadrature
Components 122
4.3.2 Representation of N.B Noise Interms of
Envelope and Phase Components 124
4.4 Noise Factor (F) 126
4.4.1 Spot Noise Factor 127
4.5 Noise Factor of Amplifier in Cascade (FRIIS Formula) 128
4.5.1 Noise Temperature 130
4.6 Noise Equivalent Bandwidth (BN ) 131
4.7 Noise Performance in AM System 132
4.7.1 Noise in DSB-SC System Using Coherent
Detector 132
4.7.2 Noise in AM System Using Envelope Detection 135
4.7.3 Noise performance in SSB-SC Using Coherent
Detector 138
4.8 Noise in FM System 141
4.9 FM Threshold Effect 144
4.10 Pre Emphasis and DE Emphasis 145
4.10.1 Pre Emphasis 145
4.10.2 De-Emphasis Circuit 146
4.10.3 Pre-Emphasis and De-Emphasis in
FM System 147
UNIT 5
Sampling and Quantization 149
5.1 Functional Description of Digital Communication
System 149
5.1.1 Bandpass Modulator 150
5.1.2 Performance Measure 150
5.2 Digital Communication System-Functional Description 150
i i
i i
i i
“FM” — 2018/12/17 — 16:26 — page xiv — #10

i i
xiv . Contents
5.3 Sampling (Discretization in Time) 151

5.3.1 Sampling Theorem 151
5.3.2 Proof of Sampling Theorem 151
5.3.3 Quadrature Sampling of Bandpass Signal 156
5.3.4 Reconstruction of Bandpass Signal g(t) (or)
Reconstruction of Message from it’s Sample (or)
Proof of Reconstructions of Message for
its Samples with Zero Error 156
5.3.5 Types of Sampling 158
5.3.6 Limitations of Sampling 160
5.3.7 Calculation of Aliasing Error 161
5.4 Quantization 163
5.4.1 Quantization Error (or)
Quantization Noise Power 163
5.4.2 Types of Quantization 166
5.4.3 Companding 167
5.5 Pulse Modulation 169
5.5.1 Introduction 169
5.5.2 Pulse Amplitude Modulation (PAM) 171
5.5.3 Pulse Width Modulation (PWM/PDM) 174
5.5.4 Pulse Position Modulation 175
5.6 Frequency Division Multiplexing (FDM) System 179
5.6.1 FDM Transmitter 179
5.6.2 FDM Receiver 180
5.7 Pulse Code Modulation 181
5.7.1 Temporal Waveform Encoding 182
5.7.2 Pulse Code Modulation 182
5.7.3 PCM Bandwidth: (Transmission B.W BT ) 185
5.7.4 PCM Word Size 185
5.7.5 Noise in PCM System 186
5.7.6 SNR of a PCM System 186
5.8 Time Division Multiplexing (TDM)
[Digital Circuit Switch] 189
5.8.1 Operation 189
5.8.2 LPF 189
i i
i i
i i
“FM” — 2018/12/17 — 16:26 — page xv — #11

i i
Contents . xv
5.8.3 Pulse Modulator 189

5.8.4 Pulse Amplitude Demodulator 189
5.8.5 De Commutator 189
5.8.6 Types of TDM 191
5.8.7 Performance Measure of Digital
Communication System 191
University Questions 193
i i
i i
i i
“FM” — 2018/12/17 — 16:26 — page xvi — #12

i i
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 1 — #1

i i
Introduction
COMMUNICATION
It’s the process of conveying (or) transferring manages from one point to
another. Generally it can be classified into two types:
(i) Communication within line of sight.
(ii) Communication beyond the line of sight from point to point.
For example, line telephony and line telegraphy, radio telephony, and
line telegraphy, radio broadcasting, point to point communication and
mobile communication, computer communication, radar communication,
television telecasting, radio telemetry, radio aids to navigation, radio aids to
aircraft landing etc.
BASIC COMMUNICATION SYSTEM

Modulation Demodulation/detection
(encoding) (decoding)
Information
Transmitter Channel Receiver Destination
Source
{
Transmitter Receiver
section Noise
section
{
Communication
Medium
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 2 — #2

i i
2 . Introduction
MODULATION
Modulation is the process of changing the characteristics of high frequency

carrier signal in accordance with the instantaneous value of a modulating
value of a modulating signal (or) manage signal.
The three characteristics of the carrier signal are: (i) Amplitude (ii)
Frequency and (iii) Phase. The result of modulation process produces the
modulated signal (or) passband signal.
CLASSIFICATION OF MODULATION PROCESS

Modulation
Analog Modulation Digital Modulation
Continuous Wave Pulse Modulation

Modulation (carrier in pulse)
(carrier in continuous)
PAM PWM PPM

AM FM PM
Digital pulse Keying

Modulation techniques
PCM DM DPCM ASK FSK PSK
NEED FOR MODULATION/ADVANTAGES OF

MODULATION
Modulation is extremely necessary in communication system due to the

following reason:
(i) Reduction in height of antenna.
(ii) Avoids mixing of signals.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 3 — #3

i i
COMMUNICATION THEORY . 3
(iii) Multiplexing is possible.

(iv) To overcome equipment limitation.
(v) Improved quality.
CLASSIFICATION OF COMMUNICATION
(i) Line (or) Wired Communication:

The medium of transmission is pair of conductors called Txion line.
Transmitter and receiver are connected through a wire (or) lines.
Drawbacks: Installation and maintenance of txion line is costly and
it over crowch open space.
(ii) Wireless (or) Radio Communication:
Manage is ted through open space by EM waves called as radio waves.
Advantages:
• Cost effectiveness
• Possible long distance communication
• Simplicity
FREQUENCY RANGE IN COMMUNICATION SYSTEMS
Types of Signal Frequency Range Application

VLF 3 to 30KHz Long distance, Point to point
Communication
LF 30 to 300KHz Radio Navigation
MF 300 to 3MHz Broadcasting, Maine application
HF 3 to 30MHz Radio telephony
VHF 30 to 300MHz FM broadcasting, TV,
Mobile, radio, radio navigation
UHF 300 to 3000MHz FM broadcasting, TV,
Mobile, radio, radio navigation
EHF 3 to 30GHz Multichannel telephony links
(microwave) Radar, Satellite communication
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 4 — #4

i i
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 5 — #5

i i
Unit
Amplitude Modulation 1
1.1 Amplitude Modulation/Double Side
Band-Full Carrier (DSB-FC)
1.1.1 Definition
Amplitude modulation may be defined as the process by which the ampli-
tude of the high frequency carrier signal varies in accordance with the
instantaneous amplitude of the modulating (or) message signal.
1.1.2 Mathematical Expression

Let em (t) be the modulating signal,
em (t) = Em sin 2π fm t (1.1)
ec (t) be the carrier signal,
ec (t) = Ec sin 2π fc t (1.2)
where
Em → maximum amplitude of the modulating signal.
Ec → maximum amplitude of the carrier signal.
fm → frequency of the modulating signal.
fc → frequency of the carrier signal.
According to the definition, the amplitude of the carrier signal is changed
after modulation. Let EAM be the amplitude of the modulated signal,
EAM = Ec + em (t)
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 6 — #6

i i
6 . Unit 1: Amplitude Modulation
EAM = Ec + Em sin 2π fm t (1.3)

Em
= Ec 1 + sin 2π fm t
Ec
EAM = Ec [1 + ma sin 2π fm t] (1.4)
where ma = EEmc is modulation index (or) depth of modulate. Therefore the

instantaneous value of the amplitude modulated wave can be written as,
eAM (t) = EAM sin 2π fc t (1.5)
Substitute eqn. (1.4) in (1.5),
eAM (t) = Ec [1 + ma sin 2π fm t] · sin 2π fc t

eAM (t) = Ec sin 2π fc t + Ec ma sin 2π fm t · sin 2π fc t
1.1.3 Waveform of AM
Figure 1.1: AM waveform.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 7 — #7

i i
AM wave has a time-varying amplitude called as the envelope of the AM

wave. The unique property of AM wave is that the envelope of the modulated
carrier has the same shape as the message signal.
1.1.4 Modulation Index (ma )

The ratio of maximum amplitude of the modulating signal to the maximum
amplitude of the carrier signal is called modulation index (or) depth of
modulation (or) co-efficient of modulation. It’s represented by ma .
Em
ma =
Ec
1.1.5 Percentage of Modulation

When the modulation index is expressed in percentage, it is called as the
percentage of modulation and is denoted by ‘M’.
Em
M= × 100
Ec
Modulation index is also known as depth of modulation, degree of mod-
ulation (or) modulation factor. The higher the percentage of modulation,
the greater the side band power and the stronger and more intelligible the
transmitted and received signal.
1.1.6 Calculation of Modulation Index from an AM Wave

From the figure
2Em = Emax − Emin

1
Em = [Emax − Emin ] (1.6)
2
Ec = Emax − Em (1.7)
Substituting the value of Em in above eqn.

Emax − Emin
Ec = Emax −
2
1
Ec = [Emax + Emin ] (1.8)
2
We know that,
Em
ma =
Ec
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 8 — #8

i i
Em
Ec Em Emax
Emin
t
⫺
Figure 1.2: Graphical representation of AM signal.
Substitute eqn. (1.8) and (1.9) in above expression,

Emax − Emin
ma = (1.9)
Emax + Emin
Equation (1.9) gives the modulation index interms of maximum and
minimum amplitude of AM wave.
1.1.7 Degree of Modulation

(i) Critical Modulation: [ma = 1, Em = Ec ]
When Em = Ec modulation goes to 100% this situation is known
as critical modulation. The envelope of the modulated signal just
reaches the zero amplitude axis. The manage signal remain reserved
as shown below.
Figure 1.3: Critical modulation.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 9 — #9

i i
(ii) Under Modulation: [ma < 1, Em < Ec ]

The envelope of the AM signal doesn’t reach the zero amplitude axis.
Therefore the manage signal is fully preserved in the AM envelope.
Figure 1.4: Under modulation.
(iii) Over Modulation: [m > 1, Em > Ec ]

The amplitude of the modulating signal is greater than the amplitude
of the carrier signal. Therefore portion of the envelope of the mod-
ulated signal crosses the zero amplitude axis, due to this envelope
distortion occurs as shown below.
Figure 1.5: Over modulation.
1.1.8 Frequency Spectrum and Bandwidth of AM

The experiment for an AM wave is given by,
eAM (t) = [Ec + Em 2π fm t] sin 2π fc t (1.10)
We know that,
Em
ma = (1.11)
Ec
Em = ma Ec (1.12)
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 10 — #10

i i
Substitute eqn. (1.12) in (1.10)
eAM (t) = [Ec + ma Ec sin 2π fm t] sin 2π fc t

= Ec [1 + ma sin 2π fm t] sin 2π fc t
eAM (t) = Ec sin 2π fc t + ma Ec sin 2π fc t sin 2π fm t (1.13)
Expanding the eqn. (1.13) using the trigonometrical relation,

1
sin A sin B = [cos(A − B) − cos(A + B)]
2
We get,
Ec ma
eAM (t) = Ec sin 2π fc t + [cos 2π(fc − fm )t − cos 2π(fc + fm )t]
2
ma Ec ma Ec
eAM (t) = Ec sin 2π fc t + cos 2π(fc − fm )t − cos 2π(fc + fm )
2 2
(1.14)
The above eqn. (1.14), contain three frequency components as follows:

1. Unmodulated carrier signal.
2. Lower side band (fc − fm ) having amplitude ma Ec
2
.
3. Upper side band (fc + fm ) having amplitude ma Ec
2
.
Amp (V)
Ec
Emma Ecma
2 2
Em
f (MHz)
fm fc⫺fm fc fc⫹fm
Figure 1.6: Frequency spectrum of AM waves.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 11 — #11

i i
Eqn. (1.14), can be also be written with the corresponding frequency term,
i.e., fLSB = fc − fm and fUSB = fc + fm
mEc mEc
eAM (t) = Ec sin 2π fc t + cos 2π fLSB t − t
cos 2π fUSB (1.15)
2 2
1.1.9 Bandwidth of AM
The bandwidth of AM is equal to the difference between the highest upper
side frequency and lowest lower side frequency. This is the frequency range
one which the information signal is transmitted.
Bandwidth = fUSB − fLSB
fLSB = fc − fm and fUSB = fc + fm
Bandwidth = (fc + fm ) − (fc − fm )

= fc + fm − fc + fm
Bandwidth = 2fm Hz
1.1.10 Power Distribution of AM

It has been shown that the carrier component of the modulated wave has
the same amplitude as the unmodulated carrier. That is, the amplitude of the
carrier is unchanged, energy is either added (or) subtracted.
The modulated wave contain extra energy in the two sideband compo-
nents. Therefore, the modulated wave contains more power than the carrier
had before modulation took place.
Since the amplitude of the sidebands depends on the modulation index
( EEmc ), it is anticipated that the total power in the modulation wave will
depend on the modulation index also.
The total power in the modulated wave will be,
Pt (AM) = Pc + PUSB + PLSB (1.16)
To Find Carrier Power (Pc ):

The average power of the unmodulated carrier is equal to the rms carrier
voltage squared divided by the load resistance (generally antenna resistance).
Mathematically, power in an unmodulated carrier is,
2
Ec
√
2
Pc =
R
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 12 — #12

i i
Ec2
Pc = (1.17)
2R
where
Pc → Carrier power in watts.
Ec → Peak carrier voltage in volts.
R → Load resistance in ohm.
AM wave equation is given as,
ma Ec ma Ec
eAM (t) = Ec sin 2π fc t + cos(fc − fm )2π f − cos(fc + fm )2π t
2 2
From the above expression, the peak amplitude of both sidebands is ma2Ec .
The power generated by the upper side frequency component will be equal
to the lower side frequency component as they have some amplitude. Hence
the upper and lower sideband powers are mathematically expressed as,
2
ma Ec
√2
2
PLSB = PUSB =
R
m2a Ec2
PLSB = PUSB =
8R
PLSB → Lower side band power (watts).
PUSB → Upper side band power (watts).
Total power of the AM wave
Pt (AM) = PC + PLSB + PUSB

Ec2 ma Ec2 ma Ec2
= + +
2R 8R 8R
2
Ec ma m2a
2
Pt (AM) = 1+ +
2R 4 4

E2 m2
= c 1+ a
2R 2

m2a
Pt (AM) = Pc 1 +
2
PE
Pc = 2
1 + m2a

m2a
Pc 1 + = Pt
2
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 13 — #13

i i
Pt m2
= 1+ a
Pc 2
Pt m2
−1 = a
Pc 2

Pt − 1
m2a = 2
Pc

Pt − 1
ma = 2
Pc
The above expression represent the modulation index of AM interms of

power.
1.1.11 Power Spectrum of AM
Pt (AM) = Pc + PLSB + PUSB

Ec2 Ec2 m2a Ec2 m2a
= + +
2R 8R 8R
1.1.12 Current Distribution of AM

In general the power and current in the antenna are related by P = I 2 R. We
know that,

m2a
Pt (AM) = Pc 1 +
2
Power
Ec2
2R
Ec2ma2
Ec2ma2
SR
SR
freq (Hz)
fc⫺fm fc fc⫹fm
LSB USB
Figure 1.7: Power spectrum of AM.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 14 — #14

i i

m2
It2 R = Ic2 R 1 + a
2
2
m
It2 = Ic2 1 + a
2

m2
It = Ic 1 + a
2
To Find Modulation Index (ma ):

It m2a
= 1+
Ic 2
It2 m2
2
= 1+ a
Ic 2
m2a It2
= 2 −1
2 Ic
2
It
ma = 2 2 − 1
2
Ic

I2
ma = 2 t2 − 1
Ic
Modulation index (ma ) interms of current.
1.1.13 Transmission Efficiency (η) of AM

It can be defined as the ratio of power in sidebands to total power because
side bands only contain the useful information.
Power in side bands

η= × 100
total power
PLSB + PUSB
= × 100
Pt (AM)
m 2a E 2c m 2a E 2c
+ 8R
= 8R
× 100
E 2c 2
2R
[1 + m2a ]
m 2a E 2c
= 4R
E 2c 2
2R
[1 + m2a ]
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 15 — #15

i i
E 2c m 2a
·
η= 2R 2
E 2c 2
2R
[1 + m2a ]
m2a
=
2 + m2a
if ma = 100% and ma = 1, therefore,
1
%η = × 100
2+1
1
= × 100
3
%η = 33.33%
1.1.14 Generation of AM Waves [AM Modulators]

Generation of AM waves
Non-linear Linear modulator

modulator (large signal modulator or
(small signal modulator or high level modulator)
low level modulator)
Square law Product Balanced

modulator modulator modulator
Transistor Switching
modulator modulator
Collector Base Emittee
Generation of AM Wave Using Non-Linear Modulators

(i) Square law modulator:
Any device operated in non-linear region of its output characteristics
is capable of producing amplitude modulated waves.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 16 — #16

i i
Figure 1.8: Square law modulator.
The square law circuit consists of,

(i) Non-linear device.
(ii) Band Pass Filter (BPF).
(iii) Message be carrier source.
The modulating signal and carrier signal are connected in series with
each other, their sum V1 (t) is applied at the input of the non-linear
device as shown in above figure.
V1 (t) = Em sin 2π fm t + Ec sin 2π fc t
The input/output relationship for any non-linear device is given by,
V2 (t) = aV1 (t) + bV12 (t)
where a and b are constants.
V2 (t) = a[Em sin 2π fm t + Ec sin 2π fc t]

+b[Em sin 2π fm t + Ec sin 2π fc t]
= a[Em sin 2π fm t + Ec sin 2π fc t] + b[Em
2
sin2 2π fm t
+Ec2 sin2 2π fc t + 2Em Ec sin 2π fm t sin 2π fc t]
= aEm sin 2π fm t + aEc sin 2π fc t + bEm
2
sin2 2π fm t
+bEc2 sin2 2π fc t + 2bEm Ec sin 2π fm t sin 2π fc t]
V2 (t) = aEm sin 2π fm t + aEc sin 2π fc t + bEm
2
sin2 2π fm t
+bEc2 sin2 2π fc t + bEm Ec [cos 2π(fc − fm )t
− cos 2π(fc + fm )t]
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 17 — #17

i i
fifth term is AM wave with only sidebands. This is passed through

the band pass filter which is tuned to allow only fc , (fc − fm ) and
(fc + fm ). Therefore other than these tuned frequencies, the rest of
the frequencies appearing in the frequency spectrum are filtered.
V0 (t) = aEc sin 2π fc t + bEm Ec [cos 2π(fc − fm )t − cos 2π(fc + fm )]

V0 (t) = aEc sin 2π fc t + bEm Ec cos 2π(fc − fm )t
−bEm Ec cos 2π(fc + fm )
(ii) Square law modulator using transistor:

The circuit is connected in common emitter configuration as shown
in below figure. The transistor is used as a switching transistor which
operates in the non-linear characteristics. It is biased in such a way
that it operates in the class A mode and efficiency is low. This circuit
is used in the low level modulation. The modulation signal is applied
to the emitter and RF carrier at the base of the transistor.
Vcc
T2
C2 V0(t)
R1
Q1
T1
RF
Carrier C1 R2
R3 C3
Modulating
signal
Figure 1.9: Square law modulator using transistor.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 18 — #18

i i
(iii) Balanced modulator:

Here two non-linear transistors are connected in the balance mode.
It is assumed that the two transistors are identical and the circuit is
symmetrical. The operation of the transistor used are confined to
operate in the non-linear region of its transfer characteristics. The
carrier voltage across the box winding of the center tap transformer
as shown in figure are equal and opposite in phase that is
ec (t) = ec (t)
where
ec (t) = Ec sin ωc t (1.18)
Let the manage signal be,
em (t) = Em sin ωm t (1.19)
Input voltage across the transistor θ1 is given by,
Vbe = ec (t) + em (t)

Vbe = Ec sin ωc t + Em sin ωm t (1.20)
ic
T1
Q1
ec(t)
T2
Vbe
Receiver
signal AM output
ecsin ␻0t V0(t)
T3
Manage V⬘be
signal
e⬘c(t)
Q2
i⬘c
Figure 1.10: Balanced modulator.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 19 — #19

i i
Input voltage across the transistor θ2 is given by,

Vbe = ec (t) + em (t)

Vbe = −Ec sin ωc t + Em sin ωm t (1.21)
By the non-linearity relationship of the transistor the collector

current can be written as,
ic = aVbc + bVbc
2
(1.22)
ic = aVbc
2
+ bVbc (1.23)
Substitute eqn. (1.20) in (1.22), we get
ic = a[Ec sin ωc t + Em sin ωm t] + b[Ec sin ωc t + Em sin ωm t]

ic = aEc sin ωc t + aEm sin ωm t + bEc2 sin2 ωc t + bEm
2
sin2 ωm t
+2bEm Ec sin ωc t sin ωm t (1.24)
ic = a[−Ec sin ωc t + Em sin ωm t] + b[Em sin ωm t − Ec sin ωc t]2

= −aEc sin ωc t + aEm sin ωm t + bEm
2
sin2 ωm t + bEc2 sin2 ωc t
−2bEm Ec sin ωc t sin ωm t (1.25)
The output AM voltage (V0 ) is given by,
V0 = k(ic − ic ) (1.26)
This is because both the current flows in the opposite direction k is a

constant depending on impedance and other circuit parameters.
Substitute eqn. (1.24) and (1.25) in eqn. (1.26)
V0 = k[(aEc sin ωc t + aEm sin ωm t + bEc2 sin2 ωc t + bEm

2
sin2 ωm t
+2bEm Ec sin em t sin ωc t) − (−aEc sin ωc t + aEm sin ωm t
+bEm
2
sin2 ωm t + bEc2 sin2 ωc t − 2bEm Ec sin ωc t sin ωm t)]
V0 = k[aEc sin ωc t + aEm sin ωm t + bEc2 sin2 ωc t + bEm
2
sin2 ωm t
+2bEm Ec sin ωm t sin ωc t + aEc sin ωc t − aEm sin ωm t
−bEm
2
sin2 ωm t − bEc2 sin2 ωc t + 2bEm Ec sin ωm t sin ωc t]
V0 = k[2aEc sin ωc t + 4bEm Ec sin ωm t sin ωc t]
V0 = 2kaEc sin ωc t + 4kbEm Ec sin ωm t sin ωc t
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 20 — #20

i i

2bEm
V0 = 2ka Ec 1 + sin ωm t sin ωc t
a
2bEm
ma = is the modulation index of AM
a
∴ V0 = 2ka Ec [1 + ma sin ωm t] sin ωc t
The advantages of balanced modulator for generating AM is it is

simple and undesired non-linear term are eliminated automatically
without any filters.
(iv) Square law modulator using diode:
Manage signal (AF) and carrier signal (RF) applied at the input are
super imposed each other and makes the diode more forward biased
Diode D
⫹ ⫺
Emsin ␻0t ~
V R C V0(t)
Ecsin ␻ct ~
Figure 1.11: Square law modulator using diode.
Diode
current
(mA)
Operating
point
Diode voltage (volts)

Figure 1.12: Waveform.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 21 — #21

i i
during the (+ve) half cycle of input and less forward biased during
(−ve) half cycle of manage signal.
Thus the magnitude of the carrier component is greater during
the (+ve) half cycle of the modulating voltage and linear during the
(−ve) half cycle of the modulating signal.
Linear Modulator: Switching Modulator: Diode ‘D’ is acts as a switch

driven at the carrier frequency. RLC circuit is turned to resonate at fc , so
that the switching action causes the tank circuit to ring sinusidally.
Case (i) [During absence of message signal]
V0 (t) = ec (t) = Ec sin ωc t
Case (ii) [During presence of message signal]

Adding manage to the input through the transformer gives an
output of,
V0 (t) = [Ec sin ωc t + NEm sin ωm t · sin ωc t]

NEm
V0 (t) = Ec sin ωc t + [cos(ωc − ωm )t − cos(ωc + ωm )t]
2
where N is turns ratio of transformer.
V0 (t) = eAM (t)
Emsin ␻mt
em(t)
T2 D
1:N
Ecsin ␻ct V0(t)

ec(t) R L C
T1
Figure 1.13: Switching modulator.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 22 — #22

i i
Desired modulation can be obtained by adjusting the values of the input

voltage and turns ratio.
(i) Low-level AM Modulator: [Emitter Modulator]

A simple class ‘A’ amplifier shown in below figure is for the generation
of AM signal. In the absence of modulating signal, the circuit simply
operates as linear class ‘A’ amplifier. When the modulating signal is
applied to an emitter, the gain of the amplifier varies according to the
voltage of modulating signal. Gain of the transistor stage is depends
upon the emitter current at quiescent point.
ie = Ic + Em sin ωm t (1.27)
Vcc
T2
Rc
R1
C2
C1 Modulated
R output (V0)
Q1
eAM(t)
Carrier signal
ec(t)
RE C3
R2
Modulated
em(t)
Figure 1.14: Low-level modulator.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 23 — #23

i i
Voltage gain of the transistor is proportional to the emitter current.
Av α ie
Av = kie (k is constant) (1.28)
V0
Av = (1.29)
Vi
Av Vi = V0 (1.30)
Vi = Ec sin ωc t (1.31)
Av = k(Ie + Em sin ωm t) (1.32)
Substitute eqns (1.31) and (1.32) in (1.30),
V0 = k(Ic + Em sin ωm t)Ec sin ωc t

= kIe Ec sin ωc t + kEm Ec sin ωm t sin ωc t
kEm Ec
V0 = kIe Ec sin ωc t + [cos(ωc − ωm )t − cos(ωc + ωm )t]
2
Thus AM is generated.
Output waveform
amp (volts)
em(t)
Vc1
t
eAM(t)
Drawback: Low efficiency, since class ‘A’ amplifier is used.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 24 — #24

i i
(ii) High Level AM Modulator: [Collector Modulator]

The basic circuit for BJT collector modulator is shown below. The
RF drive is a carrier signal used for AM. This carrier amplitude is
such that it drives transistor in conduction over part of its cycle. The
modulating signal is passed through the power amplifier and applied
to the collector through a low frequency transformer. This modulating
voltage is in series with the supply voltage.
Hence collector voltage becomes Vcc = Vcc + Vm (t). The tuned
LC circuit associated tuned transformer on the collector receives the
AM signal. Because of modulating voltage, the net supply voltage of
transistor changes according to slow variation in Vm (t). Hence the RF
carries signal amplitude is also varied according to variation in Vm (t).
Thus AM signal is produced across LC circuit.
Case (i) Without message signal:
Circuit will acts as a class ‘c’ tuned amplifier.
V0 (t) = Vcc Ec (t)

V0 (t) = Vcc Ec sin ωc t
Vcc
Modulating Power
signal amplifier
RF bypass
C1 capacitor
C2 Modulated
RF output
RF drive
Q1
~
Bias
Figure 1.15: Circuit diagram of collector modulator.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 25 — #25

i i
Case (ii) With message signal:

Supply voltage will change with respect to manage signal.
Vcc = Vcc + Vm sin ωm t

V0 (t) = Vcc Vc sin ωc t
= (Vcc + Vm sin ωm t)Vc sin ωc t
V0 (t) = Vcc Vc sin ωc t + Vm Vc sin ωc t sin ωm t
Vm Vc
V0 (t) = Vcc Vc sin ωc t +
2
×[cos(ωc − ωm )t − cos(ωc + ωm )t]
Thus the AM wave is generated.

Advantages:
(i) Efficiency is high because class ‘c’ amplifier.
(ii) High output power.
1.1.15 Detection of AM Wave/AM Demodulating Circuits/

AM Detector Circuits
Demodulation/Detection: It is the process in which the modulating voltage
is recovered back from the modulated signal. This is the reverse process of
modulation taking place in the receiver.
AM detectors
Non-linear detectors Linear detectors
Synchronous Envelope Square-law Envelope

(or) (or) detector detector
Coherent Non-coherent
detector detector
Square-Law Detector If utilizes the non-linear region of voltage current

dynamic characteristics of a diode. This dynamic characteristics is high
non-linear, particularly in the low voltage region.
In square law modulator the filter is a bandpass filter, whereas in a square
law detector, a low pass filter is used. In the circuit the DC supply voltage
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 26 — #26

i i
Diode D
⫹ ⫺
⫹ Detected
AM
signal C output
⫺ R (modulating
signal)
VA
⫹ ⫺
Figure 1.16: Square-law detector.
VA is used to get a fixed operating point in the non-linear portion of the

V-I characteristics. Since the operation is limited to the non-linear region of
the diode characteristics, the lower half portion of the modulated waveform
is compressed. This produces envelope applied distortion. Due to this, the
average value of the diode-current is no-longer constant, rather it varies with
time.
Non linear V-I relationship between input and output,
V2 (t) = aV1 (t) + bV12 (t) (1.33)
V1 (t) is input of diode (AM wave) and V2 (t) is output of diode
V1 (t) = Ec [1 + ma sin ωm t] sin ωc t (1.34)
V2 (t) = a[Ec (1 + ma sin ωm t)] sin ωc t + b(Ec [1 + ma sin ωm t] sin ωc t)2

= aEc sin ωc t + ama Ec sin ωm t sin ωc t + b(Ec [1 + ma sin ωm t])2 sin2 ωc t
= aEc sin ωc t + ama Ec sin ωm t + [bEc2 + Ec2 b2 ma sin ωm t
(1 − cos 4π fc )
+bEc2 m2a sin2 ωm t]
2
Out of these terms only desired term is extracted by using low pass filter.
V0 (t) = bEc2 ma sin ωm t
Thus message signal recovered the output of detector.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 27 — #27

i i
Envelope Detector/Diode Detector (Linear Detector) It is the fact that a

diode operating in linear region of its V-I characteristics can extract the
envelope of an AM wave. This type of detector is known as envelope detector
(or) a linear detector.
A simple half-wave rectifier circuit using a diode operating in linear
region and the RC filter combination as shown in above figure forms the
envelope detector.
When the capacitor is absent the output is obtained across the load
resistor R. When the capacitor is connected in parallel to the resistor, the
capacitor charges to the peak value of the carrier signal for the positive half
cycle of the AM signal.
However, for a −ve half-cycle, the diode is reverse biased and doesn’t
conduct. This means that the input carrier voltage is disconnected from the
RC circuit. Therefore, the capacitor starts discharging through the resistance
‘R’ with a time constant t = RC.
If the time constant t = RC is suitably chosen, the voltage across the
capacitor ‘C’ will not fall appreciably during the small period of negative
half cycle, and by that time the next (+ve) cycle appear. The (+ve) cycle
again charges the capacitor ‘C’ to the peak value of the carrier voltage and
then their process repeats again and again.
The above figure shows the resulting detected modulating/base band
signal. We can reduce the spikes to a neglible amount by keeping the time
constant RC large so that the capacitor ‘C’ discharges negligible amount
Diode D
⫹ ⫺
T1
AM C1
signal
R V0(t)
C2
Figure 1.17: Envelope detector.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 28 — #28

i i
by keeping the time constant RC large so that the capacitor ‘C’ discharges
negligibly small amount.
Envelope of
modulating signal
i i
V V
Amplitude modulated
output (AM signal)
Figure 1.18: Characteristics of linear diode detection.
V
However, the large value of t = RC produces another problem known
as diagonal clipping. Thus we can not increase the time-constant beyond a
certain limit. Therefore, the time constant is an important consideration for
envelope detector.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 29 — #29

i i
Distortion in the Envelope Detector
(1) Diagonal Clipping:
• This type of distortion occurs when the RC time constant of load

circuit is too large.
• If RC time constant is too large, it can not follow the fast charges in
the modulating envelope.
• As a result of distortion in diagonal of detected signal is clipped out.
Amp (V)
Actual output
showing the diagonal
clipping
Output without
diagonal clipping
(2) Negative-Peak Clipping:

Modulation index of detected manage is greater than modulation
index of input AM.
Input AM (ma ) < Detected manage (ma )
Amp (V) Amp (V)
Tx Modulated
wave
Rx Modulated
wave
t t
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 30 — #30

i i
If modulation index (ma ) of manage is high
Em > Ec ma > 1
that is overmodulation. Hence negative peak clipping will take place

as a result of over modulation.
Selection of Time Constant:

To avoid distortion part in envelope detector, RC time constant should
be as high as 1/fc and as low as 1/fm .
1 1
<< RC <<
fm fc
Disadvantages of AM (DSB-FC):
• Power wastage takes place in DSBFC
• Bandwidth inefficient
• AM wave gets affected due to noise
Application of AM:
• Radio broad casting
• Picture transmission in TV system
1.2 Double Sideband-Suppressed Carrier System
The important parameters of a communication system are transmitting

power and the bandwidth. Hence saving of power and bandwidth are highly
desirable in a communication system.
Courier Antenna
signal
ec(t)
Baseband/
modulating Product
signal modulator
em(t) Modulated
signal (or)
DSB-SC signal
eDSB-SC(t)
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 31 — #31

i i
In AM with carrier scheme, there is wastage in both transmitted power

and the bandwidth. In order to save the power in AM, the carrier is
suppressed because it doesn’t contain any useful information.
This scheme is called as the double sideband suppressed carrier amplitude
modulation (DSB-SC-AM).
It contain only LSB and USB term, resulting in a transmission bandwidth
which is twice the frequency of the manage signal.
Let
• Modulating signal, em (t) = Em sin ωm t
• Carrier signal, ec (t) = Ec sin ωc t
When multiplying both the carrier and manage signal, the resulting signal
in the DSB-SC-AM signal.
eDSB-SC-AM (t) = em (t) · ec (t)
eDSB-SC (t) = Em sin ωm tEc sin ωc t
= Em Ec sin ωm t sin ωc t
Em Ec
= [cos(ωc − ωm )t − cos(ωc + ωm )t]
2
The derived expression shows only two side bands and the unmodulated
carrier signal is removed. The spectrum of DSB-SC signal is shown in the
following figure, which contains only the two side bands.
eDSB-SC(t)
EmEc LSB USB

2
2fm
Frequency
␻c⫺␻m ␻c⫹␻m
Figure 1.19: Spectrum of DSB-SC signal.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 32 — #32

i i
1.2.1 Power Relation for DSB-SC Signal

Total power of DSB-SC
Pt = PUSB + PLSB
m2a Ec2
PUSB − PLSB =
8R
m2 E2 m2 E2
Pt = a c + a c
8R 8R
2 2
m E
Pt = a c
4R
if R = 1
m2a Ec2
Pt =
4
E m2
2
= c · a
2 2
E2
Pc = c
2
Hence,
m2a
Pt = · Pc
2
For 100% modulation ma = 1
Pc
Pt(PSB−SC) =
2
1.2.2 Transmission Efficiency/Power Saving from DSB-SC System
PAM − PDSB-SC
Power Saving =
PAM

m2a
PAM = Pc 1 +
2
m2
PDSB-SC = Pc · a
2

m2 m2
Pc 1 + 2a − Pc · 2a
Power Saving = 2

Pc 1 + m2a
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 33 — #33

i i

m2 Pc m2 a m2a
Power Saving = Pc + Pc a − Pc 1 +
2 2 2
Pc
= 2

Pc 1 + m2a
2
=
2 + m2a
For 100% modulation, ma = 1,
2
Power Saving =
2+1
2
= = 0.66%
3
% of power saving = 66.7%
1.2.3 Generation of DSB-SC

• Balanced modulator
• Ring modulator
Balanced Modulator Using FET: When the modulating signal is not

applied, the FET current due to the carrier are equal but opposite in direction
so they cancel each other. Hence there is no output at T3 , then suppressing
the carrier.
When the modulating signal is applied, current id and id flow in the
primary of T3 due to carrier signal and modulating signal.
Since the modulating signal is applied 180◦ out of phase at the gates,
they are equal in magnitude and in direction, so they will not cancel each
other since the modulating signal windings are equal and opposite across
the windings of T1 we can write,
em (t) = em (t)
The voltage across the FFT greater can be written as,
Vgs = ec (t) + em (t)

= Ec sin ωc t + Em sin ωm t (1.35)
Vgs = Ec sin ωc t − Em sin ωm t (1.36)
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 34 — #34

i i
id
Q1
T1 em(t)
T3
V⬘gs
Modulating C1
signal AM output
em(t) V0(t)
Carrier
signal
ec(t) V⬘gs
e⬘m(t)
Q2
i⬘d
Figure 1.20: Balanced modulator using FET.
The drain currents of FET can be written as,
id = a1 vgs + a2 vgs
2
(1.37)
id = a1 vgs + a2 vgs 2 (1.38)
Now the output of DSB-SC signal V0 (t) is,
V0 (t) = k(id − id ) (1.39)
Substitute eqn. (1.37) and (1.38) in eqn. (1.39), we get
V0 (t) = k[(a1 vgs + a2 vgs

2
) − (a1 vgs + a2 vgs 2 )]
= k[a1 (vgs − vgs ) + a2 (vgs
2
− vgs 2 )]
= k[a1 (Ec sin ωc t + Em sin ωm t − Ec sin ωc t + Em sin ωm t)]
+a2 [(Ec sin ωc t + Em sin ωm t)2 − (Ec sin ωc t − Em sin ωm t)2 ]
= k[2a1 Em sin ωm t + a2 (4Em Ec sin ωm t sin ωc t)]

2a2 Ec sin ωc t
= 2ka1 Em sin ωm t 1 + · · sin ωm t
a1 sin ωm t
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 35 — #35

i i
therefore,

2a2 Ec
V0 (t) = 2ka1 Em sin ωm t 1 + sin ωc t (1.40)
a1
From the V0 (t) we can say that the carrier is suppressed and we have only
the two sidebands at output of balanced modulator.
This is output, V0 (t) is applied to the bandpass filter whose centre
frequency is ωc . Then we will get the output as two side band signals.
1.2.4 Ring Modulate/Diode Bridge Modulator

A ring modulator using for diodes are shown in figure In a ring modulator
circuit four diodes are connected in the form of a ring in which all four diodes
point in the same manner.
All the four diodes in the ring are contributed by a square have carrier
signal applied through a centre-tapped transformer.
Circuit Operation: The circuit operation can be explained for two different
conditions,
(1) With only the carrier signal applied to the circuit without the menage
signal.
(2) With a sinusoided menage signal applied to the circuit along with the
carrier signal.
Figure 1.21: Ring modulator.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 36 — #36

i i
(1) With only the carrier signal applied to the circuit without the
message signal.
(i) For (+ve) half-cycle of the carrier signal

Diodes D1 and D2 are forward biased and they conduct produces
the output current. The output currents from the diodes D1 and
D2 divides equally in the upper and lower portion of primary
winding of T2 . This produces a magnetic field which is equal in
magnitude but opposite in direction hence they cancel each other
producing no output at the secondary of T2 . Thus the carrier is
suppressed.
(ii) For −ve half cycle of the carrier signal
Diodes D1 and D2 are reverse biased and the diodes D3 and D4 are
forward biased. Hence the magnetic fields in primary winding of
T2 are equal in magnitude and opposite in direction cancelling
each other thus the carrier is suppressed.
(2) With a sinusoidal message signal is applied to the circuit along with
the carrier signal.
amp (V)
Modulating
signal
O
Carrier
signal
O
D1 & D2 D3 and D4
conduct DSB-SC
conduct signal
O
D1 and D2
D3 and D4 conduct
conduct
Figure 1.22: Waveform.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 37 — #37

i i
When the modulating signal is applied to the primary of T1 , it will

appear across the secondary of T1 .
(i) For (+ve) half-cycle of the modulating signal

Diodes D1 and D2 are forward biased and they will connect the
secondary of T1 to the primary of T2 . As a result of the modulating
signal at the secondary of T1 is applied to primary of T2 through
diodes D1 and D2 .
(ii) During negative half cycle
Diodes D3 and D4 are forward biased and they will connect the
secondary of T1 to the primary of T2 with revere connection.
1.2.5 Demodulation of DSB-SC/DSB-SC Detection

(1) Synchronous detector/coherent detector
(2) Costas PLL/Costas loop
(1) Synchronous detector/coherent detector:

This type of detector uses locally generated carrier for detection. This
technique uses the carrier which is in phase coherence with that of
transmitter.
Ec ma
S(t) = sin ωm t sin ωc t
2
y(t) = s(t) · ec (t)
Ema
= sin ωm sin ωc tEc sin(ωt + φ)
2
Ec Ec1 ma
= sin ωm t sin ωc t[sin ωc t + cos φ + cos ωc + sin φ]
2
DSB-SC Product v(t) Law pass

signal SCT modulator filter Output v0(t)
ec(t) . E⬘csin(␻ct⫹␾)
Local
oscillator
Figure 1.23: Synchronous detector.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 38 — #38

i i
ma Ec Ec
y(t) = sin ωm t sin2 ωc t cos φ
2
ma Ec Ec2
+ sin ωm t sin2 ωc t cos ωc t sin φ
2

ma Ec Ec1 1 − cos 2ωc t
= sin ωm t cos φ
2 2
ma Ec Ec1 sin 2ωc t
+ sin ωm t · sin φ
2 2
The output of LPF,
ma Ec Ec
V0 (t) = sin 2m t cos φ
4
If φ = 0 then, V0 (t) is maximum. φ = 90, then V0 (t) is minimum.
This is known as quadrative null effect. For continuous charge in phase,
fading occurs at the output.
(2) Costas-PLL/Costas loop:
1/2Em(t)
Product Low-pass Eccos ␾ Demodulator
filter signal
modulator
Eccos (2␲fct⫹␾)
Voltage Phase
DSB-SC controlled discriminator
signal oscillator
Em(t)cos 2␲fct
⫺90⬚
phase shifter
Ecsin (2␲fct⫹␾)
1/2Em(t)
Product Low-pass Ecsin ␾
modulator filter
Figure 1.24: Costas loop.
The costas receiver is the practical synchronous receiver system

suitable for demodulating DSB-SC. It consists two coherent detectors
which practically solves the problem of synchronization. One coherent
detector is called in-phase detector is also called as I-channel detector
and the other detector is called as quadratum phase detector is also
known as Q-channel detector.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 39 — #39

i i
To the I-channel detector the modulated signal and the carrier signal
are given as it is θ to the Q-channel detector the carrier signal is phase
shifted by 90 degrees and given as the input along with the modulated
signal.
The phase discrimination detects any phase difference and produces
error control voltage which controls the phase error by using the
voltage controlled oscillator. Therefore the error voltage will decide the
feedback in controlling the phase of the local carrier so that the carrier
signal is in synchronous with the carrier of the modulating signal.
1.3 Hilbert Transform
Hilbert transform of a signal x(t) represented as,
xn (t) (or) H(x(t))
It’s used to provide −π

2
(or) −90◦ phase shift for every component present in
the signal x(t).
Mathematically,
ˆ
1 α x(τ )
H[x(t)] = xn (t) = · dτ
π −α t − τ
1 1
xn (t) = · x(t)
π t
1.3.1 Characteristics of Phase Shifting System

1. Magnitude of the frequency components present in x(t) remains
unchanged, when it is passed through the signal, that is, H(ω) = 1.
2. The phase of the (+ve) frequency components is shifted by −π 2
(or)
−90◦ . Hence, H(ω) = −jX(ω)sgn(ω).
Xh (ω) = X(ω) · H(ω) (1.41)

H(ω) = X(ω) − jsgn(ω) (1.42)
x(t) ⫺␲/2 phase shifter xh(t)

X(␻) transfer fh⫽x(t) X (t)
h
Figure 1.25: Phase shifting system.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 40 — #40

i i
Xh (ω) = −jX(ω)sgn(ω) (1.43)
Derivation of H(ω):

H(ω) = ejθ(ω) ∵ H(ω) = |H(ω)| · ejθ(ω) = 1 · ejθ(ω) (1.44)

π
for ω < 0
θ (ω) = 2
−π
2
for ω < 0
Hence,

ejπ/2 , ω<0
H(ω) = −jπ/2
(1.45)
e , ω>0

π π
ejπ/2 = cos + j sin = j
2 2
−jπ/2 π π
e = cos − j sin = −j
2 2

j, ω < 0
H(ω) =
−j, ω > 0

H(ω) 1, ω < 0
=
j −1, ω > 0
H(ω) = −jsgn(ω)
∴ xh (ω) = −jX(ω) · sgn(ω)
Taking inverse Fourier transform,
F −1 [xh (ω)] = xh (t)

= F −1 [−jX(ω)sgn(ω)]
1
xh (t) = x(t) ·
πt
1.3.2 Properties of Hilbert Transform

1. A signal x(t) and its Hilbert transform xh (t) have the same amplitude
spectrum and same auto correlation function.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 41 — #41

i i
2. x(t) and xn (t) are orthogonal. That is,

ˆ α
x(t)xh (t) = 0
−α
3. Hilbert transform of xh (t) is −x(t).
1.3.3 Pre Envelope (or) Analytic Signal

Useful is deriving the general expression of the SSB-SC signal. The preenve-
lope of real valued signal x(t) is defined as,
xp (t) = x(t) + jxh (t)
Obviously, the pre-envelop xp (t) is a complex valued signal. The real part of
xp (t) is x(t) and the imaginary part is its Hilbert transform xh (t).
1.3.4 Complex Envelope

The new quality based on the analytic signal, called the complex envelope is
defined as,
x(t) = x(t)ej2π fc t
Taking Fourier transform,
F[x(t)] = X(ω)
⎧
⎪
⎨2X(f + fc )
⎪ for f > 0
X(ω) = H(0) for f = 0
⎪
⎪
⎩0 for f < 0
The complex envelope is just the low pass signal, part of the analytic signal.
The analytic low pass signal has been multiplied by the complex exponential
at the carrier signal.
1.4 Single-Side Band-Suppressed Carrier System
Both the sidebands in the DSB-SC are carrying the same information hence
only one sideband is sufficient to convey the message. So we can suppress
one sideband and transmit the other. This is called as single sideband with
suppressed carrier (or) Single sideband (SSB) system.
In this case, the power saving increases by eliminating any one sideband
in addition with the carrier component.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 42 — #42

i i
Advantages:
(i) Efficiency is increased (83.33%).
(ii) Bandwidth is reduced (fm ).
(iii) Fading effect which arises because of the interference of carrier and
two sidebands is removed in SSB.
1.4.1 Power Distribution

Ec2 m2a
Pt(SSB−SC) = PUSB =
8R
Power saving of SSB-SC = PtAM − PtDSB−SC
Ec2 m2a
Pt(SSB−SC) = ·
2R 4
If R = 1
Ec2 m2a
Pt(SSB−SC) = ·
2 4
2
m
Pt(SSB−SC) = Pc · a
4
For 100% modulation, ma = 1,
Pc
Pt(SSB−SC) =
4
1.4.2 Power Saving of SSB-SC with Respect to AM Signal
Pt(AM) − Pt(SSB−SC)
Power saving =
Pt(AM)

m2 m2
Pc 1 + 2a − Pc · 4a
= 2

Pc 1 + m2a
m2 m 2a
Pc + Pc 2a − Pc
= 2
4
Pc 1 + m2a
P m2
P c + c2 a
= 2

Pc 1 + m2a
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 43 — #43

i i
4Pc (4 + m2a )
Power saving = 2
Pc (1 + m2a )
4 + m2a 2
= ×
4 2 + m2a
4 + m2a
=
2(2 + m2a )
For 100%ma = 1,
5
%η = = 0.8333
6
% of power saving = 83.33%
1.4.3 Power Saving with DSB-SC
Pt(DSB−SC) − Pt(SSB−SC)
η=
Pt(DSB−SC)
P c m 2a P m2
− c4 a
= 2
p c m 2a
2
2
Pc ma 2
η= ·
4 Pc m2a
1
=
2
= 0.5
% of η = 50%
1.4.4 Time-Domain of SSB-SC
␻ ␻
xp*(t) xp(t)
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 44 — #44

i i
A(V)
fc⫺fm fc fc⫹fm fc (Hz)

DSB-SC
A(V)
fc⫺fm fc fc⫹fm fc (Hz)
SSB-SC
Complex
envelope
Pre-envelope
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 45 — #45

i i
S(t) = xp (ω) + x∗p (ω) (1.46)

1
xp (ω) = xp (t) · e−jωc t (1.47)
4
1
x∗p (ω) = x∗p (t) · ejωc t (1.48)
4
xp (t) = x(t) + jxh (t) (1.49)
x∗p (t) = x(t) − jxh (t) (1.50)
1
xp (ω) = [x(t) + jxh (t)]e−jωc t (1.51)
4
1
Xp∗ (ω) = [x(t) − jxh (t)]ejωc t (1.52)
4
Substitute eqn. (1.52) and (1.53) in (1.47), we get
1 1
S(t) = [x(t) + jxh (t)]e−jωc t + [x(t) − jxh (t)]ejωc t
4 4
1
S(t) = [x(t)e−jωc t + jxh (t)e−jωc t + x(t)ejωc t − jxh (t)ejωc t ]
4
1
S(t) = [x(t)[e+jωc t + e−jωc t ] + jxh (t)[e−jωc t − e−jωc t ]]
4
1
= [x(t) · 2 cos ωc t − j(2j sin ωc t)xh (t)]
4
1
S(t) = [x(t) cos ωc t + xh (t) sin ωc t]
2
Substitute x(t) = sin ωm t, xh (t) = cos ωm t
1
S(t) = [sin ωm t cos ωc t + cos ωm t sin ωc t]
2
1
S(t) = [sin(ωc + ωm )t]
2
Substitute x(t) = cos ωm t, xh (t) = sin ωm t
1
S(t) = [cos ωm t cos ωc t + sin ωm t sin ωc t]
2
1
S(t) = [cos(ωc − ωm )t]
2
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 46 — #46

i i
1.4.5 Single Side Band Suppressed Carrier (SSB-SC)

Since lower side band (LSB) and upper side band (USB) contains the same
information, one of the side band can be suppressed the resultant signal is
called as single side band suppressed carrier.
The generation of SSB-SC can be done in three ways namely:
(1) Filter method.
(2) Phase shift (or) phasing method.
(3) Weavers method.
Filter Method: The block diagram of filter method is been shown below:
Side band Balanced Linear
Carrier Balanced
suppresion mixer Amplifier
signal modulator
filter
Modulating Antenna
Matching
signal
network
Figure 1.26: Filter method.
• In the above diagram linear amplifier amplifies the magnitude content

of the amplitude modulated signal.
• Side band suppression filter passes only the USB, LSB that is being
suppressed.
• Matching network boosts frequency content of amplitude modulated
wave.
Phase Shift Method/Phasing Method/Hartley Method: Let us consider the

below given block diagram.
Em (t) = x(t) = Em sin ωm t

V1 (t) = x(t)Ec sin ωc t
V2 (t) = xh (t)Ec cos ωc t
V0 (t) = V1 (t) + V2 (t)
= Ec [x(t) sin ωc t + xh (t) cos ωc t]

1 1
V0 (t) = [x(t) sin ωc t + xh (t) cos ωc t] ∵ Ec =
2 2
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 47 — #47

i i
Balanced Antenna
modulator
EC sin␻ct
v1(t)
Crystal
osciliator
90 phase Linear Matching
Summer network
shift Amplifier
90 phase
shift
EC cos␻ct v2(t)
Balanced
modulator
Figure 1.27: Phase shift method.
sin␻m(t) V1(t) V3(t)

Balanced
Balanced LPF modulator 3
modulator
V⬘3(t)
2sin(␻0⫹90⬚) 2sin␻ct
90⬚ phase
shift RF crystal
network DSC
2sin␻0t 2sin␻ct Summer
circuit Vc(t)
90⬚ phase
AF crystal
shift
OSC V⬘4(t)
network
2sin␻0t 2sin(␻ct⫹90⬚)
Balanced Balanced
modulator V (t)
LPF modulator 4
2 V4(t)
Figure 1.28: Weavers method.

Weavers Method

◦ 1
V1 (t) = sin ωm t2 sin(ω0 t+90 ) ∵ sin A sin B= [cos(A−B) + cos(A+B)]
2
= cos(ω0 t + 90◦ − ωm t) − cos(ω0 t + 90◦ + ωm t) (1.53)
From LPF 1,
V0 (t) = cos(ω0 t + 90◦ − ωm t) (1.54)
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 48 — #48

i i
From BM 2,
V2 (t) = sin ωm t2 sin ω0 t

= cos(ω0 − ωm )t − cos(ω0 + ωm )t
From LPF 2,
V4 (t) = cos(ω0 − ωm )t (1.55)
From BM 3,
V3 (t) = 2V3 (t) sin ωc t

V3 (t) = 2[cos(ω0 t + 90◦ − ωm t) sin ωc t]
V3 (t) = sin(ωc t − ω0 t − 90◦ − ωm t) + sin(ωc t − ω0 t + 90◦ − ωm t)
(1.56)
Output of BM 4,
V4 (t) = V4 (t)2 sin(ωc t + 90◦ )

= cos(ω0 − ωm )t × 2 sin(ωc t + 90◦ )
= sin(ωc t + 90◦ + ω0 t − ωm t) + sin(ωc t + 90◦ − ω0 t + ωm t)
(1.57)
Output of BM (1.55),
V0 (t) = V3 (t) + V4 (t)

= sin(ωc t + ω0 t + 90◦ − ωm t) + sin(ωc t − ω0 t + 90◦ − ωm t)
+ sin(ωc t + 90◦ + ω0 t − ωm t) + sin(ωc t + 90◦ − ω0 t + ωm t)
= sin(ωc t + ω0 t − ωm t + 90◦ ) + sin(ωc t − ω0 t + ωm t − 90◦ )
+ sin(ωc t + ω0 t − ωm t90◦ ) + sin(ωc t − ω0 t − ωm t + 90◦ )
V0 (t) = 2 sin(ωc t + ω0 t − ωm t + 90◦ )
Resembles LSB fc − fm .
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 49 — #49

i i
1.4.6 Demodulation of SSB-SC
Coherent Product
synchronous LPF
modulator v0(t)
detector
Ec1sin␻ct
Local
oscillator
Ec Em
VSSB−SC (t) = cos(ωc t − ωm t)
2
Ec Em
V1 (t) = cos(ωc t − ωm t)Ec1 sin ωc t
2
Ec Ec1 Em
V1 (t) = [sin(2ωc t − ωm t) + sin(ωm (t))]
2
From LPF,
Ec Ec1 Em
V0 (t) = [sin ωm t]
2
1.5 Vestigial Sideband Suppressed Carrier (VSB-SC)
Desired side band is allowed to pass completely whereas small portion

(vestige/base) of undesired sideband also transmitted. Transmitted vestige
of undesired sideband compensates the loss of desired sideband.
1.5.1 Frequency Spectra of (VSB-SC)

Amp (V)
Ec
Ecma Ecma
2 2
Frequency (Hz)
LSB fv fc USB
fc⫺fm fc⫹fm
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 50 — #50

i i
Bandwidth = fm + fv
1. Bandwidth of VSB-SC is more than DSB-SC but less than SSB-SC.

2. Efficiency of VSB-SC is more than DSB-SC but less than SSB-SC.
3. Power consumption is less than DSB-SC but more than SSB-SC.
1.5.2 Generation of VSB-SC
Product m(t) VSB s(t)

x(t)
modulator m(f) filter H(f) s(f)
EC cos2␲fct fm⫹fv
Oscilliator
Figure 1.29: Generation of VSB-SC.
Output of product Modulator: m(t)
m(t) = x(t)Ec cos 2π fc t (1.58)
Output of VSB filter
S(t) = m(t)H(f ) (1.59)
In frequency Domain:
S(t) = S(f ) = M(f )H(f ) (1.60)

EC
m(t) = M(f ) = X(f ) [δ(f + fc ) + δ(f − fc )] (1.61)
2
Ec
M(f ) = [x(f + fc ) + x(f − fc )] (1.62)
2
Substituting eqn. (1.62) in (1.60),
Ec
S[f ] = [x(f + fc ) + x(f − fc )]H(f ) (1.63)
2
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 51 — #51

i i
1.5.3 Demodulation of VSB-SC (Coherent Detector)
V(t) = δ(t) cos 2π fct

S(f )
V(f ) = [δ(f + fc ) + δ(f − fc )]
2
1
V(f ) = [S(f + fc ) + S(f − fc )] (1.64)
2
To find S(f + fc )
Ec
S(f ) = [X(f + fc ) + X(f − fc )]H(f ) (1.65)
2
Replace f → f + fc in eqn. (1.63)
Ec
S[f + fc ] = [X(f + fc + fc ) + X(f + fc − fc )]H(f + fc )
2
Ec
S[f + fc ] = [X(f + 2fc ) + X(f )]H(f + fc ) (1.66)
2
Replace f → f − fc in eqn. (1.63),
Ec
S(f − fc ) = [X(f ) + X(f − 2fc )]H[f − fc ] (1.67)
2
Substitute eqns (1.65) and (1.66) in (1.64),
Ec
V(f ) = [X(f + 2fc )H(f + fc ) + X(f )H(f + fc )]
4
+X(f )H(f − fc ) + X(f − 2fc )H(f − fc )]
Ec
= [X(f )[H(f + fc ) + H(f − fc )] + X[f + 2fc ]H[f + fc ]
4
+X(f − 2fc )H(f − fc )]
Ec
V(f ) = [X(f )]
4
Vc Vc
V1 (f ) = [M(f )]
4
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 52 — #52

i i
s(t) Product v(t) LPF

modulator fm x(t)
cos2␲fct
Local
oscilliator
Figure 1.30: Demodulation of VSB-SC.
Magnitude Response of USB Filter:
H(f)
LSB USB
1
fc⫺fv fc fc⫹fv fc⫹fm f (Hz)
Vestige portion of LSB
Figure 1.31: Magnitude response of VSB filter.
Properties:
• Frequency response of the filter is normalized |H(fc )| = 1/2.
• Sum of values of the magnitude response at any two frequencies
equally displaced above and below ‘fc ’ is unity.
H(f − fc ) + H(f + fc ) = 1
fc − fv ≤ f ≤ fc + fv
• Phase response is linear.

• fc to fc − fm is LSB, in this fc − fv , fc is transmitted.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 53 — #53

i i
1.6 AM Super Heterodyne Receiver
Performance of a transformer is poor for higher radio frequencies.
1.6.1 Heterodyning
The process of mixing two signals having difference frequencies to produce
a new frequency is called as “Heterodyning”. In Addition it consists of:
(i) Mixer
(ii) Local oscillator
(iii) If amplifier
The five sections of super heterodyning:
1. RF section
2. Mixer/Converter section
3. IF section
4. Audio detector section
5. Audio amplifier section
Principle: RF signal is converted to IF.
RF −→ IF (Intermediate Frequency) −→ 455 KHz
(1) RF section: These section having two types (i) preselector and (ii)
RF amplifier (determines sensitivity). Preselector is broad tuned BPF
with adjustable centre frequency that is tuned to the desired carrier
frequency of the incoming signal. Provide “Enough initial Band
limiting” to prevent unwanted radio frequency to image frequency.
Local
Antenna Gang tuning oscillator
fLO fRF+fLO
Pre RF IF
amplifer Mixer BPF amplifer
selector
Loud Audio Audio

speaker amplifier detector
Figure 1.32: Block diagram of super-heterodyne receiver.
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 54 — #54

i i
(2) Mixer/Converter section: Combination of mixer and local oscillator

provides heterodyning function
Mixer + local oscillator => First detector

Detector (or) demodulator => Second detector
Image frequency: Unwanted signal at frequency fsi is known as image

frequency. Image of the signal frequency ‘fs ’
fsi = fs + 2fi
fo = fs + fi (1.68)
fo → local oscillator frequency

fs → signal frequency
fi → intermediate frequency
fsi = fo + fi (1.69)
Substitute eqn. (1.67) in (1.68)
fsi = fs + fi + fi
fsi = fs + 2fi
Image frequency rejection ratio (IFRR): It is defined as the ratio of the gain
at the signal frequency to the gain of the image frequency.
Frequency changing and tracking (mixing):
1. TRACKING: Is a process in which the local Oscillator frequency tracks
the signal frequency to have a correct frequency difference.
2. IF: Mixer produces fo , fs , fo + fs , fo − fs out of this fo − fs is selected
remaining are rejected. This is called IF.
RF amplifier Mixer Output
Local oscillator
Major Factors Influencing the Choice of IF:

1. If IF is high, the poor selectivity and poor adjacent channel rejection.
Tracking difficulties high.
2. If is low, the high selectivity frequency stability is low.
i i
i i
i
i
i
i
Table 1.1: Compare the characteristics of DSBFC, DSB-SC, SSBSC, VSB schemes.
Parameter AM/DSBFC DSB-SC SSBSC VSB
Method Carrier and both sidebands Only sidebands Only one side band One side band and part of other
side band (undesired side band)
Bandwidth 2fm 2fm fm (fm + fv ) (i.e.,) fm < BW < 2fm
m2 m 2a P c m 2a
Power Pt = Pc 1 + 2a Pt = Pc 2
; Pt = [0.5Pc ]ma =1 Pt = 4
; Pt = [0.25Pc ]ma =1 Pt(SSB) < Pt(VSB) < Pt(DSB)
transmitted Pt = [1.5Pc ]ma =1
Efficiency η = 33.3% η = 66.6% η = 83.3% with respect to AM. 33.3% < η < 100%
η = 50% with respect to DSB-SC
Full name Double side band full carrier Double side band suppressed Single side band suppressed Vestigal side band transmission.
carrier carrier
Generation Easy Easy Complex Complex
Selective Heavy distortion More distortion compared to Least distortion Received signal is distorted.
fading SSBSC
Application AM broadcasting Carrier telephony Wireless Television broad casting.
Generation Done using modulators, Using (i) Filter Method Using product modulator and
methods Non linear modulator (i) Balanced Modulator (ii) Phase shift (or) phasing bandpass filter.
(i) Square law modulator (ii) Ring Modulator method
(ii) Product modulator (iii) Weavers method
(iii) Balanced modulator
Linear Modulators
(i) Transistor modulator
(ii) Switching modulator
AM Modulators
* Low Level/Emitter Mod−r
* High Level/Collector Mod−r
“Unit-1” — 2018/12/15 — 16:11 — page 55 — #55
Demodulation (i) Envelope detector/Diode (i) Costas loop detection Coherent SSB demodulator Demodulation is done using
Technique detector (ii) Synchronous coherent detector.
(ii) Negative peak clipping detector/coherent detector
(iii) Diagonal clipping in diode
detector.
i
i
i
i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 56 — #56

i i
IF Amplifier:
• Relatively high gain tuned amplifier
• Similar to RF amplifier
• Except operate over a relatively narrow fixed frequency band
Demodulator:
1. Output of IF section applied to a demodulator which recovers the
base-band signal.
2. Also said to be audio detector.
Summary of Super-Heterodyne Principle:
(i) Select the desired frequency fs by tuning RF amplifier
(ii) Local oscillator is tuned to frequency ‘fo ’ (fo > fs )
(iii) Mixer produces IF. fi = fo − fs
(iv) IF amplifier amplifies IF
(v) Detector will demodulate this signal to recover the modulating signal
(vi) Audio amplifier amplify AF signal and apply it to loud speaker
Problem 1: Carrier wave of frequency 10 MHz and peak value of

10 V is amplitude modulated by 5 KHz sine wave of amplitude 6 V.
Determination the modulation index, percent modulation, lower side
frequency, upper side frequency and draw the frequency spectrum of the
modulated signal.
Solution:
Given data: fc = 10MHz, Vc = 10V, fm = 5 kHz, Vm = 6V
(i) Modulation index:
Vm 6
ma = = = 0.6
Vc 10
ma = 60%
(ii) Lower side frequency:
LSF = fc − fm = 10MHz − 5KHz

LSF = 9.995MHz
(iii) Upper side frequency:
USF = fc + fm = 10MHz + 5KHz

USF = 10.005MHz
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 57 — #57

i i
(iv) Frequency spectrum:

Carrier Vc⫽10V
MaVc MaVc
LSB⫽ ⫽3V USB⫽ ⫽3V
2 2
f (MHz)
fc⫺fm fc fc⫹fm
(9.995 MHz) (10 MHz) (10.005 MHz)
Problem 2: How many AM broadcast station can be accommodated

in 10 KHz bandwidth if the highest frequency modulating a carrier is 5
KHz?
Given data: Bandwidth=100 KHz, fm = 5 KHz
Solution:
Total bandwidth
Number of stations accommodated =
Bandwidth per station
100KHz
=
5KHz
Total number of stations accommodated = 10 stations
Problem 3: Certain transmitter radiates 9 kW with carrier unmod-

ulated and 10.125 kW, when the carrier is sinusoidally modulated.
Calculate the modulation index. If another sine wave, corresponding
to 40% modulation is transmitted simultaneously. Determine the total
power.
Given data:
Pt = 9 kW
Pc = 10.125 kW
ma = 40% or 0.4
Solution: Total power:

m2a
Pt = Pc 1 +
2
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 58 — #58

i i

Pt
m2a=2 −1
Pc

9k
=2 −1
10.125k
ma = 0.5
(ii) If two signals are simultaneously modulated,

mt = m21 + m22

mt = 0.52 + 0.42 = 0.64
(iii) Total power

mt 2
Pt = Pc 1 +
2
Pt = 9 × 103 (1 + 0.205)
Pt = 10.84 kW
Problem 4: A message signal
m(t) = (2 cos 100π t + 3 sin 24π )
this message modulates a carrier of frequency 12 MHz and amplitude of

10V. Determine:
(i) Time domain representation of AM wave
(ii) Determine total modulation index
(iii) Total transmission power
(iv) Plot spectrum of AM wave
(v) Find total power of load resistance across 100.
Given: m(t) = 2 cos 100π t + 3 sin 24π t; fc = 12MHz, Vc = 10V

(i) Time domain representation: AM wave represent as
VAM (t) = [Vc + Vm sin 2π fm t] sin 2π fc t

VAM (t) = [10 + 2 cos 100π t + 3 sin 24π t] sin 2π fc t
(ii) Total modulation index:

mt = m21 + m22
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 59 — #59

i i
Vm1 2
m1 = = = 0.2
Vc 10
Vm2 3
m2 = = = 0.3
Vc 10

mt = (0.2)2 + (0.3)2
mt = 0.36
(iii) Total transmitted power:

m2t
Pt = Pc 1 +
2
Vc2
Pc =
2R
(10)2
=
2 × 100
= 0.5W
Pc = 0.5W

0.362
Pt = 0.5 1 + = 0.532W
2
1.5 10V 1.5
1 1
f (MHz)
11.9999 11.999 12 MHz 12.000012 12.00005
Problem 5: A DSB-SC signal with fc = 1000 kHz and fm = 10 KHz.

What is the bandwidth of the signal?
Given data: fc = 1000 KHz; fm = 10 KHz
Solution:
Band width of DSE-SC = 2fm = 2 × 10KHz

Band width of DSE-SC = 20KHz
i i
i i
i i
“Unit-1” — 2018/12/15 — 16:11 — page 60 — #60

i i
Problem 6: For an AM broadcast-band super heterodyne receiver with

IF, RF and local oscillator frequencies of 455 KHz, 600 KHz and 105 KHz
respectively. Determine (a) Image frequency and (b) IFRR far pre-selector
Q of 100.
Given data: fRF = 60 KHz, fIF = 455 KHz, Q = 100, fLO = 1055 KHz.
Solution:
(a) Image frequency:
fIM = fLO + fIF

fIM = fRF + fIF + fIF
fIM = fRF + 2fIF
fIM = 1055KHz + (455KHz)2
fIM = 1510KHz
(b)

fIM fRF
IFRR = 1 + Q2 p2 where P = − = 2.113
fRF fIM

IFRR = 1 + 1002 (2.113)2
IFRR = 212.15
(IFRR)dB = 10 log IFRR
(IFRR)dB = 23.25dB
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 61 — #1

i i
Unit
Angle Modulation 2
2.1 Introduction
The angle (frequency or phase) of the carrier signal is varied according to

the manage signal, then it is called angle modulation, here the amplitude of
the carrier signal is constant. These are two types:
1. Frequency Modulation (FM)
2. Phase Modulation (PM)
2.1.1 Frequency Modulation (FM)

When frequency of the carrier signal varies as per the amplitude variation
of modulating signal, then it is called frequency modulation FM. In FM,
amplitude of the modulated carrier remains constant.
2.1.2 Phase Modulation (PM)

When the phase of the carrier varies as per amplitude variation of modulat-
ing signal, then it is called phase modulation (PM). In PM, amplitude of the
modulated carrier remains constant.
2.1.3 Time Domain Representation for Phase Modulation
EPM (t) = Ec cos θi (t) (2.1)
θi (t) is instantaneous phase
θi (t) = (ωc t + kp Em (t))
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 62 — #2

i i
62 . Unit 2: Angle Modulation
kp is phase sensitivity deviation/phase deviation
θi (t) = [ωc t + kp Em cos ωm t]

θi (t) = [ωc t + mp cos ωm t] (2.2)
EPM (t) = Ec cos[ωc t + mp cos ωm t] (2.3)

mp = kp Em
mp is modulation index of phase modulation. Therefore
mp αEm
Eqn. (2.3) represents time domain represents of phase modulated signal.

Therefore
EPM (t) = Ec cos[ωc t + mp sin ωm t]
Phase Deviation The phase angle of the carrier signal varies from its
unmodulated signal during modulation process is known as phase deviation.
mp
kp = , kp (or) θ
Em
2.1.4 Time-Domain Representation of FM

EFM (t) = Ec cos fi (t) (2.4)
fi (t) is instantaneous frequency

ˆ t
fi (t) = ωc t + 2π kf Em (t) · dt
0
ˆ t
= ωc t + 2π kf Em cos ωm t · dt
0
ˆ t
= ωc t + 2π kf Em cos ωm t · dt
0
t
sin ωm t
= ωc t + 2π kf Em
ωm 0
Em
= ωc t + 2π kf · sin ωm t
ωm
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 63 — #3

i i
2π kf Em
fi (t) = ωc t + sin ωm t
2π fm
therefore

kf Em
fi (t) = ωc t + sin ωm t (2.5)
fm

kf Em
EFM (t) = Ec cos ωc t + sin ωm t
fm
EFM (t) = Ec cos[ωc t + mf sin ωm t] (2.6)
kf is frequency deviation sensitivity (or) frequency deviation (or) deviation

sensitivity (a constant)
f = kf · Em ⇒ Frequency deviation
kf Em f
mf = = ⇒ Modulation index of frequency modulation
fm fm
2.1.5 Deviation Ratio (DR)
Maximum frequency deviation

DR =
Maximum modulating index
f (max)
=
fm (max)
f (max) is constant value is 75 kHz. If fm (max) = 15kHz
75kHz
DR = =5
15kHz
2.1.6 Percentage of Modulation of FM
f (actual)
% of modulation = × 100
f (max)
for example, if f (actual) = 20kHz and f (max) = 75kHz (constant)
20kHz
Percentage of modulation = × 100
75kHz
= 26.67%
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 64 — #4

i i
2.2 Generation of FM from PM
兰Em(t)
Em(t)
兰 PM FM wave
Eccos ␻ct
Figure 2.1: Generation of FM from PM.
VPM (t) = Ec cos[ωc t + kp Em cos ωm t]

ˆ t
VFM (t) = Ec cos ωc t + kp Em cos ωm t
0

kp Em
VFM (t) = Ec cos ωc t + sin ωm t (2.7)
ωm
2.3 Generation of PM from FM
d V (t)
Vm(t) d dt m
FM PM wave
dt
Vccos ␻ct
Figure 2.2: Generation of PM from FM.

kf Em
VFM (t) = Ec cos ωc t + sin ωm t
ωm

kf Em d
VPM (t) = Ec cos ωc t + (sin ωm t)
ωm dt

kf Em
= Ec cos ωc t + cos ωm tωm
ωm

VPM (t) = Ec cos ωc t + kf Ec cos ωm t (2.8)
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 65 — #5

i i
2.4 Types of FM
Depending on the value of modulation index, FM is classified into two types.

(1) Narrow band FM and (2) Wide band FM.
2.4.1 Narrowband FM
When the value of modulation index mf is smaller than one radian in a FM.
It is called as narrow band FM (i.e.,) mf 1.
Let the message signal be represented as,
em (t) = Em sin ωm t
Let the carrier signal be represented as,
ec (t) = Ec sin(ωc t + φ)
ec (t) = Ec sin θ
where θ = (ωc t + φ) and dθ

dt
= ωc is angular frequency of carrier signal.
After frequency modulation,
ωc = ωc + kem (t)
= ωc t + kEm sin ωm t
The frequency derivation is maximum, when
sin ωm t = ±1
hence ωi = ωc ± kEm .
The frequency derivation is proportional to the amplitude of modulating
voltage, hence it can be written as,
2π f = kEm
ωi = ωc ± 2π f sin ωm t
ˆ
φi = ωi dt
ˆ
= (ωc ± 2π f sin ωm t)dt
f
φi = ωc t ± sin ωm t
fm
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 66 — #6

i i
eFM (t) = Ec sin φi t

f
= Ec sin ωc t ± sin ωm t
fm

= Ec sin ωc t ± mf sin ωm t
eFM (t) = Ec sin ωc t · cos(mf sin ωm t)
+Ec cos ωc t · sin(mf sin ωm t)
For narrow band FM assume the modulation index mf is small compared to

one radian, hence ωc may use the following approximation,
cos(mf sin ωm t) = 1
sin(mf sin ωm t) = mf sin ωm t
because cos θ = 1 and sin θ = θ , if ‘θ ’ is small.

ˆ f m /2
cn = fm Ec ejm f sin ωm t · e−j2π fm nt · dt
−f m /2
Let y = 2π fm t, hence limits will change from −π to π
dy
= 2π fm
dt
dy
dt =
2π fm
ˆ π
dy
cn = fm Ec ej(mf sin y−ny) ·
−π 2π fm
ˆ π
fm Ec
cn = · e(mf sin y−ny) · dy
2π fm −π
The above integral is known as the nth order Bessel function of the first kind
and its given,
cn = Ec Jn (mf ) (2.9)

α

x(t) = Ec Jn (mf )ej2π fm nt (2.10)

n=−α
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 67 — #7

i i

α

VFM (t) = Re Ec Jn (mf )e j2π f m nt

·e jωc t
n=−α
α

= Ec Re [Jn (mf )ej2π(nfm +fc )t ]

n=−α
Hence we get,
eNBFM (t) = Ec cos ωc t + Ec cos ωc t(mf sin ωm t)
2.4.2 Wideband FM
VFM (t) = Ec cos(ωc t + mf sin ωm t) (2.11)
Expression form,
VFM (t) = Re[Ec ej(ωc t+mf sin ωm t) ] (2.12)

x(t) = Ec ejmf sin ωm t (2.13)
VFM (t) = Re[x(t)ejωc t ] (2.14)
x(t) can be expressed with the help of Fourier series,

α

x(t) = cn ej2π fm nt (2.15)

n=−α
where
ˆ f m /2
cn = fm x(t)e−j2π fm nt · dt (2.16)
−f m /2
α

= Ec Jn (mf ) cos[2π(fc + nfm )t]

n=−α
α

VWBFM (t) = Ec Jn (mf ) cos[2π(fc + nfm )t] (2.17)

n=−α
Taking Fourier series of eqn. (2.17),

α
Ec
VWBFM (t) = Jn (mf ){δ(f − (fc − nfm )) + δ(f + (fc + nfm ))}
2 n=−α
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 68 — #8

i i
2.4.3 Generation of NBFM
⫺ Narrow
Modulating Product
Integrator ⌺ band
wave modulator
FM wave
⫹
Ecsin␻ct
Carrier
⫺90⬚ phase
wave
shifter
Eccos␻ct
Figure 2.3: Generation of NBFM.
The modulated signal produced by NB modulator differs from these two

ideal condition. They are
(i) The envelope contain a residual amplifier modulation. Therefore, the
amplitude varies with time.
(ii) Angle θi (t) contain harmonic distortion in the form of 3rd and higher
harmonics of modulating frequency fm .
The NBFM requires the same transmitter bandwidth as that of the AM signal
and has only one upper side band and one lower side band.
2.4.4 Comparison of WBFM and NBFM

S.No Narrow band FM Wide band FM
1. Modulation index mf < 1 Modulation index mf > 1
2. Spectrum contains Spectrum contains infinite
2-sidebands and carrier number of sidebands and carrier
3. BW= 2fm BW= 2(δ + fm )
4. Maximum deviation=5 kHz Maximum deviation= 75 kHz
5. Range of modulating Range of modulating frequency
frequency 30 to 3 kHz 30 kHz to 5 mHz
6. Noise suppression is less Noise suppression is more.
7. It’s used for mobile It’s used for broad casting and
communication entertainment
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 69 — #9

i i
Carson’s Rule It states that bandwidth of FM signal is equal to twice the sum
of frequency deviation and maximum modulating frequency.
Bandwidth = 2(f + fm )
where f is frequency deviation and fm maximum modulating frequency.
2.5 Generation of FM Waves
The primary requirement of FM generator is a variable output frequency

with variation proportional to the instantaneous amplitude of the modulat-
ing voltage.
The other requirement is that the frequency deviation which is indepen-
dent of modulating frequency.
Metods of FM generation
Direct method Indirect method
Amstrong
Reactance Vaiactor diode method
modulator modulator
2.5.1 FET Reactance Modulator

Reactance of the drain and source contributed by gate voltage (Vg ) (or)
modulating signal voltage.
v
Z= (2.18)
gm Vg
where Vg is gate voltage and gm is bass conductance.
To find Vg ,
Vg = I1 R (2.19)
V
I1 = (where R < c)
R + jωc
1
I1 = vjωc (2.20)
Vg = vjωcR (2.21)
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 70 — #10

i i
A
I1 Id
C
D
G V
L C
I1
S Z
Modulating
Vg
signal R
B
Figure 2.4: FET reactance modulator.
v
z=
gm jωv.CR
1
z= where Ceq = gm CR
jω · Ceq
1
fc =
2π LCeq
• If voltage of modulating signal varies, (z) reactance of FET varies 1.

• If z- varies Ceq varies.
• If Ceq varies then Fc is varies, hence FM wave is generated.
2.5.2 Frequency Modulation Using Varactor Diode

Case (i) If modulating signal voltage increases
Capacitance of varactor diode decreases
k
cd = √ VD = v + Vm (t)
VD
Frequency of oscillator circuit increases,
1
fc = , Ceq = c1 + cd
2π LCeq
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 71 — #11

i i
⫹V
R1
To carrier
C3 RFC C2
AF VD⫽V⫹Vm(t) oscillator
input
Vm(t) D L1
R2 C1
Figure 2.5: Frequency modulation using varactor diode.
Case (ii) If voltage of modulating signal decreases

Capacitance of varactor diode increases. Frequency of oscillator
circuit decreases.
Capacitance of varactor diode:
k
Cd = √ (2.22)
VD
k is constant and VD is instantaneous voltage across the diode
VD = V0 + Vm (t) (2.23)
VD = V0 + Vm sin ωm t (2.24)
V0 is supply voltage. The total capacitance of the oscillation tank circuit is

(C0 + Cd ) and the instantaneous frequency of oscillator ω,
1
ωi = √ (2.25)
L0 (C0 + Cd )
1
ωi = (2.26)
L0 C0 + √KV
0
Thus the frequency modulated signal is generated.

The radio frequency choke (RFC) has high reactance at the carrier
frequency to present carrier signal from getting into the modulating signal.
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 72 — #12

i i
Application:
(1) Automatic Frequency Control (AFC)
(2) Remote tuning
2.5.3 Indirect Method of FM Generation (Armstrong Method)

Drawbacks in direct method
(i) LC oscillator circuits are used to generate carrier signal.
(ii) The carrier signal generated from LC oscillator is not stable for
broadcasting and communication purpose.
(iii) Due to non-linear characteristics of FET and varactor diode distor-
tion will occur in the output.
The indirect method is made up of two stages
(i) Generation of NBFM using phase modulation.
(ii) NBFM is converted into WBFM by using frequency multiplier and
mixer.
Step 1 Step 2
t
Vm(t)
兰 Vm(t)dt Phase NBFM Frequency WBFM
0
Integrator modulator modulator
Vmcos ␻mt (PM)
Vccos ␻ct
Capital
oscillator
Generation of NBFM Signal from phase modulation:
Vccos ␻ct Vccos ␻ct Combining NBFM

Capital
oscillator Buffer network (or)
Mixer
Vccos ␻ct Vccos[2␲fct+mf sin ␻ct]
Audio Vmcos ␻ct Phase 90⬚ Phase

amplifier modulator shift
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 73 — #13

i i
Frequency Multiplier:
Group of Mixer Group of

NBFM muliplier muliplier
NBFM
with with
fm and mf nfm and nmf
Crystal
oscillator
After frequency multiplier:
VFM (t) = Vc cos[2π nfc t + nmf sin 2π nfm t]
2.6 FM Demodulators/Detectors
To detect the FM signal, it is necessary to have a circuit whose output voltage

varies linearly with a frequency of input signal. The circuit used is called
frequency discriminator which converts FM signal into its corresponding
AM signal. Then the modulating signal is obtained from AM signal by
envelope detector.
Types of FM demodulators
Frequency discriminator Phase discriminator PLL FM

detecor
Balanced slope detector (i) Foster-seely discrminator

(or) (ii) Ratio detecor
Round Travis detector
2.6.1 Slope Detectors

Single slope circuit (or) Single tuned circuit.
• This circuit consist tuned circuit, which is slightly tuned with carrier
frequency ‘fc ’.
• It converts FM into a AM signal.
• AM signal is detected from envelope detector.
Principle A FM signal is applied at the input of the tuned circuit.
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 74 — #14

i i
FM input output
C
L
Center frequency of FM signal is ‘fc ’ is frequency deviation is f . Resonant

frequency of the tuned circuit is adjusted to (fc + f = f0 ).
Simple Slope Detector
(fc⫹⌬f) D1
fc
FM output C VAM C Output

R
Discriminator Envelope detector
Operation
• The circuit is tuned.
• When the signal frequency increases above ‘fc ’ with modulation, the
amplitude of the carrier voltage drops when the signal frequency
decreases below fc the carrier voltage rises.
• The change of voltage results because of the change in the magnitude of
the impedance in the tuned circuit as a function of frequency and results
in an effective conversion of frequency modulation into amplitude
modulation.
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 75 — #15

i i
Simple Slope Detector Charcterstics
VAM
fc f
Disadvantages
• It is linear only over a limited frequency range.

• It is difficult to adjust as the primary and secondary windings of the
transformer must be tuned to slightly different frequencies.
• A good discrimination circuit should respond only to frequency varia-
tions and not to amplitude variations.
2.6.2 Balanced Slope Detector

• Balanced slope detector consists of two slope detector.
• Input transformer has a center tapped secondary. Hence input voltages
to the two slope detector is 180◦ out of phase.
• Three tuned circuit, primary is tuned at ‘fc ’ R1 C1 is tuned at (fc + f )
and R2 C2 tuned at (fc − f ).
• V01 and V02 are output of two slope detector final output voltage ‘V0 ’ is
obtained by V0 = V01 − V02 .
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 76 — #16

i i
Characteristics of Balanced Slope detector (S-Curve)
Output voltage
(fc⫺⌬c)
fc (fc⫹⌬c)
Frequency of
input
(fc⫺⌬f) < fin < fc fin ⫽ fc fc < fin < (fc⫹⌬f)

Input of D1 and D1
Input of D1< D1 are equal Input of D1> D1
V0< V02 V0 ⫽ V01 V0 > V02
V0 is negative V0 ⫽ 0 V0 is postive
Operation
• For fin = fc :
Instantaneous frequency fin is fc the induced voltage ‘T1 ’ is winding is
induced voltage in T2 . Thus the input voltage to both diodes D1 and D2
are same (V01 = V02 ) but opposite in phase so output voltage is σ .
• For fc < fin < (fc + f )
(i) Voltage induced by ‘T1 ’ is > T2 .
(ii) Therefore input of D1 is > D2 , V02 > V01 . Hence output voltage V0
is +ve.
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 77 — #17

i i
• For (fc − f ) < fin < fc

(i) Voltage induced by ‘T1 ’ is < T2 .
(ii) Therefore input of D1 is > D2 , V01 > V02 . Hence output voltage V0
is −ve.
Advantages More efficient and better linearity than slope detector.
Disadvantages
• Difficult to tune because 3 tuned circuit should tune at different
frequencies.
• Amplitude limiting is not provided.
2.6.3 Foster-Seeley Discrimination (Phase Discrimination)

Principle At centre carrier frequency the 1◦ voltage leads the 2◦ voltage
by 90◦ and as the input frequency decreases the phase shift increases. This
discrimination is also known as center tuned discriminator.
1◦ voltage is coupled to 2 winding of the transformer by using coupling
capacitor and RFC (Radio Frequency Choke). The voltage of 2◦ is V2 which
is equally divided across upper and lower hay of the secondary coil.
• Voltage across diode D1 is VD1 = V1 + 0.5V2 .
• Voltage across diode D2 is VD2 = V1 − 0.5V2 .
• Voltage at the output terminals are
V01 = VD1 V02 = VD2
• Output voltage (V0 ) = |V01 | − |V02 |.
Cc RFC
D1
V01
C1 R1
Vout output
FM signal C V1 C3 message
signal
D2
V02
C2 R2
Figure 2.6: Foster-Seeley discriminator for FM demodulation
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 78 — #18

i i
Operation
Case (i) fin = fc (at resonance)
Voltage across VD1 and VD2 will be equals as V2 will have 90◦
phase shift with V2 . Hence output is zero. The vector addition
of VD1 and VD2 is shown in the below figure.
Vout = V01 − V02

Vout = 0
Hence net output is zero.
0.5V2 V01⫽V2⫹0.5V2
90 phase shift
V1 between V1and
V2 at fe
⫺0.5V2 V02⫽V2⫺0.5V2
Case (ii) fin > fc (above resonance):
Vout = V01 − V02

= (V1 + 0.5V2 ) − (V1 − 0.5V2 )
Vout = V2
Hence net output is positive.
0.5V2 V01
phase shift
is reduced due to
increase in frequency
V1
⫺0.5V2 V02
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 79 — #19

i i
Case (iii) fin < fc (below resonance)
Vout = V01 − V02

= (V1 − 0.5V2 ) − (V1 + 0.5V2 )
Vout = −V2
Hence net output is negative.
0.5V2 V01
phase shift
is reduced due to V1
decrease in frequency
⫺0.5V2 V02
Advantages
(1) Better linearity.
(2) Easy to align.
(3) Only two-tuned circuits needed.
Disadvantage Need separate amplitude limiting circuit.
2.6.4 Ratio Detector

Compare with foster seeley discriminator. This circuit is identical except the
following changes.
(1) The direction of diode D2 is reversed.
(2) A long value capacitor C5 has been included in the circuit.
Operation The polarity of voltage V02 is reversed, since connection of diode

D2 are reversed. Hence the voltages V01 and V02 across two capacitors add.
V0 = V01 + V02
when V01 increases, V02 decreases and vice versa.
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 80 — #20

i i
Input voltage due to diode D1 :

VR
V0 = V01 − (2.27)
2
VR = V01 + V02 (2.28)

V01 + V02
V0 = V01 − (2.29)
2
Output voltage due to diode D2 :
VR
V0 = −V02 + (2.30)
2
VR = V01 + V02 (2.31)

−V02 + V01
V0 = (2.32)
2
Output voltage of ratio detector:
Add eqn. (2.29) and (2.32)

V01 − V02 −V02 + V01
2V0 = +
2 2
2V0 = V01 − V02
V01 − V02
V0 =
2
Figure 2.7: Ratio detector.
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 81 — #21

i i
Since V01 ≈ |V0 | and V02 ≈ |VD2 |

1
V0 = (|VD1 | − |VD2 |)
2
output voltage V0 is half compared to that of foster-seeley circuit. When
input frequency increases above fc and |VD1 | > |VD2 |, hence output is ‘+ve’.
Similarly, if frequency decreases below fc and |VD1 | < |VD2 |, hence output is
negative.
– Merits: Reduced fluctuations in the output voltage compare to
foster-seeley circuit. Very good linearity, due to linear relationship
between primary and secondary.
– Demerits: The ratio detector may not tolerate the long period variation
in signal strength, hence AGC requires.
2.6.5 Block Diagram of PLL

FM wave e(t) Loop
filter V(t)
S(t)
b(t)
VCO
FM demodulator using PLL:
Input Ve Vc
Phase Low pass Error
FM detector filter amplifier
Demodulate
signal FM signal
output
fc Voltage VCO control
control
oscillator
2.7 Frequency Spectrum Analysis of Single

Tone Sinusoidal FM
2.7.1 Introduction
Frequency spectrum of FM wave is analysed by Fourier methods. But
analysis of FM and PM spectrum is quite difficult. So the analysis can be
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 82 — #22

i i
carried out with the help of BESSEL FUNCTION. FM wave
VFM (t) = Vc cos[ωc t + mf sin ωm t]
The above FM equation can be written using Bessel function as,
VFM (t) = Vc {J0 (mf ) sin ωm t + J1 (mf )[sin(ωc + ωm )t − sin(ωc − ωm )]

+J2 (mf )[sin(ωc + 2ωm )t − sin(ωc − 2ωm )t]
+J3 (mf )[sin(ωc + 3ωm )t − sin(ωc − 3ωm )t] + · · · so on}
where J0 , J1 , J2 , J3 , . . . are called as Bessel function. The value of Bessel func-

tion depends on modulation index (mf ). The above equation represents,
– 1st term: J0 (mf )Vc cos ωc t represents carrier.

– 2nd term: J1 (mf )Vc cos(ωc ± ωm )t represents pair of side frequencies.
– 3rd term: J2 (mf )Vc cos(ωc + 2ωm t) represents second pair of side
frequencies.
– 4th term: J3 (mf )Vc cos(ωc + 3ωm t) represents third pair of side fre-
quencies.
The amplitude coefficient Jn (mf ) are known as Bessel functions of order

n. Bessel function amplitude of carrier is at n = 0 and sidebands at n =
1, 2, 3, . . .. The values of Bessel functions are shown in the table given below.
Note: Frequency Modulation (FM) wave is analysed by a single frequency

sinusoidal wave which produces a peak phase deviation of (mf ) radians.
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 83 — #23

i i
VcJ0(mf)
LSB USB
VcJ1(mf) VcJ1(mf)
VcJ2(mf) VcJ2(mf)
VcJn(mf) VcJn(mf)
f
fc⫺2fm fc⫺fm fc fc⫹fm fc⫹2fm fc⫹nfm
Figure 2.8: Frequency Spectrum of FM.
Bandwidth of FM:
B.W = high frequency − low frequency

B.W = fc + nfm − (fc − nfm )
B.W = 2nfm Hz (2.33)
Properties of Bessel Function:

(1) Jn (mf ) = (−1)n J−n (mf )
(2) For small values of βJ0 (mp ) ≈ 1
Figure 2.9: Bessel function vs Modulation index (mf ).
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 84 — #24

i i
(3) J1 (mf ) = mf /2
(4) Jn (mf ) ≈ 0 when n > 2

(5) ∞ n=−∞ Jn (mf ) = 1
2
All the values of J-coefficients are dependent on modulation index mf and

this mf calculates many sideband components.
2.7.2 Power Calculation of FM
Total power = Pc + Psidebands

Pt = P0 + P1 + P2 + P3 + · · · Pn (2.34)
we know that
Vc2
carrier power (Pc ) =
2R
therefore
Vc2 J02 (mc )
carrier power of FM =
2R
because
J0 (mf ) = 1
Carrier power of FM
Vc2
(2.35)
2R
V02 2V12 2V22 2V 2

Pt = + + + ··· n
2R 2R 2R 2R
Here powers are written as,
∞
Vc2
2
General equation for average power = J (mf )
2R n=0 n
therefore
1 V02
Pt = + V1 + V2 + · · · Vn
2 2 2
R 2
The above equation is total power of FM wave.
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 85 — #25

i i
2.8 Comparison of FM and PM
FM PM
Frequency deviation is Phase deviation is proportion to
proportional to modulating modulating voltage.
voltage.
Frequency of the carrier is Phase of the carrier signal is
modulated with respect to message modulated with respect to message
signal. signal.
Modulation index is increased as Modulation index remains same
modulation frequency reduced. if the modulating frequency is

mf = fmf changed.
mp = Kp Vm
Better SNR Less SNR
Noise immunity is better than AM. Noise immunity is worst than FM.
FM is used widely in all application. PM is used in some mobile system.
Problem 1: The carrier is frequency modulated with the sinusoidal

signal of 2 kHz resulting in a maximum frequency deviation of 5 kHz.
Calculate the modulation index.
Solution: fm = 2kHz, f = 5kHz
f 5
mf = = = 2.5
fm 2
Problem 2: An FM wave is represented by the equation
VFM (t) = 10 sin[8 × 106 + 6 sin 3 × 104 t]
Calculate,
(i) Modulating frequency
(ii) Carrier frequency
(iii) Modulating index
(iv) Frequency deviation
Solution:

kEm
eFM (t) = Ec sin 2π fc t + · sin 2π fm t
2π fm
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 86 — #26

i i
kEm
mf =
2π fm
eFM (t) = Ec sin[2π fc t + mf sin 2π fm t]
Ec = 10, 2π fc = 8 × 106 , 2π fm = 3 × 104 , mf = 6

(i)
2π fm = 3 × 104
3 × 104
fm =
2π
= 4.474kHz
(ii)
2π fc = 8 × 106
8 × 106
fc =
2π
= 1.273kHz
(iii) Modulation index mf = 6

(iv)
f
mf =
fm
f = mf · fm
= 6 × 4, 474 × 103
= 28.644kHz
Problem 3: The carrier is frequency modulated with a sinusoidal signal

of 2 kHz resulting in a max deviation of 5 kHz. Find the Bandwidth of the
modulated signal.
fm = 2kHz
f = 5kHz
BW = 2[f + fm ]
= 2[5 × 103 + 2 × 103 ]
BW = 14kHz
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 87 — #27

i i
Problem 4: FM Modulator Signal
VFM (t) = 12 sin(6 × 108 t + 2.5 sin 1250t)
Calculate,
(i) Power of the modulated signal.
(ii) Maximum frequency deviation
(iii) Bandwidth of the modulated signal
(iv) Spectrum of frequency
Given Data:
VFM (t) = 12 sin(6 × 108 t + 2.5 sin 1250t)
Solution: General equation:
VFM (t) = Vc sin(2π fc t + mf sin 2π fm t)
Vc = 12v; mf = 2.5, 2π fc = 6 × 108 ; 2π fm = 1250, fm = 199.04 Hz

(i) Power of modulated signal.

1 V02
Total Power, Pt = + V12 + V22 + · · ·
R 2
because mf = 2.5, the corresponding Bessel values are J0 =
−0.05, J1 = 0.5, J2 = 0.45, J3 = 0.22, J4 = 0.07, J5 = 0.2
V02 = [Vc J0 (mf )]2

= (12 × 0.05)2 = 0.36
V12 = [Vc J1 (mf )]2
= (12 × 0.5)2 = 36
V22 = [Vc J2 (mf )]2
= (12 × 0.45)2 = 29.16
V32 = [Vc J3 (mf )]2
= (12 × 0.22)2 = 6.96
V42 = [Vc J4 (mf )]2
= (12 × 0.07)2 = 0.70
V52 = [Vc J5 (mf )]2
= (12 × 0.02)2 = 0.0576
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 88 — #28

i i

1 0.36
Total Power, Pt = + 36 + 29.16 + 6.96 + 0.70 + 0.0576
1 2
Pt = 73.05w
(ii) Maximum frequency deviation:

f
mf =
fm
f = fm · mf
f = 199.04 × 2.5
f = 497.6Hz
(iii) Bandwidth:
B.W = 2(f + fm )
= 2(497.6 + 199.04)
B.W = 1.39kHz

6 6
5.4 5.4
2.64 0.6 2.64
0.8 0.8
fc⫺4fm fc⫺3fm fc⫺2fm fc⫺fm fc fc⫹fm fc⫹2fm fc⫹3fm fc⫹4fm
Problem 5: An angle modulated wave is described by the equation
V(t) = 10 cos(2 × 106 π t + 10 cos 20000π t)
Find (i) Power of the modulated signal (ii) maximum frequency deviation
(iii) Bandwidth of the modulated signal.
Solution: FM wave is
VFM (t) = Vc cos[2π fc t + mf cos 2π fm t]
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 89 — #29

i i
and
V(t) = 10 cos(2 × 106 π t + 10 cos 2000π t)
Compare both equation, Vc = 10, mf = 10, 2π fm = 2000π , fm = 1kHz.
(i) Power of the modulated signal is carrier.
Vc2
Pc =
2R
102
=
2×1
Pc = 50w
(ii) Maximum frequency deviation.

f
mf =
fm
f = mf × fm
f = 10 × 1000kHz
f = 10kHz
(iii) Bandwidth:
B.W = 2(f + fm )
= 2[10 × 103 + 1 × 103 ]
B.W = 22kHz
i i
i i
i i
“Unit-2” — 2018/12/15 — 16:16 — page 90 — #30

i i
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 91 — #1

i i
Unit
Random Process 3
3.1 Random Process
If an outcome λ is associated with time t, then a function of λ and time t is

formed i.e., X(λ, t). Then the function X(λ, t) is known as Random Process.
3.1.1 Random Variables

Definition: A random variable is a mathematical function that maps the
outcomes of random experiments to numbers. That is, numerical value to
each possible outcome of an experiment. For example,
• A student’s grade in the mid term exam.

• Price of Google’s stock.
• Average temperature of each day.
• Number of heads when tossing 10 fair coins.
Random variables classified as,
(i) Discrete random variable

(ii) Continuous random variable
3.1.2 Discrete Random Variable

A random variable whose set of values is finite numbers.
Example: Experiment of tossing three-coins simultaneously. There are

eight-possible outcomes in a sample space(s) and random variable as
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 92 — #2

i i
92 . Unit 3: Random Process
number of heads and represented by ‘X’.
S = HHH HHT HTH THH HTT THT TTH TTT

X = {x1 = 3, x2 = 2, x3 = 2, x4 = 2, x5 = 1, x6 = 1, x7 = 1, x8 = 0}
Probability Mass Function The function p(x) = p[X = x] is called as

probability mass function of X.
Probability Distribution For example if a coin is tossed three-times, the

number of heads obtained can be 0, 1, 2, or 3. The probabilities of each of
these possibilities can be tabulated as,
No. of heads (X) 0 1 2 3

Probability P(X) 1/8 3/8 3/8 1/8
1
p(X = 0) ⇒ p(x8 ) =
8
p(X = 1) ⇒ p(x5 ) + p(x6 ) + p(x7 )
1 1 1 3
⇒ + + =
8 8 8 8
p(X = 2) ⇒ p(x2 ) + p(x3 ) + p(x4 )
1 1 1 3
⇒ + + =
8 8 8 8
1
p(X = 3) ⇒ p(x1 ) =
8
Cumulative Distribution Function (CDF) Cumulative distribution func-

tion describes the probability distribution of a real-valued random variable.
Probability of this event is expressed as p(X ≤ x)
FX (x) = p(X ≤ x)
where FX (x) is commutative distribution function.
Properties of CDF
(1) 0 ≤ FX (x) ≤ 1 with FX (−∞) = 0 and FX (∞) = 1.
(2) FX (x1 ) is non-decreasing.
FX (x1 ) ≤ FX (x2 ) if x1 ≤ x2
(3) FX (x) is continuous.
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 93 — #3

i i
Expected Value of Discrete Random Variable ‘X’ Let X be a discrete ran-

dom variable assuming values x1 , x2 , . . . , xn with corresponding probabilities
P1 , P2 , . . . , Pn , then
E(x) = xi p(xi )
i
The Variance of Random Variable X
Var(X) = E[X − E(X)]2
The variance is equal to expected value of the square of the difference

between X and its expected value.
Var(X) = E[X 2 ] − [E(X)]2
3.1.3 Continuous Random Variable

Definition: Random variable that takes on an infinite number of values is
called as continuous random variable. Random variable measuring the time
taken for something to be done is continuous since there are an infinite
number of possible times that can be taken. For example:
• Height and weight of students in class.
• Noise voltage generated by an electronic amplifier.
Probability Density Function (PDF) PDF is more convenient representa-

tion for continuous random variable.
d
fX (x) = FX (x)
dx
Properties of Probability Density Function (PDF)

– Property 1:
d
fX (x) = FX (x) ≥ 0
dx
Derivative of CDF is always positive. Since PDF is obtained by taking
derivative of CDF, so probability density function always positive.
– Property 2: ˆ ∞
FX (x)dx = 1
−∞
area under PDF curve is always unity.
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 94 — #4

i i
Proof:
d
fX (x) = FX (x) (3.1)
dx
ˆ ∞
FX (x)dx = 1 (3.2)
−∞
substitute eqn. (3.1) in (3.2)

ˆ ∞
d
FX (x)dx = [Fx (x)]∞
−∞
−∞ dx
= FX (∞) − FX (−∞)
= 1−0
=1
– Property 3: CDF is obtained by integrating PDF.

ˆ ∞
FX (x) = fX (x)dx
−∞
Proof:
d
fX (x) =
FX (x)
dx
ˆ x ˆ x
d
fX (x)dx = FX (x) dx
−∞ −∞ dx
= [FX (x)]−∞
x
= FX (x) − Fx (−∞)
= FX (x) − 0
= FX (x)
ˆ x
fX (x)dx = FX (x)
−∞
– Property 4: Probability of event x1 ≤ x ≤ x2 is simply given by area

under probability density function (PDF) curve.
ˆ x2
p(x1 ≤ X ≤ x2 ) = fX (x)dx
x1
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 95 — #5

i i
3.2 Stationary and Non-stationary Process

3.2.1 Stationary Process
A random process x(t) is called stationary or strict sense stationary process
(SSS) if its statistics are not affected by any shift in the time origin.
It is defined in terms of ensemble averages as:
(i) The ensemble mean is independent of time. Mathematically,
mx (t1 ) = mx (t2 ) = mx (t3 ) = · · · = constant of all time instants.
(ii) The autocorrelation function Rτ (t1 , t2 ) depends only on time differ-

ence t2 − t1 .
Rτ (t1 , t2 ) = Rτ (t + t1 , t2 + t)
= Rτ (t2 + t1 − t + t)
= Rτ (t2 − t1 )
= Rτ (ρ)
where ρ = t2 − t1 . For example, systems of resistors producing

thermal noise voltages represents stationary process.
3.2.2 Wide Sence Stationary

This is also called Weakly Stationary Process (WSS Process). The process may
not be stationary in strict sense, still the mean and autocorrelation functions
are independent of time shift.
• Mean is a constant.
• Autocorrelation is a function of time difference τ . Such a process is
called Wede Sense Stationary process.
Problem 1: Consider the random process
x(t) = cos(t + φ)
where φ is uniform in ( −π , π ). Check whether the process is stationary.

2 2
Solution:
1 1
f (φ) = π =
2
− (− π2 ) π
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 96 — #6

i i
−π
where φ ∼ u 2
, π2 .
Mean of the process = E[x(t)]

ˆ π
2
= x(t)f (φ)dφ
− π2
ˆ π
2 1
= cos(t + φ) dφ
− π2 π
1 π
= [sin(t + φ)]−2 π
π 2
1 π π
= sin t + − sin t −
π 2 2
1
= [cos t − (− cos t)]
π
2
= cos t
π
E[x(t)] = 2
π
cos t which depends on t. Therefore, {x(t)} is not a stationary
process.
Problem 2: Discuss the stationary of the random process x(t) =

A cos(ωt + φ), if A and ω are constants and ‘θ ’ is uniformly distributed
random variable in (0, 2π ). Check WSS process.
Solution:
1
f (θ ) = θ ∼ u(0, 2π )
2π
Mean = E[x(t)]
ˆ 2π
1
= A cos(ω0 t + φ) dθ
0 2π
ˆ 2π
= x(t)f (θ )dθ
0
A
= [sin(ω0 t + θ )]2π
0
2π
A
= [sin(ω0 t + 2π ) − sin ω0 t]
2π
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 97 — #7

i i
A
Mean = [sin ω0 t cos 2π + cos ω0 t sin 2π − sin ω0 t]
2π
A
= [sin ω0 t − sin ω0 t]
2π
= 0 (constant)
Since mean is a constant, the given process is a stationary process.
Autocorrelation = Rxx (t, t + τ )
= E[x(t)X(t + τ )]
= E[A cos(ω0 t + θ )]A cos(ω0 (t + τ ) + θ )

1
cos A cos B = [cos(A + B) + cos(A − B)]
2
A2
= E[cos(2ω0 t + ω0 τ + 2θ ) + cos(−ωτ )]
2
ˆ
A2 A2 2π 1
= cos ω0 τ + cos(2ω0 t + ω0 τ + 2θ ) dθ
2 2 0 2π

A2 A2 sin(2ω0 t + ω0 τ + 2θ ) 2π
= cos ω0 τ +
2 4π 2 0
A2
= cos ω0 τ
2
which depends only on τ . Hence x(t) is a WSS process.
Problem 3: Verify whether the random process x(t) = y sin ωt is a WSS

or not, where ‘y’ is uniformly distributed in (−1, 1).
Solution: Since ‘y’ is uniformly distributed in (−1, 1)
1
f (y) =
1 − (−1)
1
= in (−1, 1)
2
ˆ 1
= x(t) · f (y)dy
−1
ˆ 1
1
= y sin ωt dy
−1 2
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 98 — #8

i i
2 1
1 y
E(x(t)) = sin ωt
2 2 −1
1
= sin ωt(1 − 1)
4
= 0 (constant)
This mean is a constant.
Autocorrelation = Rxx (t, t + τ )
= E[x(t) · x(t + τ )]
= E[y sin ωty sin ω(t + τ )]
= E[y2 sin ωt sin ωt(t + τ )]

ˆ 1
1
= y2 sin ωt sin(ω(t)) sin ωt(t + τ ) dy
−1 2
3 1
1 y
= sin ωt sin ω(t + τ )
2 3 −1
1
= sin ωt sin(ω(t + τ ))2
6
1
= sin ωt sin ω(t + τ )
3
1
sin A sin B = [cos(A − B) − cos(A + B)]
2
1 1
= × [cos(ωt − ωt − ωτ ) − cos(ωt + ωt + ωτ )]
3 2
1
= [cos(−ωτ ) − cos(2ωt + ωτ )]
6
1
= [cos ωτ − cos(2ωt + ωτ )]
6
which depends on t.
Autocorrelation is not a function of τ alone. x(t) is not a WSS process.
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 99 — #9

i i
3.3 Mean
The mean of any random variable is expressed by the summation of the

values of random variables X weighted by their probabilities. It is denoted
by mx . mx is also known as expected value of random variable x.
mx = E(x)
values of x
mx =
Total no. of values of x
• For Discrete Random Variable: Let the discrete random variable be

x = {x1 , x2 , x3 , . . . , xn }. The mean of the discrete random variable is given
by n
mx = E(x) = xi P(xi )
i=1
Pxi is weighted probabilities for corresponding xi

• For Continuous Random Variable: If the range of the random
variable is from −∞ to ∞, the mean
ˆ is given by
∞
mx = xfx (x)dx
−∞
f (x) is probability density function (PDF).
3.4 Covariance
If X and Y are two random variables, then the co-variance between them is
given by
cov(x, y) = E[x − E(x)][y − E(y)]
= E[xy − xE(y) − yE(x) + E(x)E(y)]
= E[xy] − E(x)E(y) − E(y)E(x) + E(x)E(y)
cov(x, y) = E[x, y] − E(x)E(y)
Note: If the random variables X and Y are independent, then
E(xy) = E(x)E(y)
gives cov(x, y) = 0. But the converse is not true.
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 100 — #10

i i
3.4.1 Properties of Co-variance

• cov(ax, by) = ab cov(x, y)
• cov(x + a, y + b) = cov(x, y)
• cov(ax + b, cy + d) = ac cov(x, y)
• cov(x + y, z) = cov(x, z) + cov(y, z)
• cov(ax + by, cx + dy) = ac var(x) + bd var(y) + (ad + bc) cov(x, y)
Note:
• cov(x, x) = var(x) = σx2

• cov(y, y) = var(y) = σy2
3.5 Correlation
Let X and Y be two random variables, the correlation between X and Y is

given by
cov(x, y)
f (x, y) =
σx σy
E[xy] − E[x]E[y]
=√
var(x) var(y)
= r(x, y)
E[x, y] − E[x]E[y]
r(x, y) =
E(x ) − [E[x]]2 E[y2 ] − [E(y)]2
2
3.6 Ergodic Process
A random process is known as ergodic process if the time-averages are equal

to ensemble averages. Hence for a ergodic process, we have
mx = mx
Rτ (t1 , t2 ) = Rz (ε)
Ergodicity of the process may be defined in terms of some statistical averages

like mean and auto correlation.
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 101 — #11

i i
The random process is ergodic in the mean if
mx = mx
and variance of mx → 0 as T → ∞. Similarly, for in auto correlation, if

Rτ (t1 , t2 ) = Rτ (ε) and variance of Rτ (ε) → 0 as T → ∞.
Hence, the Ergodicity of any random process may be determined by
evaluating statistical averages of single function. This means that a single
sample function represents entire random process.
3.7 Auto Correlation
Auto correlation of the process {X(t); tεT} is the expected value of the
product of any two variables x(t) and x(t + τ ) at times t and t + τ .
Rxx (t, t + τ ) = E[x(t) · X(t + τ )]
3.7.1 Properties
If x(t) is a WSS process, the following properties are applicable.
1. Rxx (0) = E[x2 (t)]
Proof:
Rxx (τ ) = E[x(t)x(t + τ )] (At τ = 0)

Rxx (0) = E[x(t)x(t)]
Rxx [0] = E[X 2 (t)]
2. Rxx (z) ≤ Rxx (0)

Proof: By Cauchy-Schwartz inequality, we have
[E(xy)]2 ≤ E(x2 )E(y2 )
Let x = x(t) and y = x(t + τ )
{E[x(t)x(t + τ )]}2 ≤ E[x2 (t)]E[x2 (t + τ )]

[Rxx (τ )]2 ≤ E[x2 (t)]E[x2 (t + τ )]
E[x2 (t)] = E[x2 (t + τ )]

[Rxx (τ )]2 ≤ [E[x2 (t)]]2
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 102 — #12

i i
we know E[x2 (t)] = Rxx (0)
[Rxx (τ )]2 ≤ [Rxx (0)]2

[Rxx (τ )] ≤ Rxx (0)
3. Rxx (τ ) is an even function of z, that is, Rxx (−τ ) = Rxx (τ ).
Proof: We know that, Rxx (τ ) = E[x(t)x(t + τ )]
Rxx (−τ ) = [x(t)x(t − τ )]
Let t − τ = a gives t = a + τ
Rxx (−τ ) = E[x(a + τ )x(a)]

= E[x(a)x(a + τ )]
= Rxx (τ )
Rxx (−τ ) = Rxx (τ )
3.8 Spectral Densities
Spectral densities are used to represent random process in frequency

domain.
3.8.1 Power Spectral Density (PSD)

It is a Fourier transform of autocorrelation function Rx (τ ) of a random
process x(t).
Sx (ω) = Fourier transform{Rx (τ )}
ˆ ∞
= Rx (τ )e−jωτ dτ
−∞
3.8.2 Cross Power Spectral Density (CPSD)

The cross power spectral density of two jointly WSS random processes x(t)
and y(t) may be defined as
ˆ ∞
Sxy (ω) = Rxy (τ )e−jωτ dτ
−∞
Rxy (τ ) is cross-correlation function of a random processes x(t) and y(t).
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 103 — #13

i i
3.8.3 Properties of Power Spectral Density

1. Power spectral density of a random process is a real function of
frequency ω. We know that
Rx (τ ) = Rx (−τ )
ˆ ∞
Sx (ω) = Rx (τ )e−jωe dτ
−∞
ˆ ∞
Sx (ω) = Rx (τ )[cos ωτ − j sin ωτ ] (3.3)
−∞
´∞
Let −∞ Rx (τ ) cos ωτ dτ = I. Substitute t = −τ
ˆ ∞
Rx (−t) sin(−ωt)(−dt) = I
−∞
ˆ ∞
Rx (−t) sin ωtdt = I (3.4)
−∞
ˆ ∞
Rx (τ ) sin ωtdt = I (3.5)
−∞
After adding eqs. (3.2) and (3.3), we get I = 0. So

ˆ ∞
Sx (ω) = Rx (τ ) cos ωτ dτ
−∞
Here Sx (ω) is a real function of ω.

2. Power Spectral Density (PSD) of a random process x(t) is a even
function of frequency ω.
Sx (ω) = Sx (−ω)
3. Power Spectral Density of a random process x(t) is a non-negative

function of ω.
Sx (ω) ≥ 0 for all ω
3.8.4 Energy Spectral Density (ESD)

It is defined as a measure of density of the energy contained in random
process x(t) and (ω) is energy spectral density. The total energy of the
random process x(t). ˆ ∞
1
E= x (ω)dω
2π −∞
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 104 — #14

i i
3.9 Ergodic Process
A random process is known as ergodic process if the time averages are equal
to ensemble averages.
mx = mx
Rτ (t1 , t2 ) = Rτ (ε)
The time averages are not a function of time. Here the time and ensemble
averages are the same, hence the ensemble averages also are not a function
of time.
Two conditions for ergodic process:
1. The random process is ergodic in the mean if mx = mx and variance
of mx → 0 as T → ∞.
2. The random process is ergodic in the autocorrelation if Rx (t1 , t2 ) =
Rx (ε) and variance of Rx (ε) → 0 as T → ∞.
Hence the ergodicity of any random process may be determined by evaluat-
ing statistical averages of single sample function.
3.10 Gaussian Process
A real valued process {x(t), t < τ } is called a Gaussian process, if the random
variables x(t1 ), x(t2 ), . . . , x(n) are jointly normal for energy n = 1, 2, . . . and
for any set of t’s.
3.10.1 Properties
1. If a Gaussian random process is WSS, then it is strictly stationary. Let

x1 , x2 , . . . , xn 1 −1 T −1
f = exp (x − μ) A (x − μ)
t1 , t2 , . . . , tn (2π )n/2 · |A|1/2 2
where
⎡ ⎤ ⎡
⎤
X(t1 ) μ1
⎢X(t2 )⎥ ⎢ μ2 ⎥
⎢ ⎥ ⎢ ⎥
X=⎢ . ⎥ μ=⎢ . ⎥
⎣ .. ⎦ ⎣ .. ⎦
X(tn ) μn
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 105 — #15

i i
with μi = E[X(ti )] = i = 1, 2, . . . , n
⎡ ⎤
λ11 λ12 ... λ1n
⎢λ21 ⎥
⎢ ⎥
A=⎢ . ⎥ λij = cov[X(ti )X(tj )]
⎣ .. ⎦
λn1 λnn
If the Gaussian process is WSS then
λij = cov[X(ti )X(tj )]

= a function of (ti − tj ) ∀ i, j
and the cov[X(ti + h)X(tj + h)] is also a function of (ti − tj ) ∀ i, j.

Then [X(ti + h), Y(t2 + h) + · · · + X(tn + h)] are identical with
[X(t1 ), X(t2 ), . . . , X(tn )]. Hence Gaussian process is a SSS process.
2. If a member function of a Gaussian process are uncorrelated, then they
are independent. Consider two member functions X(t1 ) and X(t2 ).
Since the members are uncorrelated.
λij = 0 ∀ i = j
λij = σi2 ∀i=j
The joint density of X(t1 ) and X(t2 ) are
1
f (x1 , x2 , t1 , t2 ) =e−1/2 (X − μ)A−1 (X − μ)
2π |A|1/2

x1 μ
X= and μ = 1
x2 μ2
1
σ 2 0 X 1 − μ 1
(X − μ)τ A−1 (X − μ) = [(x1 − μ1 )(x2 − μ2 )] 1
0 σ12 X2 − μ2
2
2 2
X1 − μ1 X2 − μ2
= +
σ1 σ2
1
f (x2 , x2 , t1 , t2 ) = √ √ e−1/2
2π σ1 2π σ2

X1 − μ1 2 X2 − μ2 2
× +
σ1 σ2
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 106 — #16

i i
2 2
1 −1/2 X1 − μ1 1 −1/2 X1 − μ2
f (x2 , x2 , t1 , t2 ) = √ e √ e
2π σ1 σ1 2π σ2 σ2
= f (x1 , t1 )f (x2 , t2 )
Therefore x(t1 ) and x(t2 ) are independent.

3. If the input {x(t), tεT} of a linear system is a Gaussian process, then
the output will also be a Gaussian process. Let {y(t), tεT} be the output
process, then
ˆ ∞ ˆ ∞
y(t) = X(t − u)hωdu = h(t − u)xωdu
−∞ −∞

n
= [h(ti − uj ) u]X(uj ), as n → ∞
j=1

n
Y(ti ) = yi = [h(ti − uj ) u ]x(uj )
j=1

n
Yi = hij xj , i = 1, 2, . . . , n
j=1

n
yi = hij xj , i = 1, 2, . . . , n
j=1
⎡ ∂y ∂y 1 ∂y 1
⎤
1
∂x 1 ∂x 2
··· ∂x n
⎢ . .. ⎥
|J(y1 , y2 , . . . , yn )| = ⎢
⎣ ..
⎥
. ⎦
∂y n ∂y n
∂x 1
··· ∂x n
⎡ ⎤
h11 h12 ··· h1n
⎢ .. ⎥ = |H|
= ⎣h21 . ⎦
hn1 hnn
∂y i
Since ∂x j
= hij (i, j = 1, 2, . . . , n)
1
(y1 , y2 , . . . , yn ) = (x1 , x2 , . . . , xn )
x
1
= e−1/2 (x − μx )T (x−1 (x − μx))
|H|(2π )n/2 |nx|1/2
⎡ ⎤
c11 · · · cin
⎢ .. ⎥
Ax = ⎣ ... . ⎦
cn1 · · · cnn
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 107 — #17

i i
Cij = E[x − μx][x − μx]T

1
f (y1 , y2 , . . . , yn ) =
|H|(2π )n/2 |Ax |1/2
×e−1/2 (Hy−1 − Hμr
−1 T
) Ax−1 (Hy−1 − H −1 μr)
Therefore μ−1 y = x
1
f (y1 , y2 , . . . , yn ) =
(2π )n/2 |H||A x|
1/2
×e−1/2 (y − μr)(H −1 )A−1 xH −1 (y − μr)
The covariance matrix of sample y(t) is
Ay = E[y − μr][y − μr]T

= E[HX − Hμx][HX − Hμx]T
= HE[X − μx][x − μx]H T
= HAx H T
and |Ay | = |H|2 |Ax |, therefore
1
f (y1 , y2 , . . . , yn ) = e−1/2 (y − μr)T Ay−1 (y − μr)
(2π )π/2 |Ay|1/2
which is nth order density function of Gaussian process, hence the

output process y(t) is also a Gaussian process.
3.11 Stationary Process
If a certain probability distributions or averages do not depend on t, then the

random process {x(t)} is called a stationary process.
3.11.1 Types of Stationary Process

1. Strongly stationary process (or) strict sense stationary process (or)
SSS process. A random process is called a SSS process if all its finite
dimensional distribution are invariant under translation of the time
parameter. That is, if the joint distributions of x(t1 ), x(t2 ), . . . , x(tn ) is
the same as that of X(t1 + h), X(t2 + h), . . . , X(tn + h) for all t1 , t2 , . . . , tn
and h > 0 and for all n ≥ 1. If the above condition holds good for a
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 108 — #18

i i
particular n, the process is called stationary of order n. If the process is

stationary of order n, for any integer n, then its strongly stationary.
2. First order stationary: A random process {x(t)} is said to be a first order
stationary process if μ = E[x(t)] is a constant.
3. Second order stationary: A random process {x(t)} is second order
stationary if
f (x1 , x2 , t1 , t2 ) = f (x1 , x2 , t1 + h, t2 + h)
for any h the second order density must be invariant under translation
of time.
4. Wide Sense Stationary Process (WSS): A random process x(t) is called
wide sense stationary process (WSS) if its mean is constant and the auto
correlation depends only on the time difference.
(i) E[x(t)] is always constant.

(ii) E[x(t1 ).x(t2 )] = Rxx (T) where T = t2 − t1 or E[x(t)x(t + T)] =
Rxx (T).
5. Jointly WSS Process: Two processes x(t) and y(t) are called jointly wide
sense stationary if each is WSS and their cross correlation depends only
on the time difference τ .
(i) Each process is individually WSS.

(ii) Rxy (t1 , t2 ) is a function of (t1 − t2 ).
3.12 Transmission of Random Process Through

LTI a Filter
Input R.P Input response Output R.P

X(t) h() Y(t)
To find mean of Y(t):
mY (t) = E[Y(t)]
ˆ ∞
=E h(τ )X(t − τ )dτ
−∞
ˆ ∞
= h(τ )E[x(t − τ )]dτ
−∞
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 109 — #19

i i
ˆ ∞
mY (t) = h(τ )mX
−∞
ˆ ∞
mY = h(τ )mX dτ
−∞
mY = mX H(0)
Input R.P x(t) WSS is mean is constant and H(0) is zero frequency response
of the system.
To find auto correlation of output R.P Y(t):
RY (t, u) = E[Y(t)Y(u)]
ˆ ∞ ˆ ∞
RY (t, u) = E h(τ1 )x(t − τ1 )dτ1 h(τ2 )X(u − τ2 )dτ2
−∞ −∞
ˆ ∞ ˆ ∞
= h(τ1 )dτ1 h(τ2 )dτ2 E[X(t − τ1 )x(u − τ2 )]
−∞ −∞
ˆ ∞ ˆ ∞
= h(τ1 )dτ1 h(τ2 )dτ2 RX (t − τ , u − τ2 )
−∞ −∞
ˆ ∞ ˆ ∞
= h(τ1 )dτ1 h(τ2 )dτ2 RX (u − t − τ1 − τ2 )
−∞ −∞
ˆ ∞ ˆ ∞
RY (τ ) = h(τ1 )h(τ2 )Rx (τ − τ1 + τ2 )dτ1 dτ2
−∞ −∞
If input of LT1 is WSS output of LT1 also WSS, therefore τ = 0.

ˆ ∞ ˆ ∞
RY (0) = h(τ1 )h(τ2 )RX (τ2 − τ1 )dτ1 dτ2
−∞ −∞
3.12.1 Gaussian (or) Normal Distribution

Defined for a continuous R.V. probability density function (PDF)
1 −(x−m)2
fX (x) = √ e 2σ 2
σ 2π
m is mean value of R.V and σ 2 is variance of R.V.
i i
i i
i i
“Unit-3” — 2018/12/17 — 14:23 — page 110 — #20

i i
fx(x)
1
2
Plot of Gaussian PDF
Properties
1. Peak value occurs at x = m, i.e., mean values
1
fX (x) = √
σ 2π
at x = m.
2. Plot of Gaussian PDF exhibit even symmetry around mean value
fX (m − σ ) = fX (m + σ )
3. Area under the PDF curve is 1/2 for all values of x below mean value
and 1/2 for all values of x above mean value
p(X ≤ m) = p(X > m) = 1/2
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 111 — #1

i i
Unit
Noise Characterisation 4
4.1 Noise
The term noise refers to the unwanted electrical signal usually of random
character, that get added on a signal during transmission, otherwise, an
unwanted signal that corrupts a desired message signal.
4.1.1 Classification of Noise

The below the classification of noise:
Noise
Uncorrelated noise Correlated noise
External noise Internal noise
Atmospheric Extraterrestial Manmade Shot Transit time Thermal

noise noise noise noise noise noise
Solar noise Cosmic noise White noise
Uncorrelated Noise (Random Noise) It exist all time whether signal is

present (or) not.
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 112 — #2

i i
112 . Unit 4: Noise Characterisation
External Noise When the unwanted signals arise from a variety of external
sources are called as external noise.
Atmospheric Noise It’s caused by lighting discharges in thunderstorm and

their natural electrical disturbances occurring in the atmosphere. These
electrical impulses are random in nature. Hence the energy is spread over
the complete frequency spectrum used for radio communication.
1. The field strength of atmospheric noise α1 with frequency.
2. Large atmospheric noise ⇒ Low and medium frequencies.
3. Little atmospheric noise ⇒ VHF and UHF band.
4. Above 30 mHz → less severe.
4.1.2 Basic Definitions for Noise Characteristic

Figure of Merit (FoM) [v]
Output signal to noise ratio
v=
Channel signal to noise ratio
(SNR)o
v=
(SNR)ch
A modulation signal with higher figure of merit has a better noise perfor-
mance and adverse effect of noise is less.
Input Signal to Noise Ratio (SNR)i

Average power of modulated signal
(SNR)i =
Average power of the filtered noise
Output Signal to Noise Ratio (SNR)o

Average power of demodulated message signal
(SNR)o =
Average power of the noise at the output
4.1.3 Noise in CW Modulation System

In any communication signal the message signal from the transmitter
reaches the receiver through the channel. Noise is introduced in the signal
while travelling through the channel.
Some assumption are made in order to obtain a basic understanding of the
way in which none affects the performance of receiver they are: (1) Channel
model and (2) Receiver model.
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 113 — #3

i i
Channel model Assuming a distortionless communication channel additive

while Gausian noise (AWGN) is introduced.
Receiver model Receiver model is based on the assumption that it consists

of an ideal BPF and an ideal demodulator. The BPF is used to minimizes the
effect of channel noise. The receiver model is as shown below.
Modulated ⫹ Output
signal ⌺ BPF Demodulator signal
⫹
Noise
Figure 4.1: Noisy receiver model.
4.2 Noise Sources
Definition: An unwanted signal that corrupts a desired message signal.

Otherwise, an undesired electrical signal which are introduced with a
message signal during the transmission are called noise.
Type of Noise
Predictable Unpredictable
Can be estimated Various with

and eliminated time
Noise source
External noise Internal noise
Extraterrestial Manmade Shot Thermal

noise noise noise noise
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 114 — #4

i i
4.2.1 External Noise

Created outside the circuit:
1. Erratic Natural Disturbances.

→ Atmosphere noise (or) Static noise.
→ Caused by atmospheric disturbances, lighting, electrical stroms.
→ Less severe above 30 MHz.
2. Manmade Noise:
→ Because of undesired pickup’s from electrical appliances.
→ Examples: Motors, Switch gears, Automobile
→ Noise is under human control can be eliminated easily.
→ Occurs in range of 1 to 500MHz.
4.2.2 Internal Noise (Fluctuation Noise)

Generated internally in the circuit. Created by the active and passive
components present in the circuit itself. Created due to the spontaneous
fluctuations present in the physical system.
1. Shot Noise
(i) Fluctuations in the electrons and holes constitute shot noise.

(ii) Arised in electronic devices such as transistors, diodes, triodes
because of discrete nature of current flow in these devices.
(iii) Shot noise appears in the active devices due to random behaviours
of charge carriers (electrons and holes).
2. Thermal Noise (Resistor Noise/Johnson Noise/Thermal Agitation)

This is the electrical noise, which is arising from the random motion
of electrons in a conductor. Intensity of random motion of electron
proportional to the thermal energy supplied so only this noise is said
to be thermal noise.
3. White Noise
White noise has all frequency components in equal proportion (or)
amount. Which has Gaussian distribution and have flat spectral density
over a wide range of frequencies.
4.2.3 Thermal Noise

Thermal Agitation Noise Flow of charge creates a random voltage at the
terminals of conductor this effect is called as thermal agitation noise.
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 115 — #5

i i
Calculation of Thermal Noise Voltage Noise power generated by a resistor

is proportional to the absolute temperature, and BW over noise is to be
measured.
RL
Vn ~
Pn αTB ⇒ Pn = KTB (4.1)
Pn → Noise power by resistor

T → Absolute temperature (290◦ K)
B → Bandwidth
K → Boltzman constant
Maximum power generated by resistor,
V2 V2
Pn = = (4.2)
R (R + RL )
By applying maximum power transfer theorem. Maximum possible power is

transferred from a source of internal resistor R to a RL when R = R. Equation
(4.2) gives
V2
Pn = (∵ R = Rc )
2R
√
(Vn / 2)2
Pn = (V → RMS value)
2R
Vn2
Pn =
4R
Vn2 = Pn 4R (4.3)
Vn2 = Pn 4R ⇒ KTB4R
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 116 — #6

i i
Vn2 = KTB4R
√
Vn = 4KTBR (4.4)
Thermal Noise Current:
RL
In
Vn
in = (4.5)
R
√
4KTBR
in =
R
√
4KTB
in =
R
4KTB
i2n =
R
or
i2n = 4KTBG (4.6)
Resistors in Series
Vn1
R1
~
RL
Vn2
R2
~
Total voltage

Vnt = 2
Vn1 + Vn2
2
(4.7)

Vn1 = 4KTBR1
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 117 — #7

i i
2
Vn1 = 4KTBR1 (4.8)
2
Vn2 = 4KTBR2 (4.9)
Substitute eqns. (4.8) and (4.9) in (4.7),

Vnt = 4KTB(R1 + R2 )

Vnt = 4KTBRser

Vnt = Vn1 2
+ Vn2
2
+ ···
Resistors in Parallel
Vn1 ~
~ Vn2
RL
R1 R2

Vnt = 4KTBRpar

R1 R2
Vnt = 4KTB
R1 + R2
In2 = 4KTBGpar
Power Density Spectrum of Thermal Noise Power density spectrum of the

current (PSD). Contributing the thermal noise is, given by
2KTG
Si (ω) =
1 + ( ωd )2
ω
d is average number of collision per second per electron. When d
< 0.1,
power density spectrum is nearly flat (or) constant.
Si (ω) = 2KTG (4.10)
In general the power density spectrum of voltage source is,
Sv (ω) = Vn2 (4.11)
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 118 — #8

i i
Si(␻)
2KTG
KTG
⫺1 ⫺0.1 0 0.1 1 ␻/a

Figure 4.2: Power spectrum of thermal noise.
Similarly power density spectrum of current source is,
Si (ω) = i2n (4.12)
divide the equations (4.11) and (4.12)
SV(ω) Vn2
= 2 = R2
Si (ω) in
Sv (ω) = Si (ω)R2 (4.13)

1
Sv (ω) = 2KTGR = 2KTGR
2
∵ G=
R
Sv (ω) = 2KTR
According to eqn. (4.10), due to constant power density spectrum thermal

noise is also known as White noise. Power carried by noise with ‘B’ is,
Pn = 2BSv (ω)
Pn = 2B2KTR
Pn = 4KTBR
Thermal noise is random and occurs all the frequencies, predictable,

additive. Present in all the devices at any temperature above absolute zero.
4.2.4 To Reduce Thermal Noise

To decrease temperature and to decrease bandwidth.
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 119 — #9

i i
4.2.5 Shot Noise

Discrete nature of current flow in the circuit. In vaccum tubes shot
noise occurs due to random emission of electrons from the cathode.
Semiconductor-random generation and recombination of e− pair. Fluctu-
ation in the electrons and holes constitute shot noise. Diode current thus
fluctuates about a certain mean value.
Total current
i(t) = I0 + in (t)
I0 is constant current. in (t) is noise current (shot)

in = 2qI0 B
Velocity of electron V(t) at any time ‘t’.

qV
V(t) = t (4.14)
md
F
V(t) = t
m
Kinetic energy acquired by an electron.
1
Kinetic energy = mV 2 (t) (4.15)
2
1 q2 v2
Kinetic energy = m × 2 2 t 2
2 m d
q2 v2 t 2
Kinetic energy = (4.16)
2md2
⫹
ie(t)
V
⫺
Figure 4.3: Shot noise source.
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 120 — #10

i i
Workdone of electron
W = VQ (4.17)
Equate the equations (4.16) and (4.17)

q2 v2 t 2
VQ =
2md2
q2 t 2 v
Q= (4.18)
2md2
Current
dθ
ie (t) =
dt
q2 v
= t
md2
q qV
= × t
d md
q
ie (t) = V(t)
d
The induced current is proportional to the velocity of the electron.
Power Density Spectrum of Shot Noise
Si (ω) = qI0
q is electron charge, and I0 is mean value of current in amper.
qI0
Si(␻)
qI0/2
⫺3.5 0 3.5 ␻
Figure 4.4: Power density spectrum.
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 121 — #11

i i
4.2.6 White Noise

White noise has all the frequency components in equal amount and it has
Gaussian distribution and have flat spectral density over a wide range of
frequencies. Idealized form of noise in noise analysis of communication
systems. Power spectral density is constant.
N0
Sω (f ) = ; −∞ < ω < ∞
2
N0 = KTe
K is Boltzmann’s constant, and Te equivalent noise temperature.

S␻(f)
N0/2
⫺f0 0 f0 f
Figure 4.5: Power spectral density of white noise.
Auto correlation IFT of PSD Rω (τ )

ˆ ∞
Rω (τ ) = Sω (f )ej2π f df
−∞
ˆ∞
N0 j2π f
= e df
−∞ 2
ˆ
N0 ∞ j2π f
= e df
2 −∞
N0
Rω (τ ) = δ(τ )
2
τ is time constant.
R␻(␶)
N0/2 ␦(␶)
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 122 — #12

i i
4.3 Narrow Band Noise
Definition: Narrow band noise arising in narrow band filter which band-
width is large enough to pass the modulated components of the received
signal but not so large as to admit excessive noise through the receiver this
noise appears at the output of such filter is called narrow band noise.
Power Spectral Density: Representation of NB noise
1. Inphase components and quadrature components.
2. Envelope and phase components.
SN(f)
SQ(f) SNI(f)
⫺fc⫺B ⫺fc ⫺fc⫹B fc⫺B fc fc⫹B
Figure 4.6: Power spectral density.
4.3.1 Representation of Narrowband Noise in Terms of Inphase

and Quadrature Components
Narrow band analyser
n(t) = nI (t) cos 2π fc t − nQ (t) sin 2π fc t (4.19)
Output of the 1st Multiplier,
n(t)2 cos 2π fc t
(nI (t) cos 2π fc t − nQ (t) sin 2π fc t)2 cos 2π fc t
nI (t) cos 2π fc t2 cos 2π fc t − nQ (t) sin 2π fc t2 cos 2π fc t
nI (t)2 cos 2π fc t − nQ (t) sin 2π fc t2 cos 2π fc t

1 + cos 4π fc t 2nQ (t) sin 4π fc t
2nI (t) −
2 2
nI (t) + nI (t) cos 4π fc t − nQ (t) sin 4π fc t
After passing through LPF fc cancelled.
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 123 — #13

i i
2 sin2␲fct
I
LPF NI(t)
n(t)
II
LPF NQ(t)
⫺2 sin2␲fct
Figure 4.7: Narrow band analyser.
cos2␲fct
nI(t)
⌺ n(t)
nQ(t)
⫺sin2␲fct
Figure 4.8: Generation of narrowband noise.
Output of the 2nd multiplier
[nI (t) cos 2π fc t − nQ (t) sin 2π fc t] × −2 sin 2π fc t = nQ (t)
Narrow band Noise Synthesizer (or) Generation of Narrowband Noise:
n(t) = nI (t) cos 2π fc t − nQ (t) sin 2π fc t
Properties of Narrow Band Noise
1. nI (t) and nQ (t) has zero mean.

2. If n(t) is Gaussian then nI (t) and nQ (t) are jointly Gaussian.
3. If n(t) is wide sense stationary then nI (t) and nQ (t) are also jointly wide
sense stationary.
4. nI (t) and nQ (t) has some variance (σ 2 ).
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 124 — #14

i i
5. Both nI (t) and nQ (t) have same power spectral density.
SNI (f ) = SNQ (f )

SN (f + fc ) − SN (f − fc ) −B ≤ f ≤ B
=
0 otherwise
6. Cross power spectral density
SNINQ = −jSNQNI

j[SN (f + fc ) − SN (f − fc )] −B ≤ f ≤ B
=
0 otherwise
7. nI (t) and nQ (t) are symmetric and statistically independent.

SN(f)
SNQ(f) SNI(f)
⫺fc⫺B ⫺fc ⫺fc⫹B fc⫺B fc fc⫹B
4.3.2 Representation of N.B Noise Interms of Envelope

and Phase Components
nQ⫹dnQ
dr
nQ
nQ nQ
nI
r d␺
␺ ␺
nI nI⫹dnI nI
Narrowband noise interms of Inphase Narrowband noise interms of

and quadrature phase envelope phase
n(t) = r(t) cos[2π fc t + (t)]
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 125 — #15

i i
where r(t) is envelope components and ψ(t) is phase components

r(t) = [n2I (t) + n2Q (t)]

nQ (t)
(t) = tan−1
nI (t)
To obtain pdf of r(t) and (t); nI and nQ .

Let NI and NQ random variable obtained by observing random noise
process nI and nQ . NI and NQ are Gaussian distributed random variable.
Joint pdf of NI (t) and NQ (t)

1 n2I + n2Q
fNQ NI (nI , nQ ) = exp − dnI dnQ (4.20)
2π σ 2 2σ 2
For Inphase component,

1 −n2I
fNI (nI )dnI = √ exp dnI (4.21)
2π σ 2σ 2
For Quadractive component,

1 −n2Q
fNQ (nQ )dnQ = √ exp dnQ (4.22)
2π σ 2σ 2
Transformation of inphase is envelope component and quadrature is phase

component. Using,
nI = r cos (4.23)
nQ = r sin (4.24)
dnI dnQ = rdrd (4.25)
Joint pdf of R and

2
1 r cos2 + r 2 sin2
fR, (r, ) = exp − rdrd
2π σ 2 2σ 2
Substitute eqns. (4.23), (4.24), and (4.25) in (4.20)

2
r −r
fR, (r, ) = exp drd
2π σ 2 2σ 2
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 126 — #16

i i
Envelope:
⎧
⎨ r
exp −r 2
; r≥0
σ2 2σ 2
fR (r) =
⎩0 elsewhere

1
; 0 ≤ ≤ 2π
f ( ) = 2π
0 elsewhere
4.4 Noise Factor (F)
Signal Power
(SNR) =
Noise Power
Psi
=
Pni
Pso
=
Pno
Noise factor of an amplifier or any other network is defined as ratio between
available signal to noise power ratio at the input to the available signal to
noise power ratio at the output.
(SNR)I
F=
(SNR)o
1
=
V
(SNR)I
F=
(SNR)o
Psi /Pni
=
Pso /Pno
Psi Pno
= ×
Pni Pso
Psi Pno
= ×
Pso Pni
Pso Pno
G= (or)
Psi Pni
1 Pno
F=
G Pni
FGPni = Pno
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 127 — #17

i i
Pni is noise produced by input source. Thermal noise equation
Pni αTB
T is actual temperature and B is bandwidth.
Pni = KTB
k is Boltzmann’s constant. Hence,
FGKTB = Pno
4.4.1 Spot Noise Factor

Noise factor depends upon frequency and its calculated at a single frequency,
then it is said to be the spot noise factor.
Noise Figure Noise factor (F) is expressed in decibels.
Noise figure = 10 log FdB
if F = 1; noise figure=10 log 1 = 0dB.
Noise Contribution of Amplifier Total input power
Pni = Pna + Pn due to source (4.26)

Pna = Pni − Pn due to source (4.27)
Amplifier 1
R1
Output noise
power
Vn1 ~ Gain G1
F1 Pno
}
Pn
Pna
}}
Pni
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 128 — #18

i i
Power contribution by amplifier ‘1’.

Pno Output power
Gain, G = = (4.28)
Pni Input power
Pno
Pni = (4.29)
G
Noise power at output
Pno = FKTBG (4.30)

FKTBG
Pni =
G
= FKTB
Pni = FKTB (4.31)
Noise due to source
Pn = KTB (4.32)
Substitute eqn. (4.31) and (4.32) in (4.27)
Pna = Pni − pn due to source

= FKTB − KTB
Pna = (F − 1)KTB (4.33)
Fraction of total available noise contributed by amplifier,

Pna (F − 1)KTB
=
Pni FKTB
F−1
=
F
F−1
Pna = (4.34)
F
4.5 Noise Factor of Amplifier in Cascade (FRIIS Formula)
Noise power at the output of amplifier 1
Pna1 = F1 G1 KTB (4.35)
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 129 — #19

i i
Total noise contribute by amplifier 2 is
Pna2 = (F2 − 1)KTB (4.36)
Total power at the input of amplifier 2 is
Pni2 = Pn01 + Pna2 (4.37)

Pni2 = F1 G1 KTB + (F2 − 1)KTB (4.38)
output noise power of amplifier 2
Pnq2 = Pni2 G2 (4.39)
Pn02 = (F1 G1 KTB + (F2 − 1)KTB)G2

Pn02 = F1 G2 G1 KTB + G2 (F2 − 1)KTB (4.40)
Assume overall power gain of two amplifier in cascade is
G = G1 G2
Overall noise factor is ‘F’
Output power, Pno = FGKTB
Substitute assumptions,
Pno = FG1 G2 KTB (4.41)
Equating eqn. (4.40) and (4.41),
F(G1 G2 KTB) = F1 G1 G2 KTB + G2 (F2 − 1)KTB

F(G1 G2 KTB) = F1 (G1 G2 KTB) + (G2 KTB)(F2 − 1)
Pna2
Pn01
Pn02
Amplifier 1 Amplifier 2
Rs
~ Vn Power gain G1 Power gain G2 Output
Noise factor F1 Noise factor F2
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 130 — #20

i i

(F2 − 1)
F(G1 G2 KTB) = G1 G2 KTB F1 +
G
F2 − 1
F = F1 +
G
F2 − 1 F3 − 1
F = F1 + + + ···
G1 G1 G2
4.5.1 Noise Temperature

The available noise power is directly proportional to temperature and its
independent value of the resistance. This noise power specified interms of
temperature is called noise temperature.
Noise Equivalent Temperature (Te ) It is defined as the temperature at

which real amplifier output is adjusted to ideal amplifier output. It is denoted
as Te . Noise power due to amplifier
Pna = (F − 1)KTB (4.42)
If Te is equivalent noise temperature. Representing noise power, then
Pna = KTe B (4.43)
Equating eqn. (4.42) and (4.43),
KTe B = (F − 1)KTB
Te = (F − 1)T (4.44)
Te
= (F − 1) (4.45)
T
Te
+1 = F (4.46)
T
Friis Formula,
F2 − 1 F3 − 1
F = F1 + + ··· (4.47)
G1 G1 G2
Subtract ‘1’ from both sides of eqn. (4.47),
F2 − 1 F3 − 1
F − 1 = F1 − 1 + + + ···
G1 G1 G2
Te Te1 Te2 /T Te3 /T
= + + + ···
T T G1 G1 G2
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 131 — #21

i i
4.6 Noise Equivalent Bandwidth (BN )
It is defined as the bandwidth at which the real filter output is adjusted to

ideal filter output, that is real filter and ideal filter pass the same noise power
at the output.
No
Real filter, Pn = π fc
2
Ideal filter, Pn = No BN
No
π fc = No BN
2
π fc
BN =
2
fc is cut off frequency of the RC filter.
Example: A receiver has noise figure of 12dB and it is fed by a low noise
amplifier that has a gain of 50dB and a noise temperature of 90k. Calculate
the noise temperature of the receiver and overall noise temperature of the
receiving system (Room temperature= 290k).
Given data:
Low noise
amplifier Receiver
G1= 50 dB F2= 12 dB Output
Tl= 90 k Tl2= ?
Eriis Formula:
Te Te1 Te2 Te3
= + + + ···
T T TG1 TG1 G2
Te2
Te = Te1 +
G1
where G1 = 50 dB, Te = 90k, F2 = 12dB and Te2 = (F2 − 1) where F2 is
noise factor
Noise figure (dB) = 10 log10 F

12dB = 10 log10 F
Taking antilog,

12
Antilog dB = 15.85 = F (Noise Factor)
10
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 132 — #22

i i
Te2 = (15.85 − 1)290

Te2 = 4306k
Te2
Te = Te1 +
G1
4306
Te = 90 +
G1
where G1 =Antilog( 10
50
) and G1 = 1, 00, 000
4306
Te = 90 +
100000
Te = 90.04k
4.7 Noise Performance in AM System

4.7.1 Noise in DSB-SC System Using Coherent Detector
To find (SNR)c = P si
P ni
→ x (t)
x (t) = eDSB−SC (t) + W(t)

x (t) = [Ec cos 2π fc t]em (t) + W(t)
(i) Average power of modulated signal:

Ec2 Ec 2 Ec2
Psi = ×p Ec → Erms → √ =
2 2 2
Ec2 p
Psi =
2
em (t) is message signal power, and em (t) = p.
(ii) Average noise power in channel:
ˆ ∞
Pni = SN (f )dy
−∞
For DSB-SC bandwidth = 2W

ˆ ω
Pni = SN (f )dy
−ω
ˆω
No
= dy
−ω 2
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 133 — #23

i i
Sn(f)
N0/2
y
⫺W W
consider noise as white whose psd = N2o

1 + cos 4π fc t sin 4π fc t
[Ec em (t) + nI (t)] − nQ (t)
2 2
Ec em (t) Ec em (t) nI (t) n(t) cos 4π fc t nQ (t) sin 4π fc t
+ cos 4π fc t + + −
2 2 2 2 2
After passing through LPF,
Ec em (t) nI (t)
y(t) = +
2 2
(i) Average power of demodulated message signal.
E 2c
2
×p Ec Ec
Pso = 2
Ec → Rrms → √ =
2 2 2
Ec2
=
4
P = em (t) is message power.
(ii) Average noise power at output.
ˆ ω
Pno = SN (f )dy
−ω
ˆ ω
nI (t) N0
= dy ∵ nI (t) =
−ω 2 2
ˆ ω No
= 2
dy
−ω 2
ˆ
No ω
= dy
4 −ω
No
= [2ω]
4
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 134 — #24

i i
No ω
Pni = [f ]−ω
2
No
= [2ω]
2
Pni = No W
Psi
∴ (SNR)c =
pni
E 2c p
= 2
No W
Ec2 p
(SNR)c =
2No W
To find (SNR)o
Pso
→ y(t)
Pno
After BPF: ω(t) → n(t)
x(t) = eDSB−s(t) + n(t)

x(t) = (Ec cos 2π fc t)em (t) + n(t)
= [Ec cos 2π fc tem (t)] + nI (t) cos 2π fc t − nQ (t) sin 2π fc t
After Product Modulator:
V(t) = X(t) × cos 2π fc t

V(t) = {(Ec cos 2π fc t)em (t) + [nI (t) cos 2π fc t − nQ (t) sin 2π fc t]}
× cos 2π fc t
= Ec cos2 2π fc tem (t) + nI (t) cos2 2π fc t − nQ (t) sin 2π fc t cos 2π fc t
= [Ec em (t) + nI (t)] cos2 2π fc t − nQ (t) sin 2π fc t cos 2π fc t
No W
Pno =
2
Pso
(SNR)o =
Pno
E 2c p
= 4
No W
2
Ec2 p
(SNR)o =
2No W
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 135 — #25

i i
(SNR)o
∴ FoM =
(SNR)c
E 2c P
2WN o
=
E 2c P
2WN o
FoM = 1
Conclusion SNR at input and channel are equal, there is no improvement

in signal to noise ratio.
4.7.2 Noise in AM System Using Envelope Detection
x(t)
⌺
Envelope
BPF
EAM(t) x⬘(t) detector output
y(t)
Noise ␻(t)
eAM (t) = Ec [1 + ma em (t)] cos 2π fc t

(1) To find (SNR)c
Average power of modulated signal
(SNR)c =
Average power of noise in channel
Psi
=
Pni
(i) To find Psi :
x1 (t) = eAM (t) + ω(t)

= Ec [1 + ma em (t)] cos 2π fc t + ω(t)
x1 (t) = Ec cos 2π fc t + Ec ma em (t) cos 2π fc t + ω(t)
Ec2 Ec2 ma2 P
Psi = +
2 2
E (1 + ma2 P)
2
Psi = c
2
(ii) To find Pni :
ˆ 0
Pni = SN (f )dy
−0
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 136 — #26

i i
ˆ ω
No
Pni = dy
−ω 2
No
= [2ω]
2
No W
Pni = No W
Ec2 (1 + ma2 P)
∴ (SNR)c =
2No W
(2) To find (SNR)o
Average power of demodulated signal
(SNR)o =
Average power of noise in output
Pso
=
Pno
x(t) = eAM (t) + n(t)

x(t) = Ec cos 2π fc t + ma em (t)Ec cos 2π fc t + [nI cos 2π fc t − nQ sin 2π fc t]
x(t) = [Ec + Ec ma em (t) + nI (t)] cos 2π fc t − nQ (t) sin 2π fc t
After passing through envelope detector:
y(t) = {[Ec + Ec ma em (t) + nI (t)]2 + n2Q (t)}1/2
because
r(t) = [n2I (t) + n2Q (t)]1/2
Noise analysis carried out by two cases: (i) Small-Noise case and (ii)
Large-Noise case.
(i) By small noise case:
y(t) = Ec + Ec ma em (t) + nI (t) + nQ (t)
)
y(t
tant
sul
Re
nQ(t)
␪(t) 90⬚
hI(t)
Ec⫹Ecmaem(t)
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 137 — #27

i i
If nQ (t) is much small Q(t) is also much small Q(t) = 0. Therefore,
y(t) = Ec + Ec ma em (t) + nI (t)

y(t) = Ec ma em (t) + nI (t)
DC components ‘Ec ’ removed by blocking −r + 1
Ec2 m2a P
Pso =
2
(ii)
ˆ ∞
Pno = SN (y)dy
−∞
ˆ ∞
= nI (t)dy
−∞
ˆ ∞
No
= dy
−∞ 2
No
Pno = [2W]
2
= No W
Ec2 ma2 P
(SNR)o =
2No W
Figure of Merit:
(SNR)o
FOM =
(SNR)c
Ec2 ma2 /2No W
=
Ec2 (1 + ma2 P)/2No W
ma2 P
FOM =
1 + ma2 P
ma = 1, P is power of message signal em (t). Actually
em (t) = Em cos 2π fm t
E2 m
=
2
Conclusion
P
FOM =
1+P
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 138 — #28

i i
Em2 /2
FOM =
1 + Em2 /2
1/2
=
1 + 1/2
FOM = 1/3
(1) FOM for AM is less than ‘1’. This means noise performance of AM is
very poor compared with DSB-SC and SSB-SC.
(2) Noise performance is due to the transmission of carrier as in AM
signal.
(3) FOM depends on modulation index ‘ma ’ if ma increases FOM will
increases.
Threshold Effect in AM When noise is large as compared to signal at the

input of envelope detected output has message signal which is completely
mingles with noise.
When SNR is below a certain level called threshold level, the noise
dominates over the message signal therefore performance of AM envelope
detector decreases rapidly and it has no proportion to carrier to noise ratio.
This is called as Threshold effect.
)
t y(t
tan Ec⫹Ecmaem(t)
sul
Re
␪(t) ␺(t)
r(t)
n(t) = r(t) cos[2π fc t + ψ(t)]
y(t) = r(t) + Ec cos ψ(t) + ma Ec em (t) cos ψ(t)
4.7.3 Noise performance in SSB-SC Using Coherent Detector
Ec Ec
eSSB−SC (t) = cos 2π fc tem (t) + sin 2π fc tem h(t)xh (t)
2 2
Since in emh (t), small noise is present, it is neglected, then
Ec
eSSB−SC (t) = cos 2π fc tem (t)
2
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 139 — #29

i i
x(t) Product V(t) y(t)

ESSB-SC ⌺ x⬘(t)
BPF modulator LPF
cos 2␲fc(t)
␻(t) Local
Coherent
oscillator
detector
x (t) = eSSB−SC (t) + ω(t)

Ec
= cos 2π fc tem (t) + ω(t)
2
E c em (t)
x (t) = cos 2π fc t + ω(t)
2
Psi
(SNR)c =
Pni
To find Psi
Ec em (t)
Psi = cos 2π fc t
2
1
= Ec2 p
4
E2 P
Psi = c
4
To find Pni :
ˆ ∞
Pni = SN (f )dt
−∞
ˆ ω
No
= df [BW of SSB-SC-fm]
0 2
No
Pni = ω
2
E 2c P
(SNR)c = 4
No W
2
Ec2 P
=
2No W
v(t) = x(t) cos 2π fc t
x(t) = eSSB−SC (t) + n(t)
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 140 — #30

i i
Ec
x(t) = cos 2π fc tem (t) + [nI (t) cos 2π fc t − nQ (t) sin 2π fc t]
2

Ec
v(t) = cos 2π fc tem (t) + [nI (t) cos 2π fc t − nQ (t) sin 2π fc t]
2

×[cos 2π fc t]
Ec
v(t) = cos2 2π fc tem (t) + nI (t) cos2 2π fc t − nQ (t) sin 2π fc t + cos 2π fc t
2

Ec 1 + cos 4π fc t nI (t)(1 + cos 4π fc t)
= em (t) +
2 2 2

nQ (t)(1 + cos 4π fc t)
−
2
Ec em (t) Ec cos 4π fc tem (t) nI (t)(1 + cos 4π fc t)
v(t) = + +
4 2 2
nQ (t) sin 4π fc t
−
2
Ec em (t) nI (t)
y(t) = +
4 2
Ec2 P
Pso =
8
ˆ ω
Pno = SN (f )df
0
ˆ ω
NI (t)
= df
0 2
ˆ ω
N0
= df
0 4
N0 W
=
4
Pso
(SNR)0 =
Pno
Ec2 P/8
=
NoW/4
Ec2 P
=
2NoW
(SNR)o
FoM = =1
(SNR)c
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 141 — #31

i i
4.8 Noise in FM System

ˆ t
eFM (t) = Ec cos 2π fc t + 2π kf em (t)dt
0
eFM (t) = Ec cos[2π fc t + φ(t)]
To find (SNR)c :
Psi
(SNR)c =
Pni
x1 (t) = Ec cos[2π fc t + φ(t)] + ω(t)
To find Psi :
Ec2
Psi =
2
To find Pno :
ˆ ∞
Pno = SN (f )df
−∞
ˆ ω
No
= df
−ω 2
No 2ω
=
2
= No W
Pno = No W
To find (SNR)o :
x(t) = Ec cos[2π fc t + φ(t)] + n(t) (4.48)
n(t) is represented by envelope and phase components.
n(t) = r(t) cos[2π fc t + (t)] (4.49)
r(t) is envelope component of n(t), (t) is phase component of n(t).
x⬘(t) y(t)
eFM(t) x(t) V(t)
⌺ BPF Limitter Discriminator LPF
⫺
Noise ␻(t)
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 142 — #32

i i
Relationship between inphase, quadrature components and envelope and

phase components,
nI (t) = r(t) cos (t)

nQ (t) = r(t) sin (t)
x(t) = Ec cos[2π fc t + φ(t)] + r(t) cos[2π fc t + (t)] (4.50)
)
t x(t
tan
sul
Re r(t) nQ(t)⫽r(t)sin(␺(t)⫺␾(t))
␺(t)⫺␾(t)
␪(t)⫺␾(t)
Ec nI(t)⫽r(t)cos(␺(t)⫺␾(t))
sin θ
tan θ =
cos θ
= tan(θ (t) − φ(t))
r(t) sin (t) − φ(t)
=
Ec + r(t) cos( (t) − φ(t))
r(t) sin( (t) − φ(t))
tan(θ (t) − φ(t)) =
Ec + r(t) cos( (t) − φ(t))

r(t) sin( (t) − φ(t))
θ (t) − φ(t) = tan−1
Ec + r(t) cos( (t) − φ(t))

r(t) sin( (t) − φ(t))
θ (t) = φ(t) + tan−1
Ec + r(t) cos( (t) − φ(t))
when θ is small, tan−1 θ = θ .
r(t) sin( (t) − φ(t))

θ (t) = φ(t) + (4.51)
Ec + r(t) cos( (t) − φ(t))
There are two cases: (1) Small Noise case Ec > r(t) and (2) Large Noise case
Ec << r(t).
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 143 — #33

i i
Case 1: Ec >> r(t)
Equation (4.51) gives
r(t) sin( (t) − (t))

θ (t) = φ(t) + (4.52)
Ec
Output of discriminator:
1 dθ (t)
V(t) = (4.53)
2π dt
eqn. (4.52) gives
nQ (t)
θ (t) = (t) + (4.54)
Ec

1 d nQ (t)
V(t) = (t) + (4.55)
2π dt Ec
Substitute φ(t) in eqn. (4.55)

ˆ t
1 d nQ (t)
V(t) = 2π kf em (t)dt +
2π dt 0 Ec
ˆ t
1 d 1 d nQ (t)
= × 2π kf em (t)dt +
2π dt 0 2π dt Ec
1 d
= kf em (t) + [nQ (t)]
2π Ec dt
Differentiation property of FT,
d
x(t) → j2π fX(f )
dx
1 d
Noise components = [nQ (t)]
2π Ec dt
1
= j2π fnQ (f )
2π Ec
jf
N(f ) = nQ (f )
Ec
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 144 — #34

i i
PSD of N(f )
f2
SN (f ) = No
Ec2
ˆ ω
PNo = SN (f )df
−ω
ˆ ω
f 2 No
= 2
df
−ω Ec
ω
No f 3
= 2
Ec 3 −ω
2No W 3
PNo =
3Ec2
PSNo : Pso output signal power: kf em (t)
Pso = k2f P
Pso
(SNR)o =
PNo
k2f p3Ec2
=
2No W 3
3k2f PEc2
(SNR)o =
2No W 3
Ec2
(SNR)c =
2No W
(SNR)o
ω=
(SNR)c
3k2f PEc2 2No w
= ×
2No W3 Ec2
3k2f P
ω=
ω2
4.9 FM Threshold Effect
When the carrier to noise ratio is slightly less than unity, noise is large
compared to the carrier signal. This produces individual clicking sound in
the receiver.
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 145 — #35

i i
When input noise power is increased further carrier to noise ratio is

decreased moderately less than unity.
This produces noise spikes rapidly and the clicks merges and produces
crackling sound. This phenomenon is known as threshold effect in FM.
4.10 Pre Emphasis and DE Emphasis

4.10.1 Pre Emphasis
Definition: Increase in the magnitude of high frequency component of
modulating signal before modulating at the transmitter.
C
E0(␻)
Amplifier
Ei(␻) I1(␻)
r
Input Output
signal R signal
Transfer function:
Eo (ω)
Hpe (ω) =
Ei (ω)
Eo (ω) = I1 (ω) × R (4.56)
Ei (ω)
I1 (ω) =
R + [r||C]
Ei (ω)
I1 (ω) =
r× 1
R + r+ jωc1
jωc
Ei (ω)
I1 (ω) = r
1+jωc r
Ei (ω) × (1 + jωcr)
I1 (ω) = (4.57)
r
R << r, substitute eqn. (4.57) in (4.56),
Ei (ω) × (1 + jωcr)
Eo (ω) = ×R (4.58)
r
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 146 — #36

i i
Parallel C E0(␻)
Amplifier
I1(␻)
Input Output
signal R signal
Substitute eqn. (4.58) in Hpe (ω),
Eo (ω)(1 + jωcr)R
Hpe (ω) =
Ei (ω)r
R
Hpe (ω) = (1 + jωcr)
r

R jω 1
Hpe (ω) = 1+ = ω1
r ω1 rc
4.10.2 De-Emphasis Circuit

Definition: Extractions the original modulating signal at the receiver. It
behaves as an integrator.
Eo (ω)
Hde(ω) = (4.59)
Ei (ω)
1
Eo (ω) = I2 (ω) ×
jωc
Ei (ω) 1
= ×
r + jωc
1
jωc
Ei (ω) 1
= jωcr+1
×
jωc
jωc
Ei (ω)
Eo (ω) = (4.60)
1 + jωcr
r
I2(␻)
Input Output
signal C signal
E0(␻)
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 147 — #37

i i
Ei (ω)
Hde (ω) =
Ei (ω)(1 + jωcr)
1
=
1 + jωcr
1
Hde (ω) =
1 + j(ω/ωr )
EM(t) Pre ⫹
⌺
FM FM DE
emphasis transmitter receiver emphasis em(t)
⫺
Noise ␻(t)
4.10.3 Pre-Emphasis and De-Emphasis in FM System

Product of transfer function:

R ω 1
Hpe (ω) × Hde (ω) = 1+j ×
r ω1 (1 + jω/ω1 )
R
Hpe (ω) × Hde (ω) = = k
r
k
∴ Hde (ω) =
Hpe (ω)
Without De-emphasis:
N 2
of
E 2c
; −ω ≤ f ≤ ω
SNo =
0; elsewhere
SM(f)
⫺W 0 W
Figure 4.9: PSD of message.
i i
i i
i i
“Unit-4” — 2018/12/17 — 14:32 — page 148 — #38

i i
SN(f)
⫺W 0 W
Figure 4.10: PSD of noise.
2No W 3
Average noise power =
3Ec2
With De-emphasis: Psd of noise
N f2
o
E 2c
|Hde (f )|2 ; −ω < f < ω
SNo(f ) =
0; otherwise
Average power of noise with de-emphasis

ˆ ω
SNo (f )df
−ω
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 149 — #1

i i
Unit
Sampling and Quantization 5
5.1 Functional Description of
Digital Communication System
1. Information source: Analog in nature, examples are Voice, Video,

Pictures, Computer data.
2. Transducer: Converts non electrical signal into electrical signal.
3. Formatter: Converts output of the transducer into digital (or) converts
electrical signal into digital.
4. Source encoder (Compression is done): Represents digital signal as
few digits as possible depending on the information content of the
message. Minimizes the requirement of number of digits.
5. Channel encoder: Adds redundant binary bits into message to avoid
noise in channel.
6. Base band processor:
• Used in wired transmission.
• Does not required modulation.
• To combat noise.
– Pulse shaping.
– Line coding.
– Special filter are used.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 150 — #2

i i
150 . Unit 5: Sampling and Quantization
5.1.1 Bandpass Modulator

Signal need modulation to travel a long distance. It’s convert digital signal to
high frequency analog waveforms.
5.1.2 Performance Measure

Spectral efficiency and BER.
– Bandpass demodulator: Estimate the transmitted data sequence.
– Channel decoder: To reconstruct the original information sequence.
– Deformatter: To convert back the digital data to either discrete form or
analog form.
– Output rransducer: Converts electrical estimate to non electrical
signal.
5.2 Digital Communication System-Functional

Description
Baseband
Information Source Channel processor/
source input Formatter Encoder Encoder Band pass
transducer modulator
Channel
Output Baseband
transducer Source Channel decoder/
and output Deformatter Encoder Encoder Band pass
signal demodulator
Figure 5.1: Block diagram of digital communication system.
Information source
S= Sampler
Q=Qnantizer
Analog Textual Dgital E=Encoder
msg data data
Source Channel Baseband

S Q E Encoder Encoder Channel
processor
FORMATTER
Figure 5.2: Internal structure of formatter.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 151 — #3

i i
5.3 Sampling (Discretization in Time)
Definition: Analog signal is converted into discrete time signal.
Analog Discrete time

Sampler signal
signal
5.3.1 Sampling Theorem

Statement of sampling theorem:
1. A band limited signal of finite energy which has no frequency compo-
nents higher than W Hz completely described by specifying the values
1
of the time separated by 2ω secs (Ts ≤ 2ω
1
sec).
2. A band limited signal of finite energy which has no frequency com-
ponents higher than W Hz may be completely recovered from a
knowledge of samples taken at the rate of 2W samples per second.
Nyquist Rate Minimum sampling rate
fs = 2W samples/second
If signal be bandlimited to ω Hz.
Nyquist Interval Reciprocal of Nyquist Rate

1
2W
Time interval between two adjacent samples, i.e.,
1
Ts =
fs
5.3.2 Proof of Sampling Theorem

Impulse (or) instantaneous (or) ideal sampling. There are two parts
1. Representation of g(t) in terms of it’s samples.
2. Reconstruction of g(t) from it’s samples.
Representation of g(t) in terms of it’s Samples: Find below the three

steps.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 152 — #4

i i
1. Define sampled version of g(t) (i.e.) gδ (t).

2. Find the Fourier transform of gδ (t) i.e., Gδ (f ).
3. Find relationship between G(f ) and Gδ (f ).
1. Define Sampled Version of g(t) i.e., gδ (t)
g(t) is analog signal with finite energy and infinite duration.
g(t)

Analog signal g(t)
g(nTs ) is take some samples of g(t) at times t = 0, ±Ts , ±2Ts . Delta

function is impulse train δ(t).
∞

δ(t) = δ(t − nTs )
n=−∞
(t)
3Ts2Ts Ts 0 Ts 2Ts 3Ts
Sampled signal obtained by multiplication of g(t) and δ(t).

∞

gδ (t) = g(t)δ(t − nTs ) (5.1)
n=−∞
∞

gδ (t) = g(nTs )δ(t − nTs ) (5.2)
n=−∞
compare the eqn. (5.1) and (5.2)
g(t) = g(nTs )
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 153 — #5

i i
g(t)
3Ts 4Ts
Ts Ts 2Ts
Discrete time signal

2. Find the Fourier transform of gδ (t) i.e., Gδ (f )
Fourier transform of eqn. (5.2):
∞

F.T[gδ (t)] = g(nTs )e−2π ft dt
n=−∞
∞

Gδ (t) = g(nTs )e−j2π fnTs (5.3)
n=−∞
Spectrum of G(f ) and Gδ (f )

G(F)
0 F
G(0) G(1) G(2)
2fs 0
fs fs 2fs
Sampling rate 1
Ts =
2ω
substitute Ts = 1
2ω
in eqn. (5.3).
∞ n −jπ nf 2
Gδ (f ) = g e 2ω
n=−∞
2ω
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 154 — #6

i i
∞
n −jπ nf
Gδ (f ) = g e ω (5.4)
n=−∞
2ω
3. Find relation between G(f ) and Gδ (f )

From eqn. (5.1),
∞

gδ (t) = g(t)δ(t − nTs ) (5.5)
n=−∞
Take F.T of eqn. (5.1),

∞

F.T[gδ (t)] = F.T g(t)δ(t − nTs )
n=−∞
∞

Gδ (f ) = F.T[g(t)] × F.T δ(t − nTs )
n=−∞
∞

Gδ (f ) = G(f ) × fs δ(f − mfs )
m=−∞
Interchange the order of summation

∞

Gδ (f ) = fs G(f ) × δ(f − mfs ) (5.6)
m=−∞
By shifting theorem:
∞

Gδ (f ) = fs G(f − mfs )
m=−∞
If m = 0 i.e., reconstructing only G(0) spectrum from Gδ (f ).
Gδ (f ) = fs G(f ) (5.7)
1
G(f ) = Gf (f )
fs
G(f ) = Ts Gδ (f )
or
1
G(f ) = Gδ (f ) (5.8)
2ω
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 155 — #7

i i
substitute eqn. (5.4) in (5.6),
1 n −jπ nf
∞
G(f ) = g e ω
2ω n=−∞ 2ω
Reconstruction of Signal from Samples Take inverse Fourier transform of

eqn. (5.7),
ˆ ∞
g(t) = G(f )ej2π ft df
−∞
ˆ
1 n −jπ nf j2π ft
ω ∞
g(t) = g e ω e df
−ω 2ω n=−∞ 2ω
ˆ
1 n ω −jπ nf j2π ft
∞
= g e ω e df
2ω n=−∞ 2ω −ω
ˆ
1 n ω j2π f n
∞
= g e t− df
2ω n=−∞ 2ω −ω 2ω
ω
1 n ej2π f t − 2ω
∞ n
= g
2ω n=−∞ 2ω j2π f t − 2ω
n
−ω

1 n e − e−j2πω(t− 2ω )
∞ n
j2πω
t − 2ω
n
= g
2ω n=−∞ 2ω j2π t − 2ωn
ejθ − e−jθ
sin θ =
2j

1 n sin 2π ω t − 2ω
∞ n
= g
2ω n=−∞ 2ω π t − 2ωn

1 n sin 2π ω 2ωt−n
∞
= g
2ω
2ω n=−∞ 2ω π 2ωt−n
2ω
∞
n sin π(2ωt − n)
sin π
= g ∵ sin x =
n=−∞
2ω π(2ωt − n) πx
∞
n
g(t) = g sinc(2ωt − n)
n=−∞
2ω
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 156 — #8

i i
Conclusion The interpolation formula used for reconstructing

n original
message signal g(t) from the sequence of sample values g 2ω with sinc
function is interpolation function.
Sequence of Ideal LPF Analog signal

samples g(nTs)
5.3.3 Quadrature Sampling of Bandpass Signal

The band-pass signal g(t) is sampled in terms of inphase and quadrature
phase components, and whose spectrum is limited to a B.W of 2ω centered
around the frequency fc . [Bandpass signal x(t) is sampled in terms of
inphase and quadrature phase components at the rate of 2ω Hz/sec is called
quadrature sampled.
Definition: The bandpass signal g(t) is sampled in terms of inphase and

quadrature phase components ar the rate of 2ω samples per second. This
form of sampling is called Quadrature Sampling.
Generation of inphase and quadrature phase components
LPF 1/2 gI(nTs)

Sampled at rate
cos2fc(t) of 2
g(t)
LPF 1/2 gQ(nTs)
sin2fc(t)
5.3.4 Reconstruction of Bandpass Signal g(t) (or) Reconstruction of

Message from it’s Sample (or) Proof of Reconstructions of
Message for its Samples with Zero Error
Message signal is x(t) with autocorrelation Rx (τ ) and PSD Sx (f ). Assume
that Sx (f ) = 0. Consider reconstructed message x (t).
∞
n
x (t) = X sin c(2ωt − n) (5.9)
n=−∞
2ω
Mean square value of error ()

= E[(x(t) − x (t))2 ]
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 157 — #9

i i
Reconstruction
gI(nTs)
filter
⫹
cos2␲fc(t) ⌺
g(t)
Reconstruction ⫺
gQ(nTs) filter
sin2␲fc(t)
Figure 5.3: Reconstruction of BPS g(t).
x(t) is original message and x (t) is reconstructed message.

= E[x2 (t)] − 2E[x(t)x (t)] + E[(x (t))2 ] (5.10)
1st term in above equation
E[x2 (t)] = Rx (0) (5.11)
2nd term in above equation

∞
n

E[x(t)x (t)] = E x(t) x sin c(2ωt − n)
n=−∞
2ω
∞
n
= E x(t) × sin c(2ωt − n)
n=−∞
2ω
∞
n
= Rx t − sin c(2ωt − n)
n=−∞
2ω
Input x(t) is WSS it does not vary with time t = 0 and n = 0.
RX (0) sin c(0)

RX (0)(1)
therefore
E[x(t)x (t)] = Rx (0) (5.12)
3rd term in above equation

∞

n ∞
k

E[(x (t)) ] = E
2
· sin c(2ωt − n)+ · sin c(2ωt − k)
n=−∞
2ω 2ω
k=−∞
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 158 — #10

i i
where n, k and t are time constants.

∞
∞

n k
sin c(2ωt − n) sin c(2ωt − k)E x x
n=−∞
2ω 2ω
k=−∞
∞
∞

n−k
sin c(2ωt − n) sin c(2ωt − k)Rx
n=−∞
2ω
k=−∞
Input signal x(t) is WSS time constant n, t and k = 0.
E[x(t)x (t)] = Rx (0) (5.13)
substitute (5.11) to (5.13) in (5.10)
= Rx (0) − 2Rx (0) + Rx (0) = 0
Conclusion If stationary message contains no frequency components

higher than W Hz. Then it may be reconstructed with zero mean square
error.
5.3.5 Types of Sampling

There are three types of the sampling (i) Impulse sampling, (2) Flat top
sampling and (3) Natural sampling.
x(t) X(f)
0 f
t
(f)
(t)
f
t
x(t) X(f)
f
t 3 2 2 3
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 159 — #11

i i
x(t) → Input analog signal.

xδ (t) → Sampled version of input signal.
δ(t) → Impulse train.
x(f ) → Spectrum of input signal.
δ(f ) → Spectrum of impulse train.
xδ (f ) → Spectrum of sampled signal.
Impulse Sampling An analog waveform x(t) is sampled by a sequence of

unit impulses Xδ (t) (or) dirac delta functions.
Flat Top Sampling Flat top rectangular pulses are used to sample the analog
signal. Top of a each pulse in the sampled signal Xs (t) is flat.
x(t)
xp(t)
t
xs(t)
Figure 5.4: Flat top sampling.
Natural Sampling Flat top rectangular pulses are used to take samples. Top
of each pulse in the sampled signal Xs (t) retains the shape of the original
signal.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 160 — #12

i i
x(t)
Original signal
t
Narrow
xp(t) rectangular
pulses
t
xs(t)
Sampled signal
Figure 5.5: Natural sampling.
5.3.6 Limitations of Sampling

Practically band limiting is not possible because if signal is bandlimited
overlapping of signals occurs.
Aliasing Interference of high frequency component with low frequency

component in the spectrum of sampled version is called Aliasing.
Effects of Aliasing Due to aliasing, original signal cannot be reconstructed

properly and distortion occurs. Causes of aliasing contains two types (1)
Undersampling (fs < 2fm) and (2) Oversampling (fs >> 2fm).
To Avoid Aliasing
Step 1: Prior to sampling, low pass antialiasing filter is used to attenuate
the high frequency components.
Step 2: Filtered signal is sampled by selecting a sampling rate slightly
greater than the Nyquist rate.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 161 — #13

i i
fs > 2fm
Low pass
G(f) anti-aliasing Sampler G(f)
filter
5.3.7 Calculation of Aliasing Error

= |g(t) − g (t)| (5.14)
we know that g(t) is original message signal, and g (t) reconstructed message
signal.
ˆ ∞
g(t) = G(f )ej2π ft df
−∞
or
∞ ˆ
(m+1/2)f
g(t) = G(f )ej2π ft df (5.15)
m=−∞ (m−1/2)f
Reconstructed message:
∞

Gδ (f ) = fs G(f − mfs ) (5.16)
m=−∞
∞
n −jπ nf
Gδ (f ) = g e ω (5.17)
n=−∞
2ω
compare eqn. (5.16) and (5.17),

1 n −jπ nf
∞ ∞
G(f − mfs ) = g e ω
m=−∞
fs n=−∞ 2ω
−f s
multiply by ejπ ft df on left hand side and integrate with respect to ‘f ’ from 2
to +f2 s . Therefore
∞ ˆ fs
2

g (t) = G(f − mfs )ej2π ft df
−f s
m=−∞ 2
replace f with f − mfs

∞ ˆ
(m+ 12 )f s

g (t) = ej2π(f −mfs )t df
1
m=−∞ (m− 2 )f s
∞
ˆ (m+ 12 )f s
g (t) = e−j2π mfs t G(f )ej2π ft df (5.18)
m=−∞ (m− 12 )f s
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 162 — #14

i i
where f gives
fs
= f − mfs
2
fs
f = + mfs
2

1
f = fs m +
2
−fs
= f − mfs
2

1
f = fs m −
2
substitute (5.15) and (5.18) in (5.14),

∞ ˆ
(m+ 12 )f s
= G(f )ej2π ft df
(m− 1 )f
s
m=−∞ 2
∞ ˆ (m+ 1 )fs

−j2π mf s t
2
− e G(f )e j2π ft
df
m=−∞
1
(m− 2 )f s
∞ ˆ (m+ 1 )fs

j2π ft
2
−j2π mf s t
= [1 − e ] G(f )e df
(m− 1 )f s
m=−∞ 2
absolute value is
1 − e−j2π mft = 2
therefore
ˆ (m+ 12 )f s ˆ fs
2
G(f )e j2π ft
df ≤ |G(f )|df
−f s
(m− 12 )f s 2
ˆ fs
2
≤2 |G(f )|df
−f s
2
G(f) G(f)
G(0) G(1)
fs fs/2 0 fs/2 fs 3fs/2 2fs
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 163 — #15

i i
5.4 Quantization
Definition: Discretization in amplitude:
Sampled signal Sampled signal

Quantizer discretised in amplitude
discretised in time
and time
L Level Linear Quantizer
Quantile increases Vp
Vp2
Vp32
52
32
step size
2
0 Vpp
32
52
Vp 32
Vp 2 Vp
Figure 5.6: Quantizer.
L is level of the quantizer, is step size, Vpp is peak to peak voltage, Vp is

peak voltage.
Vpp = Vp − (−Vp )/L

Vpp = 2Vp /L
5.4.1 Quantization Error (or) Quantization Noise Power

Definition: Difference between the original analog signal and quantized
signal called Quantization Noise. Maximum quantization error occurs
between −2
to 2 .
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 164 — #16

i i
Derivation:
Vp − (−Vp ) 2Vp
= =
L L
is step size of quantizer. L is Number of levels. Quantization error (or)
Quantization noise power
[E(Q)]2 = σq2
Sp
SNR =
Np
Sp
=
[E(Q)]2
P
=
[E(Q)]2
ˆ ∞
[E(Q)] =
2
q2 fθ (q)dq (5.19)
−∞
1
fθ (q) = (5.20)

fQ(q)
q
/2 /2
ˆ
2 1
[E(θ )] =
2
q2 dq
−
2

1 q3 2
[E(θ )] =
2
3 −
2
2
[E(θ )]2 = (5.21)
12
2Vp
= (5.22)
L
L = 2R (5.23)
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 165 — #17

i i
is step size, L is levels of quantizer and R is PCM word size (or) length.
2Vp
= (5.24)
2R
Vp2
[E(θ )]2 = (5.25)
3L2
Vp2
[E(θ )]2 =
3 × L2
P × 3L2
(SNR) =
Vp2
SNR = 3L2 P (∴ Vp = 1) (5.26)
L = 2R (5.27)
R is word size or word length. Substitute eqn. (5.27) in (5.26), therefore
SNR = 3P22R
Case (1): Normalized input P = 1, therefore
SNR = 3P22R = 3 × 22R
In dB,
(SNR)dB = 10 log10 [3 × 22R ]dB
= 10 log10 3 + 2R10 log 2
Normalized input (SNR)dB = (4.8 + 6R)dB
Case (2): Sinusoidal input signal P = 1

2
1
SNR = 3P22R = 3 × × 22R
2
3
SNR = × 22R
2

3
dB = 10 log10 ×2 2R
dB
2

3
= 10 log10 + 10 × 2R log10 2 dB
2
Sinusoidal input (SNR)dB = (1.8 + 6R)dB
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 166 — #18

i i
5.4.2 Types of Quantization

There are two type of quantization: (1) Uniform quantization: and (2)
Non-Uniform quantization.
Uniform Quantization Step size between two quantization levels remains

constant over the complete amplitude range. There are three types of uniform
quantization. (1) midtread, (2) midrise and (3) biased.
1. Midtread Quantizer: (−)
Quantizer output is zero when input is zero. Origin lies in the middle
of a tread of the stair case.
Output level
2
ⴚ4 ⴚ2
0 2 4 Input level
ⴚ2
ⴚ4
2. Midrise Quantizer: (l)

Quantizer output is ± 2 when input is zero. Origin lies in the middle
of rising part of the staircase like a graph.
Output level
⌬ ⌬/2
⌬/2 Input level
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 167 — #19

i i
3. Biased Quantizer:
Quantizer output is zero when the input is between 0 to .
Output
levels
0 Input
levels
Non Uniform Quantization Step size between two quantization levels are
non uniform throughout complete amplitude range.
Application → Human Speech → Preponderance of Low Amp Signals
Quantize lower amplitude portion of speech more precisely than larger

amplitude portion.
[m(t)]n
(Peak value/rms value) = Kcr =
E(m)2
Crest Factor Describes how strong the signal defined as a ratio of peak value
of the signal and RMS value of the signal.
Probability Output
Statistics of
human speech
Non uniform
quantizer
I Input
5.4.3 Companding
Definition: Compressed and quantized signal is transmitted through the
channel and it can be undistorted at the receiver by the same algorithm
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 168 — #20

i i
(expansion). The processing pair compression and expansion is collectively

reffered to as companding.
Two Laws of compression

-Law A-Law
μ-Law: Algorithm
• used in DCS of North America and Japan.
• used to reduce dynamic range of speech signal.
loge (1 + μ|m|)
|V| =
loge (1 + μ)
Reciprocal slope of compression curve,
d|m| log(1 + μ)
= (1 + μ|m|)
d|v| μ
• Linear at low input levels.

• Logarithmic at high input levels.
• A is compression factor is 87.7.
A-Law companding
A|m| 1
|V| = ; 0 ≤ |m| ≤
1 + loge A A
1 + loge (A|m|) 1
|V| = ; ≤ |m| ≤ 0
1 + loge A A
Reciprocal curve of compression curve,
d|m| 1 + log A 1
= ; 0 ≤ |m| ≤
d|v| A A
d|m| 1
= (1 + A)|m|; ≤ |m| ≤ 1
d|v| A
μ compression factor is 255. It is used in European DCS.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 169 — #21

i i
Normalized output |V|
Normalized output |V|

0
10 00

1
A

5 2

0 A
A1
0 0.2 0.4 I0.6 0 0.2 0.4 0.6 I

Normalized input |m| Normalized input |m|

-law A-law
Transfer Characteristics of Quantizer Two fold effects of quantizer:

1. Peak to peak range of input sample value is subdivided into a finite set
of decision levels or decision thresholds that are aligned with the risers
of the staircase.
2. The output is assigned a discrete value selected from a (finite set of)
representation levels or reconstruction values that are aligned with
treads of the staircase.
Output level
Represent at 3⌬
levels
2⌬
⌬
ⴚ3⌬Ⲑ2 ⴚ⌬Ⲑ2
0 3⌬Ⲑ2 5⌬Ⲑ2 Input level
ⴚ⌬
Decesion levels
ⴚ2⌬
ⴚ3⌬
5.5 Pulse Modulation

5.5.1 Introduction
In analog modulation message signal and carrier signal both are analog in
nature. In case of pulse modulation message signal is analog but train of
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 170 — #22

i i
discrete pulses act as carrier. Like analog carrier pulse carrier also have some
parameters such that amplitude, width and position.
Definition: “Characteristics or Parameters of pulse carrier is varied in

accordance with amplitude of analog message signal.”
Type of Pulse Modulation
Analog Pulse Modulation Digital Pulse

Modulation
PAM PTM
PWM PPM
PCM DM ADM DCP
• In analog pulse modulation, some parameters of pulse carrier is varying

with respect to sample value of message signal.
• Transmission of takes place at discrete times.
• In digital pulse modulation, message signal is transmitted in the form
of code words.
• Digital pulses are combined together to form a code words. Thus the
transmitted signal in digital pulse communication is a digital signal.
Analog Pulse Modulation

• Analog pulse modulation is a technique in which each sample of analog
message signal systematically changes the characteristics of a pulse
carrier (Amplitude, Width, Position).
• Analog pulse modulation classified in to three types:
1. Pulse Amplitude Modulation (PAM)
2. Pulse Width Modulation (PWM)
3. Pulse Position Modulation (PPM)
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 171 — #23

i i
5.5.2 Pulse Amplitude Modulation (PAM)

Definition: Amplitude of the pulse carrier is varied with respect to the
amplitude of the message signal. Where as width and position of the pulse
carrier remains constant.
Generation of PAM The message signal m(t) is given to the low pass filter.
Low pass filter (LPF)

• LPF is used to band limit the message signal m(t) to the maximum
frequency ‘fm ’.
• It attenuate the frequency higher than ‘fm ’
• LPF also act as anti aliasing filter to avoid aliasing.
Pulse Train Generator: It produces a periodic pulse train of frequency fs ,
where fs > 2fm hence Nyquist criteria is satisfied.
Multiplier Sampling takes place at the block of multiplier and it generates

PAM signal.
Multiplier
Message LPF (fm) PAM signal
signal m(t)
Periodic pulse
train generator (fs)
Figure 5.7: Black diagram of PAM generator.
Message signal modulating signal m(t) is multiplied with pulse train

mp (t) to produce PAM signal [mPAM (t)].
Detection of PAM Reconstruction low pass filter act as a PAM detector. It

detects the message signal form PAM.
Reconstructed
PAM Reconstruction message
signal LPF signal
Figure 5.8: PAM detector.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 172 — #24

i i
Amp (V)
(a) Message signal m(t)

Amp (V)
(b) Periodic pulse train mp(t)

Amp (V)
(c) PAM signal (mPAM(t))
Figure 5.9: Waveforms of PAM.
Transmission Band Width of PAM (BT) To transmit a PAM signal, band-

width will be greater or equal to maximum frequency ‘fmax ’ is required.
BT ≥ fmax (5.28)
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 173 — #25

i i
ON
time OFF
time
Ts ␶
Figure 5.10: PAM signal with same on and off time.
The pulse duration ‘τ ’ of the PAM signal is very small compared to time
period ‘Ts ’ between two samples.
Maximum frequency of PAM signal can be written as
1
f=
TON + TOFF
1
=
τ +τ
1
=
2τ
1
fmax = (5.29)
2τ
Sub eqn. (5.29) in (5.28)
1
BT ≥ (5.30)
2τ
we know that τ is very small compared to Ts , that is, τ << TS
1
T << (5.31)
2fm
since ‘TS = 2fm . Combine eqn. (5.30) and (5.31) we can write
1
BT ≥ >> fm (5.32)
2τ
therefore transmission band width of PAM is
BT >> fm
where, Ts is sampling time period and fm is maximum frequency of message

signal (or) modulating signal.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 174 — #26

i i
Advantages
• PAM is the basics of other pulse modulation techniques like DM, ADM,
PCM.
• Generation of PAM is very easy because a multiplier act as a PAM
modulator.
Disadvantages
• A kind of distortion aperture effect occurs.

• It requires huge band width.
• Poor noise immunity.
• Interference of noise is maximum since the amplitude of PAM pulse is
varied.
• Power requirement is not constant.
5.5.3 Pulse Width Modulation (PWM/PDM)

Definition: Width of the pulse carrier is varied with respect to amplitude of
message (or) modulating signal, where as amplitude and position of pulse
remains constant. Pulse width modulation is also referred as pulse duration
modulation (PDM).
Generation of PWM
• Message signal, that is, modulating signal m(t) is applied to

non-inverting terminal of comparator.
• Saw tooth generator produces saw tooth signal. It act as a carrier and
applied to inverting terminal of a comparator.
PWM
output
Message
signal m(t) Monostable PPM
multivibrator output

Comparator
Saw tooth
generator
Figure 5.11: Block diagram of PWM generation.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 175 — #27

i i
• Output of the comparator remain high as long as amplitude of m(t) is

higher than that of the saw tooth signal.
• Output of comparator is PWM signal.
• This block diagram is also used for PPM Generation.
Demodulation of PWM The combination of product detector and a low

pass filter act as PWM detector. Carrier and PWM signal are given to the
product detector, and then a sequence of pulses having the width inversely
proportional to the width of PWM pulses appears at the output. When
the ‘Ea ’ signal passes through the low pass filter. At the output of LPF
demodulated signal is obtained.
Product Ea
PWM detector
signal LPF Demodulator
(Multiplier) signal
Carrier signal
Figure 5.12: PWM detector.
Output of comparator is high when instantaneous value of m(t) is higher
than that of saw tooth signal as shown in above figure. The leading edge of
PWM signal occurs at fixed time period and trailing edge of the output of
comparator depends on the amplitude of message signal m(t). When the saw
tooth voltage is greater the amplitude of m(t) at that instant, compute output
remains zero.
Advantages
• More noise immunity since amplitude kept constant.
• Synchronization is not required.
Disadvantages
• Transmission bandwidth for PWM signal is very high than PAM.
• Power requirement is not constant. Variable pulse width causes variable
power content.
5.5.4 Pulse Position Modulation

Definition: Position of the pulse carrier varies with respect to amplitude of
the message (or) modulating signal. Where as amplitude an width of pulse
carrier is kept constant.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 176 — #28

i i
Figure 5.13: PWM and PPM wave forms.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 177 — #29

i i
Amp (V)
Carrier
signal
t
PWM
signal
t
Ea
t
Figure 5.14: PWM detector output waveform.
Message
signal m(t) ⫹ PWM Monostable PPM
multivibrator signal
⫺
Comparator
Saw tooth
generator
Figure 5.15: Generation of PPM.
Generation of PPM Message signal m(t) and saw tooth signal from saw
tooth generator is given to the comparator as input signal. Comparator
produces PWM wave form. This PWM signal act as trigger input to the
monostable multivibrator. Output of the monostable multivibrator remains
zero till the trigger input is given. Multivibrator is triggered on the falling
edge of PWM pulse and output of multivibrator switches to positive level.
This voltage remains high for fixed duration and it goes low level. Output
waveform is nothing but PPM waveform.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 178 — #30

i i
PPM PWM
R signal Reverse
signal PWM
Q original
FF Detector
signal
Triggering S
pulses
Synchronization Reference
signal pulse
generator
Figure 5.16: Block diagram of PPM demodulator.
Demodulation of PPM PPM pulses are converted into PWM pulses by

SR flip flop. As shown in figure the flip flop is set to ON when is receives
reference pulse from reference pulse generator. Flip flop remains in ON
state till the leading edge of PPM pulse and turns OFF after finite duration.
PWM pulses are obtained at the output of FF. Then the PWM pulses are
demodulated by PWM demodulated to get reconstructed to get message
signal.
Amp (V)
Position of
modulated
pulses time
Triggering
pulses time
PWM
pulses time
Figure 5.17: PPM demodulated waveforms.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 179 — #31

i i
Transmission Bandwidth (BT ) of PPM and PWM

1
BT ≥
2Tr
where BT is transmission bandwidth and Tr is rise time.
Advantages
• The width of the pulse is constant hence transmitter power remains
constant.
• Because of constant amplitude pulses noise interference is less.
Disadvantage
• It needs synchronization between transmitter and receiver.
5.6 Frequency Division Multiplexing (FDM) System
The operation of frequency division multiplexing (FDM) is based on sharing

the available bandwidth of a communication channel among the signals to
be transmitted. This means that many signals are transmitted simultaneously
with each signal occupying a different frequency slot within a common
bandwidth. Each signal to be transmitted modulates a different carrier. The
modulation can be AM, SSB, FM or PM. The modulated signals are then
added together to form a composite signal which is transmitted over a single
channel. The spectrum of composite FDM signal has been shown in Figure
5.18. Generally, the FDM systems are used for multiplexing the analog
signals.
Figure 5.18: Spectrum of FDM signal.
5.6.1 FDM Transmitter

Figure 5.19 shows the block diagram of an FDM transmitter. The signals
which are to be multiplexed will each modulate a separate carrier. The type
of modulation can be AM, SSB, FM or PM. The modulated signals are then
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 180 — #32

i i
Figure 5.19: FDM transmitter.
added together to form a complex signal which is transmitted over a single

channel.
Working Operation of the FDM Transmitter Each signal modulates a

separate carrier. The modulator outputs will contain the sidebands of the
corresponding signals. The modulator outputs are added together in a linear
mixer or adder.
The linear mixer is different from the normal mixers. Here the sum and
difference frequency components are not produced. But only the algebraic
addition of the modulated outputs will take place. Different signals are thus
added together i the time domain but they have a separate identity in the
frequency domain. This is as shown in Figure 5.19. The composite signal at
the output of mixer is transmitted over the single communication channel as
shown in Figure 5.19. This signal can be used to modulate a radio transmitter
if the FDM signal is to be transmitted through air.
5.6.2 FDM Receiver

The block diagram of an FDM receiver is shown in Figure 5.20. The
composite signal is applied to a group of bandpass filters (BPF). Each BPF
has a center frequency corresponding to one of the carriers. The BPFs have
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 181 — #33

i i
Figure 5.20: FDM receiver.
an adequate bandwidth to pass all the channel information without any

distortion. Each filter will pass only its channel and rejects all the other
channels. The channel demodulator then removes the carrier and recovers
the original signal back.
5.7 Pulse Code Modulation
The last block of formatter is encoder. Encoder represent a quantized discrete

signal to digital signal. Represent discrete sample as bits. Discrete samples
are converted to coded pulses. There are three types of encoder.
1. Temporal waveform encoding
2. Spectral waveform encoding
3. Model based encoding
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 182 — #34

i i
5.7.1 Temporal Waveform Encoding

Definition: Encoder captures the tempored characteristics of source wave-
form, speech (or) music signal (time varying signal). Time domain wave
form is encoded. Bit rate is high compared to signal BW.
Types of Temporal Waveform Encoding
PCM DPCM DM
ADM
APCM ADPCM
– PCM: Pulse Code Modulation

– APCM: Adaptive PCM
– DPCM: Differential PCM
– ADPCM: Adaptive Differential PCM
– DM: Delta Modulation
– ADM: Adaptive Delta Modulation
5.7.2 Pulse Code Modulation

Definition: In PCM, PAM signal is quantized and converted to a digital code
this signal is called as pulse code modulated signal. This process is called as
pulse code modulation. In PCM message is represented the form of coded
pulses.
Three major blocks:
1. Transmitter.
2. Transmission path.
3. Receiver.
(i) Transmitter: Source of continuous time message signal. Produces

continuously varying message signal with time.
(ii) Low Pass Filter: Act as anti aliasing filter, attenuate high frequency
components.
(iii) Sampler: Filtered signal is sampled with the help of rectangular
pulses. Emitter follower circuit is used as sampler.
i i
i i
i
i
i
i
Transmitter
Source of
continuous Low pass
time MSG Sampler Quantizer Encoder PCM
filter
signal
Regenerative Regenerative
repeater repeater
Transmission
path
Regenerative Reconstruction Destination

Decoder
circuit filter
“Unit-5” — 2018/12/17 — 14:52 — page 183 — #35
Figure 5.21: Block diagram of PCM.
i
i
i
i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 184 — #36

i i
Input Vcc
signal
Sampled
R2 signal
Clock signal
Quantizer Sampled signal is quantized so that resultant signal is discrete in

both time and amplitude. There are two different types:
(i) Uniform Quantization (Step size is equal), these are two types: (1)
midtread quantizer and (2) midrise quantizer.
(ii) Non Uniform Quantization (Step size is not equal), there are two
types: (1) A law and μ-law.
Encoding is discrete samples are converted to coded pulses. Presence or
absence of a pulse is a symbol. Two types: (1) Unipolar NRZ and (2) Polar
NRZ.
0 1 1
A 0
⫺A
⫺A
Regenerative Repeater
– Equalizer is shapes the received pulses.
– Time circuit is provides a periodic pulse train for sampling.
– Decision Making Device is sample extracted is compared to a predeter-
mined threshold.
– Decoding is regenerated PCM are decoded into quantized PAM.
– Reconstruction filter is message signal is recovered by passing the
decoder output to reconstruction filter whose cut-off frequency is
message bandwidth ‘ω’.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 185 — #37

i i
Distorted Amplifier Decesion

PCM equalizer making device
Timing
circuit
Figure 5.22: Regenerative repeaters.
5.7.3 PCM Bandwidth: (Transmission B.W BT )

For a sampling frequency ‘fs ’ the symbol rate is
Rs = lfs (5.33)
Case 1: To Transmit Sinc pulse.
Rs
Transmission bandwidth ≥ (5.34)
2
Maximum frequency of input signal is ‘B’.
fs ≥ 2fm ||fs = 2fm ||fs = 2B (5.35)
Rs = l2B (5.36)
BT (or) Tx−n B.W ≥ lB
Case 2: For Rectangular Pulses

Transmission BW is Rs
BT (or) Transmission B.W = 2lB
5.7.4 PCM Word Size

Definition: Number of bits per sample. Choice of wordsize depends on how
much quantization noise we are willing to tolerate with PCM format.

1
l ≥ log2 bits
2p
p is tolerance in quantization error (or) fraction of the peak to peak analog
voltage Vpp .
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 186 — #38

i i
5.7.5 Noise in PCM System

1. Aliasing noise: Occurs due to improper sampling.
2. Quantization noise: Introduced in the transmitter. Signal dependent
and disappears when message signal is switched OFF. Completely
known from the specifications of quantizer.
3. Channel noise: Introduces bit error in the transmission path. To
minimize BER, additive white Gaussian channel is used.
Intersymbol interference is communication channel is band limited, hence
waveform passing through it is always spreads.
Case 1: Spreading of pulse is small.
Channel B.W > Pulse B.W
Case 2: Spreading will exceed a symbol duration and it causes signal

pulses to overlap. This overlapping is known as ISI.
Channel B.W=pulse B.W
5.7.6 SNR of a PCM System

Quantization noise alone present.
m̂(t) = m(t) + eq (t)
m(t) is original message signal. m̂(t) is reconstructed message signal at

receiver.
S Sq
=
N Nq

Sq = m 2 (t)
Quantized signal power is 2nd moment of original signal.

m2p
Nq = e2q =
3L2
Quantization noise power is number of levels of quantizer
Sq
(SNR)q =
Nq

m 2 (t)
= 3L2
m2 p
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 187 — #39

i i

m 2 (t)
(SNR)q = 3L2
m2 p
Channel Noise
ec = mq − m̂q
Error probability of channel noise:

2Eb
Pe = Q
η
where Eb is energy of binary pulse. η is one sided PSD of AWGN.
So
Eb =
2Bl
⎡ ⎤
So
2 2Bl
Pe = Q ⎣ ⎦
η

So
=Q
Blη

rb s0
Pe = Q ∵ rb =
l Bn
rb is baseband signal to noise power ratio. l is PCM word size.
Error Probability
1. No repeaters

rb
Pe = Q
l
2. For one repeaters

Krb
Pe = Q
l
3. Sequence of repeater is used

Krb
Pe = KQ
l
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 188 — #40

i i
Threshold and Saturation Effect of PCM When rb is small, SNR changing

very fast. When rb is large, SNR crosses threshold value and reaches
saturation region.
SNR (dB)
Threshold
l⫽8
⫺30
⫺20 l⫽4
Saturation region
⫺10
0 10 20 30 40 50 (dB)
Band width expansion factor:
BT
b=
B
BT is transmission bandwidth of PCM. B bandwidth of input signal.
Advantages
– Inexpensive circuitry.
– Better noise immunity.
– Secured communication.
– Error probability can be reduced by appropriate coding technique.
– Easy multiplexing.
Disadvantages
– Requires larger bandwidth.

– If Bandwidth is reduced ISI will occur.
Application
– Delta Modulation.
– DPCM.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 189 — #41

i i
5.8 Time Division Multiplexing (TDM)

[Digital Circuit Switch]
Definition: Combining various digital signals on to a single transmission

medium. A technique of separating the signals in time.
5.8.1 Operation
Incoming signals are divided into equal fixed length time slots. After
multiplexing these signals are transmitted over a shared medium. And
resembled into their original format after de-multiplexing.
5.8.2 LPF
Message signal is band limited to ‘w Hz’. LPF act as a anti aliasing filter
and remove frequencies that are non essential. Commutator is implemented
using electronic switch.
Two Functions
(1) Take a sample of each message signals at a rate
fs > 2ω
(2) Sequentially inter leave these samples inside a sampling interval
1
Ts =
fs
5.8.3 Pulse Modulator

It’s transform the multiplexed signal into suitable form for transmission over
the communication channel.
5.8.4 Pulse Amplitude Demodulator

Performs the reverse operation of the pulse modulator.
5.8.5 De Commutator
Distributes the appropriate pulses to the respective reconstruction filters.
There are two types of TDM.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 190 — #42

i i
Figure 5.23: Time division multiplexing.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 191 — #43

i i
5.8.6 Types of TDM

1. Synchronous TDM:
– Uses same sampling rate for all signals.
– Minimum sampling rate is equal to twice the maximum frequency of
all the signals.
– Total number of samples per second is equal to ‘N’ times the sampling
rate Nfs
– Transmission B.W is Nf2 s
2. Asynchronous TDM:
– Uses different sampling rate for different signals
– Sampling rate of a signal is equal to twice the maximum frequency of
that signal.
– Total number of samples transmitted per second is equal to sum of
samples plus sync pulses.
– Transmission B.W is half the total number of samples transmitted.
– Design of commutator/de commutator is difficult.
5.8.7 Performance Measure of Digital Communication System

BER can be improved by,
– Increasing transmitted signal power.
– Improving frequency filtering techniques.
– Modulation and demodulation techniques.
– Coding and decoding techniques.
Figure of Merit
– For analog signal
SNR
– For digital signal
Eb
No
Eb is bit energy, No is noise power spectral density. Because in digital
communication system symbol (collection of bits) transmitted and
received. Symbol energy is more useful parameter. So a Energy related
parameter interms of bit level is required to compare differ systems.
i i
i i
i i
“Unit-5” — 2018/12/17 — 14:52 — page 192 — #44

i i
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 193 — #1

i i
University Questions
Unit I - Amplitude Modulation
Part A
1. Define modulation? (May/June 2007)
Modulation is a process by which some characteristics of high frequency
carrier signal is varied in accordance with the instantaneous value of the
modulating signal.
2. What are the types of analog modulation? (Nov/Dec 2005)
Amplitude modulation and Angle Modulation
– Frequency modulation
– Phase modulation.
3. Define depth of modulation
It is defined as the ratio between message amplitude to that of carrier
amplitude.
m = Em /Ec
4. What are the degrees of modulation?
Under modulation m < 1
Critical modulation m = 1
Over modulation m > 1
5. What is the need for modulation?

(OR)
What are the advantages of converting the low frequency signal into
high frequency signal? (May/June 2013)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 194 — #2

i i
194 . University Questions
Needs for modulation:

– Ease of transmission
– Multiplexing
– Reduced noise
– Narrow bandwidth
– Frequency assignment
– Reduce the equipments limitations.
6. What are the types of AM modulators?
There are two types of AM modulators. They are
• Linear modulators
• Non-linear modulators
Linear modulators are classified as follows
• Transistor modulator
There are three types of transistor modulator.
• Collector modulator
• Emitter modulator
• Base modulator
• Switching modulators
Non-linear modulators are classified as follows
• Square law modulator
• Product modulator
• Balanced modulator
7. Give the classification of modulation
There are two types of modulation. They are
• Analog modulation
• Digital modulation
Analog modulation is classified as follows
• Continuous wave modulation
• Pulse modulation
Continuous wave modulation is classified as follows
• Amplitude modulation
• Double side band suppressed carrier
• Single side band suppressed carrier
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 195 — #3

i i
• Vestigial side band suppressed carrier

• Angle modulation
• Frequency modulation
• Phase modulation
Pulse modulation is classified as follows
• Pulse amplitude modulation
• Pulse position modulation
• Pulse duration modulation
• Pulse code modulation
Digital modulation is classified as follows
• Amplitude shift keying
• Phase shift keying
• Frequency shift keying
8. What is single tone and multi tone modulation?
If modulation is performed for a message signal with more than
one frequency component then the modulation is called multi tone
modulation.
If modulation is performed for a message signal with one frequency
component then the modulation is called single tone modulation.
9. The antenna current of an AM transmitter is 8A when only carrier is
sent. It increases to 8.93A when the carrier is modulated by a single
sine wave. Find the percentage modulation.
Solution:
Given: Ic = 8A, It = 8.93A, m = 0.8

Formula: 1
It = Ic (1 + m2 /2) 2
1
8.93 = 8(1 + m2 /2) 2
m = 0.701
1
It = 8(1 + 0.82/2) 2
It = 9.1A
10. Compare AM with DSB-SC and SSB-SC. (Nov/Dec 2005)

(OR)
Compare Bandwidth and Power requirement in terms of carrier
power Pc for AM, DSB-SC, and SSB? (May/June 2013).
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 196 — #4

i i
AM signal DSB-SC SSB-SC

Bandwidth = 2fm Bandwidth = 2fm Bandwidth = fm
Contains USB, LSB, Contains USB,LSB Contains LSB or USB
carrier
More power is Power required is less Power required is less
required for of2 AM.
than that than AM &2 DSB-SC
Pt = Pc 2a Pt = Pc 4a
m m
transmission
m2
Pr = Pc [1 + 2a ]
11. What are the advantages of VSB-AM? (May/June 2012)
– It has bandwidth greater than SSB but less than DSB system.
– Power transmission greater than DSB but less than SSB system.
– No low frequency component lost. Hence it avoids phase distortion.
12. Compare linear and non-linear modulators. (Nov/Dec 2010)
Linear modulators Non-linear modulators

Heavy filtering is not required Heavy filtering is required
These modulators are used in These modulators are used in
low level modulation high level modulation
The carrier voltage is very much The carrier voltage is very much
greater than modulating signal greater than carrier signal
voltage. voltage.
13. How will you generating DSBSC-AM?
There are two ways of generating DSBSC-AM such as (1) balanced
modulator and (2) ring modulators.
14. What are advantages of ring modulator?
1. Its output is stable.
2. It requires no external power source to activate the diodes.
3. Virtually no maintenance.
4. Long life.
15. Define demodulation.
Demodulation or detection is the process by which modulating voltage
is recovered from the modulated signal. It is the reverse process of
modulation.
16. What are the types of AM detectors?
1. Nonlinear detectors
2. Linear detectors
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 197 — #5

i i
17. What are the types of linear detectors?

1. Synchronous or coherent detector.
2. Envelope or non coherent detector.
18. Define multiplexing.
Multiplexing is defined as the process of transmitting several message
signals simultaneously over a single channel.
19. Define sensitivity. (May/June 2007)
It is defined as a measure of its ability to receive weak signals.
20. Define selectivity.
Selectivity of a receiver is defined as its ability to select the desired signals
among the various signals.
21. Define stability.
It is the ability of the receiver to deliver a constant amount of output for
a given period of time.
22. Define super heterodyne principle.
It can be defined as the process of operation of modulated waves to
obtain similarly modulated waves of different frequency. This process
uses a locally generated carrier wave, which determines the change of
frequency.
23. A transmitter supplies 8 Kw to the antenna when modulated. Deter-
mine the total power radiated when modulated to 30%.
M = 0.3;
Pc = 8kw
Pt = Pc (1 + m2 /2)
= 8.36kw
24. What are the drawbacks of emitter modulator?

1. The amplifier is operated in class A mode, thus the efficiency is low.
2. The output power is very small. Thus it is not suitable for generating
high level modulation.
25. Define frequency modulation.
Frequency modulation is defined as the process by which the frequency
ofthe carrier wave is varied in accordance with the instantaneous
amplitude of themodulating or message signal.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 198 — #6

i i
26. What do you meant by multitone modulation? (Nov/Dec 2005)

Modulation done for the message signal with more than one frequency
component is called multitone modulation.
27. Define amplitude modulation. Give the expression for AM wave.
(OR)
Represent an amplitude modulated wave as a function of
time with amplitude sensitivity of the modulator as constant.
(November/December 2013). (May/June 2014).
It is the process by which the amplitude of the carrier wave is changed
in Accordance with the instantaneous value of the message signal
VAM (t) = VC (1 + ma sin ωm t) sin ωc t
28. What are the advantages of SSB. (May/June 2007)

(OR)
State the differences between Single Side Band (SSB) and Vestigial
Side Band (VSB) transmission system (May/June 2014)?
• Fading effect is absent

• The power of the suppressed carrier and sideband is saved
• The effect of noise at the receiver circuits is reduced
29. What are the advantages of Vestigial Side Band? (April/May 2011)
1. It has bandwidth greater than SSB but less than DSB system.
2. Power transmission greater than DSB but less than SSB system.
3. No low frequency component lost. Hence it avoids phase distortion.
30. The average power of a periodic signal gp (t) is calculated using what
theorem? State the theorem. (November/December 2013, May/June
2016)
In engineering, Parseval’s theorem is used to find the power of a periodic
signal:
ˆ ∞ ˆ ∞
1
|x(t)| dt =
2
|X(ω)|2 dω
−∞ 2π −∞
ˆ ∞
= |X(2π f )|2 df
−∞
31. For an AM system, the instantaneous values of carrier and modu-

lating signal are 60 sin ωm t and 40 sin ωs t. Determine the modulation
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 199 — #7

i i
index (May/June 2014).

m = Em/Ec
m = 60/40
m = 1.5
32. Consider the signal whose spectrum is as shown in figure 1 modu-

lates the sinusoidal carrier signal of 1 MHz. Draw the spectrum of
AM signal and SSB-SC (upper sideband) signal. (Nov/Dec-2015)
33. Suggest a modulation scheme for the broad cast video transmission
and justify. (Nov/Dec-2015)
Vestigial Sideband modulation (VSB) is used for the following reasons
1. Video signal exhibits a large bandwidth and significant
low-frequency content which suggests the use of VSB.
2. The circuitry for demodulation in the receiver should be simple and
therefore cheap. VSB demodulation uses a simple envelope detection.
34. List the advantages of FM and AM. (Nov/Dec-2015)
AM has poorer sound quality compared with FM, but is cheaper and
can be transmitted over long distances. It has a lower bandwidth so it
can have more stations available in any frequency range.
FM is less prone to interference than AM. However, FM signals are
impacted by physical barriers. FM has better sound quality due to higher
bandwidth.
35. Define heterodyning. (Apr/May 2015)
In which new frequencies are created by combining or mixing
two frequencies. The two frequencies are combined in a nonlinear
signal-processing device such as a vacuum tube, transistor, or diode,
usually called a mixer.
36. What is the advantages of convolutional DSB-AM over DSB-SC and
SSB-SC AM? (Nov/Dec 2015)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 200 — #8

i i
– AM transmitters are less complex.

– AM receivers are simple, detection is easy.
– AM receivers are cost efficient.
– AM waves can travel a longer distance.
– Low bandwidth.
37. Draw the AM modulated wave for modulation index = 0.5 and its
spectra. (Apr/May-2015)
38. What is pre-envelope and complex envelope? (May/June 2016)

Useful in deriving the general expression of SSB-SC signal. The
pre-envelope of a real valued signal x(t) is defined as
xp(t) = x(t) + jxh(t)
The complex envelope of a real valued signal x(t) is defined as
x∗ p(t) = x(t) − jxh(t)
39. Draw the block diagram of SSB-AM generator. (Nov/Dec 2015)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 201 — #9

i i
Carrier Signal
Balanced Sideband Balanced To Linear

Modulator suppression Mixer Amlifier
filter
Modulating Crystal
Signal Oscillator
40. What are the advantages of superheterodyne receiver over TRF?

(April/May 2010)
The advantages of superheterodyne receiver over TRF are high selec-
tivity, improved sensitivity throughout the carrier frequency band. It
eliminates image frequency.
41. Determine the Hilbert Transform of cos ωt (Nov/Dec 2017)
Calculate the Hilbert transform of cos(ωt) by taking the Fourier
transform, applying the frequency-domain property of the Hilbert
transform, and then taking the inverse Fourier transform.
cos ωt = sin ωt
42. What is VSB? Where is it used? (Nov/Dec 2017)

Vestigial Sideband Modulation or VSB Modulation is the process
where a part of the signal called as vestige is modulated, along with
one sideband. A VSB signal can be plotted as shown in the following
figure. Along with the upper sideband, a part of the lower sideband is
also being transmitted in this technique.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 202 — #10

i i
A guard band of very small width is laid on either side of VSB in order
to avoid the interferences. VSB modulation is mostly used in television
transmissions.
The most prominent and standard application of VSB is for the
transmission of television signals. Also, this is the most convenient and
efficient technique when bandwidth usage is considered.
43. Compare and contrast DSB-SC with SSB-SC with respected to (i)
power and (ii) bandwidth (April/May 2018)
DSB-SC SSB-SC
Power requiredis less
than that Power required is less
2than
AM
m 2a ma
of AM. Pt = Pc 2 & DSB-SC Pt = Pc 4
Bandwidth=2fm Bandwidth= fm
Contains USB, LSB Contains LSB or USB
44. Mention the drawbacks of coherent detector. (April/May 2018)
(i) Perfect synchronization is a requirement
(ii) Noise should be maintained at low level through high SNR
Part B
1. Explain the generation of AM signals using square law modulator Also
derive its efficiency. (16). (Nov/Dec 2005, Apr/May-2015)
2. Explain the detection of AM signals using envelope detector. (16).
(Nov/Dec 2005, Apr/May-2015)
3. Explain about balanced modulator to generate DSB-SC signal. (16)
4. With a neat block diagram explain the SSB transmissions.
5. Explain the operation of a ring modulator. State its advantages.
(Nov/Dec 2008)
6. Explain about coherent detector to detect SSB-SC signal. (16)
7. Explain the generation of SSB using balanced modulator. (Nov/Dec
2006).
8. Draw the circuit diagram of ring modulator and explain with its
operation?
9. Discus the coherent detection of DSB-SC modulated wave with a block
diagram of detector and Explain. (16)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 203 — #11

i i
10. With a neat block diagram Explain frequency translation. (Nov/Dec

2008)
11. Explain the working of super heterodyne receiver with its parameters.
(16)
12. Draw the block diagram for the generation and demodulation of a VSB
signal and explain the principle of operation. (16) (May/June 2007)
13. Write short notes on frequency translation and FDM? (16)
14. Explain about AM transmitters. (16)
15. Define sensitivity, selectivity and image frequency of a receiver system.
16. A commercial AM station is broadcasting with an demodulated carrier
power of 10kW. The modulation index is set at 0.7 for a sinusoidal mes-
sage signal. Find the total transmitter power and efficiency. (May/June
2007)
17. With a help of a neat diagram, explain the operation of an envelope
detector. Why does negative peak clipping take place? (April/May 2011)
18. Compare the characteristics of DSBFC, DSBSC, SSBFC, SSBSC, VSB
schemes (April/May 2011)
19. Explain the concept of FDM with a suitable block diagram. (April/May
2011)
20. Draw an envelope detector circuit used for demodulation of AM and
explain its operation. 10 marks (April/May 2010)
21. How SSB can be generated using Weaver’s method? Illustrate with a neat
block diagram. (6) (April/May 2010)
22. Discuss in detail about frequency translation and frequency division
multiplexing technique with diagrams. (April/May 2010)
23. Compare amplitude modulation and frequency modulation (April/May
2010)
24. (i) With the help of a neat diagram, explain the generation of DSB-SC
using balanced modulator. (ii) Write about the coherent detection
method in detail for DSB-SC and SSB-SC. What happens when there
is phase mismatch? (Nov/Dec 2010)
25. (i) Explain the concept of Frequency Translation (ii) With aid of block
diagram explain the principle of FDM (iii) Illustrate the formation of
Basic group and super group (Nov/Dec 2010)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 204 — #12

i i
26. (i) Draw an envelope detector circuit used for demodulation of AM

and explain its operation. (ii) How SSB can be generated using Weaver’s
method? Illustrate with a neat block diagram. (April/May 2012)
27. (i) What is frequency division multiplexing? Explain. (ii) Compare
various amplitude modulation systems. (April/May 2012)
28. Discuss on the frequency components present in a periodic and non
periodic signal? (May/June 2013).
29. Derive the equation of an AM wave. Also draw the modulated AM wave
for various modulation index. (May/June 2013).
30. The antenna current of an AM transmitter is 8 ampere when only
the carrier is sent. The current increases to 8.93 A when the carrier
is modulated by a single sine wave. Find the percentage modulation.
(May/June 2013).
31. Draw the VSB spectrum and explain the significance (May/June 2013)
32. How do you demodulate AM signal? Explain (May/June 2013)
33. A 1000 kHz carrier is simultaneously AM modulated with 300 Hz, 800
Hz and 1.5 kHz audio sine waves. What will be the frequencies in the
output? (May/June 2013)
34. With suitable block diagrams and equations show how will you generate
DSBSC signals and VSB signals (November/December 2013).
35. A sine wave frequency 10Hz is applied to a product modulator together
with a carrier wave frequency of 1 MHz. The modulator output is
next applied to a resonant circuit. Determine the modulated wave
after transmission through the circuit. Assume suitable data. (Novem-
ber/December 2013)
36. Explain the need for carrier suppression in an AM system. Draw and
explain the functioning of one such system. (May/June 2014)
37. Explain the working of a AM transmitter and that of a receiver with a
suitable block scheme.
38. Explain with block diagram the super heterodyne receiver. (8)
(Apr/May-2015) (May/June 2016)
39. Explain the Hilbert Transform with an example. (8) (Apr/May-2015)
(Nov/Dec-2014)
40. (a) (i) Explain the generation of SSB SC signal using phase shift
method.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 205 — #13

i i
(ii) suggest a scheme for recovering the message signal from the sig-
nal s(t) = 2m(t) cos 2π fct. Explain the same. (Nov/Dec-2014)
(b) (i) An AM signal is generated by modulating the carrier fc = 800
MHz by the signal m(t) = sin 3000π t + 0.5 cos 5000π t the AM
signal s(t) = 100[1 + m(t)] cos 2π fct is fed to a 50 ohm load.
(Nov/Dec-2014)
41. Determine the average power in the carrier and in the sidebands.
(Nov/Dec-2014)
42. Find the modulation index and peak power delivered to the load.
(Nov/Dec-2014)
43. Explain the function of switching modulator in the generation of AM
signal.
44. With neat block diagram explain the function of super heterodyne
receiver in detail. (Nov/Dec-2015) (Nov/Dec-2014) (May/June 2016)
45. Illustrate the superiority of super heterodyne receiver over single tuned
receivers (Nov/Dec-2014)
46. With relevant diagrams, describe the process of demodulation of
DSB-SC AM signal (Nov/Dec-2015)
47. Derive the expression for DSB-SC AM and calculate its power &
efficiency. Explain a method to generate and detect it. (May/June 2016)
48. (i) Explain the operation of enveloper detector (7) (Nov/Dec 2017)
(ii) Discuss the generation of single sideband modulated signal. (6)
(Nov/Dec 2017)
49. Explain the operation of super heterodyne receiver with neat block
diagram. Dram signal at the output of each block. (13) (Nov/Dec 2017)
50. (i) Using the concept of Hilber transform, generate the SSB-SC wave
using phase shift method.
(ii) Using suitable circuit, explain the operation of envelope detector.
Comment the reason for diagonal clipping and suggest the neces-
sary conditions and expressions to overcome the same (April/May
2018)
51. (i) Defend the need of VSB modulation technique in TV broadcasting.
Also sketch its frequency spectra. (April/May 2018)
(ii) With neat block diagram, elaborate the working principle of AM
superheterodyne receiver. Also highlight how super heterodyne
receiver rectifies the drawback of TRF receiver to receiver sensi-
tivity. (April/May 2018)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 206 — #14

i i
Unit II - Angle Modulation
Part A
1. Define phase modulation.
Phase modulation is defined as the process of changing the phase of
the carrier signal in accordance with the instantaneous amplitude of the
message signal.
2. What are the types of Frequency Modulation?
Based on the modulation index FM can be divided into types. They are
narrow band FM and wide band FM. If the modulation index is greater
than one then it is wide band FM and if the modulation index is less
than one then it is narrow band FM.
3. What is the basic difference between an AM signal and a narrowband
FM signal?
In the case of sinusoidal modulation, the basic difference between an
AM signal and a narrowband FM signal is that the algebraic sign of the
lower side frequency in the narrow band FM is reversed.
4. What are the two methods of producing an FM wave? (Nov/Dec
2006).
Basically there are two methods of producing an FM wave. They are,
(i) Direct method: In this method the transmitter originates a wave
whose frequency varies as function of the modulating source. It is
used for the generation of NBFM.
(ii) Indirect method: In this method the transmitter originates a wave
whose phase is a function of the modulation. Normally it is used for
the generation of WBFM where WBFM is generated from NBFM.
5. Compare WBFM and NBFM. (April/May 2011), (Nov/Dec 2013),
(Nov/Dec 2017).
WBFM NBFM
Modulation index is greater than 1 Modulation index less than 1
Frequency deviation 75 KHz Frequency deviation 5 KHz
Bandwidth 15 times NBFM Bandwidth 2fm
Noise is more suppressed Less suppressing of noise
6. List the properties of the Bessel function. (Nov/Dec 2005)
The properties of the Bessel function is given by,
(i) Jn(b) = (−1)nJ − n(b) for all n, both positive and negative.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 207 — #15

i i
(ii) For small values of the modulation index b, we have
J0 (b) = 1
J1 (b) = b/2
Jn (b) = 0, n > 2
7. Define frequency deviation.

The maximum departure of the instantaneous frequency from the
carrier frequency is called frequency deviation.
8. State the Carson’s rule. (April/May 2011), (May/June 2013),
(Nov/Dec 2015)
An approximate rule for the transmission bandwidth of an FM signal
generated by a single tone-modulating signal of frequency fm is
defined as
B = 2Df (1 + 1/b)
9. Define the deviation ratio D for non-sinusoidal modulation.

The deviation ratio D is defined as the ratio of the frequency deviation
Df , which corresponds to the maximum possible amplitude of the
modulation signal m(t), to the highest modulation frequency.
D = Df /fm
10. What is the use of crystal controlled oscillator?

The crystal-controlled oscillator always produces a constant carrier
frequency there by enhancing frequency stability.
11. What are the disadvantages of FM system? (May/June 2012)
1. A much wider channel is required by FM.
2. FM transmitting and receiving equipments tend to be more complex
and hence it is expensive define probability.
12. What is pre-emphasis? Why is it used? (April/May 2010)
(OR)
What is the need for pre-emphasis filter. (May/June 2013) (May/June
2016)
The premodulation filtering in the transistor, to raise the power spectral
density of the base band signal in its upper-frequency range is called
pre emphasis (or pre distortion) Pre emphasis is particularly effective
in FM systems which are used for transmission of audio signals.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 208 — #16

i i
13. Define phase modulation.

It is a type of modulation, used in communication systems, in which
the phase of a carrier wave is varied by an amount proportional to the
instantaneous amplitude of the modulating signal.
14. What are the applications of phase locked loop? (Nov/Dec 2010)
Phase-locked loops are widely used in radio, telecommunications,
computer and other electronic applications.
15. State the frequency in an FM system is 500 Hz and modulating
voltage is 3 V, modulation index is 60. Calculate maximum deviation
and bandwidth.
Modulation index m = δ/fm

60 = δ/500
δ = 60 × 500
= 30 kHz
Bandwidth BW = 2δ
= 2 × 30
= 60 kHz.
16. Mention advantages of angle modulation over amplitude modula-

tion. (Nov/Dec 2015)
1. The amplitude of FM is constant. It is independent of depth of

modulation. Hence transmitter power remains constant in FM
whereas it varies in AM.
2. Since amplitude of FM is constant, the noise interference is mini-
mum in FM.
3. FM uses UHF and VHF ranges, the noise interference is minimum
compared to AM which uses MF and HF ranges.
17. A 80 MHz carrier is frequency modulated by sinusoidal signal of

1V amplitude and the frequency sensitivity is 100Hz/V. Find the
approximate bandwidth of the FM waveform if the modulating signal
has a frequency of 10KHz
2(δ + fm ) = 2(100 + 10000)

= 20.2 kHz.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 209 — #17

i i
18. What is frequency deviation in FM? (November/December 2013).

Frequency deviation is the change in frequency that occurs in the carrier
when it is acted on by a modulating signal frequency. The frequency
deviation is typically given as the peak frequency shift in Hertz (f ).
19. What is the bandwidth required for an FM wave in which the
modulating frequency signal is 2 KHz and the maximum frequency
deviation is 12 Khz? (Nov/Dec 2005)
Bandwidth = 2(δ + fm )
= 2(12 + 2)
= 28 kHz
20. A carrier wave of frequency 100 MHz is frequency modulated by a

signal 20 sin(200π × 103t). What is bandwidth of FM signal if the
frequency sensitivity of the modulation is 25kHz/v.
Bandwidth = 2(δ + fm )
= 2(500 + 100)
= 1.2 MHz
21. Define modulation index of frequency modulation. (April/May

2010), (May/June 2013)
It is defined as the ratio of maximum frequency deviation to the
modulating frequency.
B = f /fm
22. Draw a simple schematic of a PLL demodulator. (November/
December 2013)
Message
Phase signal
i FMl LPF Amplifier
detector
VCO
23. A carrier signal is frequency modulated by a sinusoidal signal of

5 Vpp and 10 kHz. If the frequency deviation constant is 1 kHz/V,
determine the maximum frequency deviation and state whether the
scheme is narrowband FM or Wideband FM. (November/December
2014)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 210 — #18

i i
24. Define lock in range and dynamic range of a PLL. (Apr/May 2015)
The range of frequencies over which the PLL can maintain lock with
the incoming signal is called the lock-in range or tracking range. It is
expressed as a percentage of the VCO free running frequency.
The range of frequencies over which the PLL can acquire lock with
an input signal is called the capture range. It is expressed as a percentage
of the VCO free running frequency.
25. A carrier is frequency - modulated with a sinusoidal signal of 2
kHz resulting in a maximum frequency deviation of 5KHz. Find the
bandwidth of the modulated signal. (Apr/May 2015)
Frequency deviation
(f ) =
Modulating frequency
5 KHz
=
2 KHz
= 2.5
26. A carrier signal is frequency modulated by a sinusoidal signal of 5

Vpp and 10 KHz. If the frequency deviation constant is 1 KHz/V,
determine the maximum frequency deviation and state whether the
scheme is narrow band FM or wide band FM. (May/June 2016)
27. A frequency modulated signal is given as
S(t) = 20 cos[2π fc t + 4 sin(200π t)]
Determine the required transmission bandwidth. (Nov/Dec 2017)

28. Differentiate narrowband FM from AM technique. (April/May 2018)
NBFM AM
Modulation index less than 1 Modulation index is less than 1, equal
to one and greater than one
Frequency deviation 5 KHz Amplitude is varied in accordance
with instantaneous value of message
signal
Bandwidth 2fm Bandwidth 2fm
Less suppressing of noise Presence of more noise
29. What is the need of limiter circuits in FM system. (April/May 2018)
Limiter circuit is used in FM receiver to remove the noise present in the
peaks of the received signal and to remove any amplitude variation in
the received signal; the output of the limiter has constant amplitude.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 211 — #19

i i
Part B
1. Explain the indirect method of generation of FM wave and any one
method of demodulating an FM wave. (16) (Nov/Dec 2006)
2. Derive the expression for the frequency modulated signal. Explain what
is meant by FM.
3. Explain any two techniques of demodulation of FM. (16) (Nov/Dec
2006)
4. Explain the working of the reactance tube modulator and drive an
expression to show how the variation of the amplitude of the input
signal changes the frequency of the output signal of the modulator. (16)
(May/June 2007)
5. Discuss the effects of nonlinearities in FM. (8)
6. Discuss in detail FM stereo multiplexing. (8)
7. (i) Derive the expression for the frequency modulated signal. Explain
what is meant by narrow-band FM and wideband FM using the
expression (Nov/Dec 2006).
(ii) Discuss the indirect method of generating a wideband FM signal.
8. Draw the frequency spectrum of FM and explain. Explain how Varactor
diode can be used for frequency modulation. (16)
9. Draw the circuit diagram of Foster-Seeley discriminator and explain its
working. (8). (Apr/May-2015)
10. Explain the principle of indirect method of generating a wide-band FM
signal with a neat block diagram. (8) (Nov/Dec 2006)
11. Differentiate narrow band and wide band FM. (May/June 2007)
12. Derive an expression for the spectrum of a FM signal with single tone
modulation. Compare the performance of AM and FM.
13. Derive the expression for the single tone frequency modulation and
draw its frequency spectrum (April/May 2011)
14. An angle modulated wave is described by the equation
V(t) = 10 cos(2 × 106 π t + 10 cos 2000π t)
Find (1) Power of the modulated signal (2) Maximum frequency

deviation (3) Bandwidth (April/May 2011) (May/June 2016)
15. With necessary diagrams explain the operation of slope detector for
demodulating FM signal (April/May 2011)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 212 — #20

i i
16. (i) Using suitable mathematical analysis show that FM modulation

produces infinite sideband. Also deduce an expression for the frequency
modulated output and its frequency spectrum. (ii) How can we generate
FM from AM and AM from FM (April/May 2010)
17. (i) Define frequency modulation. Draw the FM waveform. Derive an
expression for single tone frequency modulation. (ii) Compare narrow
band and Wide band FM (Nov/Dec 2010)
18. Draw the circuit diagram of a Foster-Seeley discriminator and
explain its working with relevant phasor diagrams. (April/May 2012)
(May/June 2016)
19. Derive the mathematical expression for FM wave (May/June 2013)
20. When the modulating frequency in an FM system is 400 Hz and the
modulating voltage is 2.4 v, the modulation index is 60. Calculate
the maximum deviation. What is the modulating index when the
modulating frequency is reduced to 250 Hz and the modulating voltage
is simultaneously raised to 3.2 V. (May/June 2013).
21. (i) Explain the Armstrong method to generate FM signal. (May/June
2013) (May/June 2016)
(ii) How is the phase and frequency modulation are related? Explain
(May/June 2013)
(iii) (a) Determine the approximate bandwidth of the FM wave by
Carson’s rule. (b) Determine the bandwidth by transmitting only
those side frequencies whose amplitude exceed 1 % of the unmod-
ulated carrier amplitude (use the universal curve/ideal condition
(November/December 2013)
23. Describe how FM wave is generated by the indirect method and give
a suitable demodulating scheme for the same (November/December
2013).
24. Explain Armstrong method of FM generation. (May/June 2014).
25. Explain the functions of any FM detector circuit. (May/June 2014).
26. Explain how FM is achieved using varactor diodes (May/June 2014).
27. Make atleast five comparisons of AM and FM systems. (May/June 2014)
28. (a) (i) Derive the expression for the single tone frequency modu-
lated signal and hence prove that is the constant envelope
modulation requiring infinite bandwidth. (Nov/Dec-2014)
(ii) Draw the typical spectrum of the FM. (Nov/Dec 2014)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 213 — #21

i i
(b) (i) Explain the indirect method of generating FM signal.

(Nov/Dec-2014) (Nov/Dec-2017)
(ii) Explain the operation of PLL as FM demodulator.
(Nov/Dec-2014)
29. Explain with diagrams the generation of FM using direct method. (8).
(Nov/Dec 2008, Apr/May-2015)
30. An angle modulated wave is described by the equation
V(t) = 10 cos(2 × 106 π t + 0.1 sin 103 π t)
Find
(1) Considering V(t) as a PM signal with Kp = 10, find m(t)
(2) Considering V(t) as a FM signal with Kp = 10π , find m(t)
(Apr/May-2015)
31. With relevant diagrams, explain how the frequency discriminator and
PLL are used as frequency demodulation? (Nov/Dec 2015)
32. Derive the expression for the single tone frequency modulated signal
and comment on the transmission bandwidth. (Nov/Dec 2015)
33. Derive the expression for the single tone frequency modulated signal
with necessary diagrams and draw its frequency spectrum. (May/June
2016)
34. List the advantages of frequency modulation over amplitude modula-
tion. (Nov/Dec 2017)
35. Explain the FM demodulation process using frequency discrimination
process. (Nov/Dec 2017)
36. (i) Obtain a mathematical expression for FM using Bessel’s function.
And also brief the method to determine the bandwidth of FM wave.
(April/May 2018)
(ii) Discuss the process of FM generation using reactance modulator.
(April/May 2018)
37. (i) Highlight the process involved in obtaining amplitude variation
from phase variation using suitable FM demodulator circuit.
(April/May 2018)
(ii) Elucidate the process of FM demodulation using PLL method.
(April/May 2018)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 214 — #22

i i
Unit III - Random Process
Part A
1. Define probability.
The probability of occurrence of an event A is defined as,
Number of possible favorable outcomes

P(A) =
Total number of equal likely outcomes
2. What are mutually exclusive events?

Two possible outcomes of an experiment are defined as being mutually
exclusive if the occurrence of one outcome precludes the occurrence of
the other.
3. Define probability density function.
Probability density function is defined as fx (x) is defined in terms of
cumulative distribution function Fx (x) as
dFx (s)
fx (x) =
dx
4. Define random variable (April/May 2010) (Nov/Dec 2015)

Random variable is defined as a rule or mapping from the original sam-
ple space to a numerical sample space subjected to certain constraints.
Random variable is also defined as a function where domain is the set
of outcomes and whose range is R, is the real line.
5. Define Random process. (May/June 2012)
A random process X(s, t) is a function that maps each element of a
samples space into a time function called sample function. Random
process is a collection of time functions.
6. Define auto correlation function. (May/June 2016)
Defined as a measure of similarity between signal or process and its
replica by available amount.
7. What is meant by Ergodic process? (Nov/Dec 2017)
8. Give the mathematical definition for random process. (April/May
2018)
Random process is defined as the ensemble (collection) of time func-
tions together with a probability rule. x1 (t) is an outcome of experiment
1 x2 (t) is the outcome of experiment 2 . . . xn (t) is the outcome of
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 215 — #23

i i
experiment n. Each sample point in S is associated with a sample

function x(t). X(t, s) is a random process.
9. What is narrow band noise. (April/May 2018)
If the bandwidth of bandlimited noise is relatively small compared to the
carrier frequency it is called as Narrow-band noise (NBN) is a type of
noise that is centered around a small range of frequencies. It is produced
by filtering a 1/3 octave range from a broad-band noise stimulus.
Part B
1. Derive the effective noise temperature of a cascade amplifier. Explain
how the various noise are generated in the method of representing them.
(16)
2. Show how a narrow band noise can be represented as
n(t) = nc (t) cos ωc t − ns (t) sin ωc t
where nc (t) and ns (t) are the in-phase and quadrature phase compo-
nents of noise respectively. (Nov/Dec 2006)
3. Derive the Friis formula. Explain noise effect on bandwidth. (Nov/Dec
2008)
4. What is noise temperature? Deduce the expression for effective noise
temperature for a cascaded system. Explain narrow band noise.
(Nov/Dec 2008)
5. Explain how the various noises are generated and the method of
representing them.
6. Explain concept of noise equivalent bandwidth (Nov/Dec 2007)
7. Write notes on noise temperature and noise figure. (8)
8. What is meant by noise equivalent bandwidth? Illustrate it with a
diagram (Nov/Dec 2006)
9. What is a narrow band noise? Discuss the properties of the quadrature
components of a narrowband noise. (Nov/Dec 2006)
10. Derive the expression for output signal to noise for a DSBSC receiver
using coherent detection.
11. Write short notes on noise in SSB receivers. (Nov/Dec 2006)
12. Derive the noise figure for cascade stages. (8)
13. Write short notes on noise in SSB. (16)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 216 — #24

i i
14. What is narrowband noise discuss the properties of the quadrature

components of a narrowband noise. (8)
15. Write short notes on
(i) Shot noise. (4)

(ii) Thermal noise. (4)
(iii) Noise figure and Noise temperature. (8) (May/June 2007)
16. What is meant by noise equivalent bandwidth? Illustrate it with a

diagram (8)
17. Derive the expression for output signal to noise for a DSB-SC receiver
using coherent detection.
18. Discuss the following: (16)
(a) Noise equivalent bandwidth (4)

(b) Narrow band noise (4)
(c) Noise temperature (4)
(d) Noise spectral density (4)
19. How sine wave plus noise is represented? Obtain the joint PDF of such
noise component.
20. State and prove four properties of Gaussian process.
21. Derive the representation of narrowband noise in terms of envelope and
phase components and list out its properties.
22. Write the details about narrow band noise and the properties of
quadrature components of narrowband noise. (April/May 2010)
23. List the different types of random process and give the definitions.
24. Write short notes on Shot Noise. (Nov/Dec 2010)
25. State and prove four properties of Gaussian process. (April/May 2011)
27. Consider two amplifiers are connected in cascade. First stage amplifier
has gain and noise figure as 10 dB and 2 dB. Second stage has noise figure
of 3 dB. Calculate total noise power. (April/May 2011)
29. List the different types of random process and give the definitions.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 217 — #25

i i
30. Write short notes on shot noise. (April/May 2012) (May/June 2013).
31. Write the definition, power spectral density and autocorrelation func-
tion for white noise and narrow band noise (filtered white noise).
32. What causes thermal noise in a material? Write the expression for RMS
value of the noise. (April/May 2012)
33. Derive relationship between noise figure and equivalent noise tempera-
ture (May/June 2013)
34. Explain the following terms mean, correlation, covariance and ergodic-
ity. (May/June 2013).
35. How do you represent narrowband noise? (May/June 2013)
36. Summarize the characteristics of various noise found in a communica-
tion channel (November/December 2013)
37. Derive the equation for finding the probability density function of a
one to one differentiable function of a given random variable. (Novem-
ber/December 2013).
38. Define and explain the following: (i) Gaussian Noise and Gaussian
Distribution (ii) Thermal Noise (iii) Shot Noise. What type of PDF does
Gaussian noise follow. (May/June 2014)
39. If X is a uniformly distributed

1/2π ; 0 ≤ x ≤ 2π
fX (x) =
0; otherwise
find E(X), E(X2), E[cos X] and E[(X − mx)2] (May/June 2014).

40. Let X and Y be real random variables with finite second moments.
Prove the Cauchy-Schwarz inequality. (E[XY])2E[X2]E[Y2] (8)
(Apr/May-2015)
41. Differentiate the strict sence stationary with that of wide sense station-
ary process. (8) (Apr/May-2015)
42. In a binary communication system, let the probability is sending a 0 and
1 be 0.3 and 0.7 respectively. Let us assume that a 0 being transmitted,
the probability of it being received as 1 is 0.01 and the probability of
error for a transmission of 1 is 0.1.
(i) What is the probability that the output of this channel is 1?
(ii) If a 1 is received, then what is the probability that the input to the
channel was 1? (Nov/Dec 2015)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 218 — #26

i i
43. What is CDF and PDF? State their properties. Also discuss them in
detail by giving examples of CDF and PDF for different types of random
variables. (Nov/Dec 2015)
44. Two random process X(t) = A cos(ωt + θ ) and Y(t) = A sin(ωt + θ )
where A and ω are constants and θ is the uniformly distributed random
variable in (0, 2π ). Find the cross correlation function. (May/June
2016)
45. Explain in detail about the transmission of random process through LTI
filter. (May/June 2016) (Nov/Dec 2017)
46. What is a random process said to be strict sense stationary (SSS), Wide
sense stationary (WSS) and Ergodic process. (May/June 2016)
47. Give a random process,
X(t) = A cos(ωt + μ)
where A and ω are constants and μ is a uniform random variable. Show

that X(t) is ergodic in both mean and auto correlation. (May/June 2016)
48. (i) Discuss the properties of Gaussian noise process. (Nov/Dec 2017)
(ii) Consider a random process defined as X(t) = A cos ωt, where ω is
a constant and A is random. Uniformly distributed over [0,1]. Find
the autocorrelation and auto covariance of X(t). (Nov/Dec 2017)
49. Distinguish between random variable and random process. Give exam-
ples to each. (Nov/Dec 2017)
50. (i) Consider the quadrature-amplitude modulated signal:
Y(t) = X(t) cos(ω0 t) − Z(t) sin(ω0 t)
where X(t) and Z(t) are zero mean independent processes with
identical autocorrelations, Rx = Rz . Determine RY (t1 , t2 ) and show
that if Rx (t1 , t2 ) = Rx (t1 − t2 ), then RY (t1 − t2 ). (April/May 2018)
(ii) Discuss the properties of autocorrelation function. (April/May
2018)
51. (i) State and explain the properties of Gaussian process. (April/May
2018)
(ii) Using suitable sketches, expression, explain the transmission of
random process through a LTI filter. (April/May 2018)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 219 — #27

i i
Unit IV - Noise Characterization
Part A
1. Define noise.
Noise is defined as any unwanted form of energy, which tends to
interfere with proper reception and reproduction of wanted signal.
2. Give the classification of noise.
Noise is broadly classified into two types. They are external noise and
internal noise.
3. What are the types of External noise?
External noise can be classified into: (1) Atmospheric noise, (2)
Extraterrestrial noises and (3) Man-made noises or Industrial noises.
4. What are types of internal noise
Internal noise can be classified into: (1) Thermal noise, (2) Shot noise,
(3) Transit time noise and (4) Miscellaneous internal noise.
5. What are the types of extra terrestrial noise and write their origin?
The two type of extraterrestrial noise are solar noise and cosmic noise
solar noise is the electrical noise emanating from the sun. Cosmic noise
is the noise received from the center part of our galaxy, other distant
galaxies and other virtual point sources.
6. Define transit time of a transistor.
Transit time is defined as the time taken by the electron to travel from
emitter to the collector.
7. Define flicker noise.
Flicker noise is the one appearing in transistors operating at low audio
frequencies. Flicker noise is proportional to the emitter current and
junction temperature and inversely proportional to the frequency.
8. State the reasons for higher noise in mixers.
1. Conversion transconductance of mixers is much lower than the

transconductance of amplifiers.
2. If image frequency rejection is inadequate, the noise associated with
the image frequency also gets accepted.
9. Define signal to noise ratio.

Signal to noise ratio is the ratio of signal power to the noise power at the
same point in a system.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 220 — #28

i i
10. Define noise figure. (May/June 2014), (May/June 2013). S/N at the
input, nose figure F = S/N at the output,
Signal power
S/N =
Noise power
11. Explain thermal noise. (Nov/Dec 2008), (Nov/Dec 2006)

Thermal noise is the name given to the electrical noise arising from the
random motion of electrons in a conductor.
12. Give the expression for noise voltage in a resistor.
The mean-square value of thermal noise voltage is given by
Vn2 = 4KTBR
K - Boltz man constant

R - resistance
T - obsolute temperature
B - bandwidth
13. Explain White Noise. (Nov/Dec 2007) (May/June 2013) (May/June
2014)
Many types of noise sources are Gaussian and have flat spectral den-
sity over a wide frequency range. Such spectrum has all frequency
components in equal portion, and is therefore called white noise. The
power spectral density of white noise is independent of the operating
frequency.
14. What is narrowband noise?
The receiver of a communication system usually includes some provi-
sion for preprocessing the received signal. The preprocessing may take
the form of a narrowband filter whose bandwidth is large enough to pass
modulated component of the received signal essentially undistorted but
not so large as to admit excessive noise through the receiver. The noise
process appearing at the output of such filter is called narrow band noise.
15. Give the expression for equivalent noise temperature in terms of
hypothetical temperature. (May/June 2012)
The expression for equivalent noise temperature in terms of hypotheti-
cal. Temperature is
Te = (F − 1)T0
where F is the noise figure and T0 absolute temperature.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 221 — #29

i i
16. Give the Friss formula in terms of noise temperature.

The Friss formula in terms of noise temperature is
Te = T1 + T2 /G1 + T3 /G1 G2 + . . .
G1 , G2 , . . . are gain of amplifiers.

17. What is called image frequency?
Image frequency is defined as the signal frequency plus twice the
intermediate frequency. This has the effect of two stations being received
simultaneously and hence it is undesirable.
fsi = fs + 2fi
where fsi is image frequency it can be eliminated by providing adequate

image signal selectivity between antenna and mixer input.
18. What is intermediate frequency? (Nov/Dec 2006)
(OR)
What is meant by image frequency in super heterodyne receiver?
Intermediate frequency (IF) is defined as the difference between the
signal frequency and the oscillator frequency. IF = fs − fo when fs > fo
(or) IF = fo − fs when fo > fs .
19. Define partition noise.
In an electron tube having one or more positive grids, this noise is
caused by irratic partition of the cathode current among the positive
electrodes. In a transistor, the partition noise is created from the random
fluctuation in the division of current between the collector and base.
20. Give the expression for noise voltage when several sources are
cascaded.

Enr = (4KTB(R1 + R2 + . . .))
where R1, R2 are the resistances of the noise resistors.

K - Boltz man constant
T - absolute temperature
B - bandwidth
21. Define equivalent B.W (May/June 2008)
It is defined as B.W of an ideal filter at which the noise power passed by
real filter and ideal filter is same.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 222 — #30

i i
22. How will you define narrow-band noise m(t) at the IF filter out-
put in terms of its inphase and quadrature components. (Novem-
ber/December 2013).
The band pass filters have narrow bandwidths in the sense that band-
width is small as compared to center frequency output of this kind of
band pass filter as narrow band noise.
n(t) = nc (t) cos 2π fc t − ns sin 2π fc t
23. DC current of 2 mA flows through the semiconductor junction.

Consider the effective noise bandwidth of 1 kHz and calculate the
shot noise component (November/December 2014)
24. Define the term noise equivalent temperature (November/December
2014) (May/June 2016)
Defined as the temperature at which a noisy resistor has to be main-
tained such that, by connecting the resistor to the input of a noiseless
version of the system, it produces the same available noise power at the
output of the system as that produced by all the source of noise in the
actual system.
Teq = (F − 1)To
25. Define the Q factor of a receiver. (Apr/May-2015)
Q = R/ωL
R- Resistance
ω-Angular frequency
L-Inductance
26. Write the equation for the mean square value of thermal noise voltage
in a resister. (Apr/May-2015)
Pn = 4KTRB
K-Boltzmann’s constant K = 1.38 × 10−23

T-Temperature
B-Bandwidth
27. State Baye’s Rule. (Nov/Dec 2015)
Bayes’ theorem is stated mathematically as the following equation
P(A/B) = (P(B/A) ∗ P(A))/P(B)
where A and B are events and P(B) = 0.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 223 — #31

i i
– P(A) and P(B) are the probabilities of observing A and B without

regard to each other.
– P(A|B), a conditional probability, is the probability of observing event
A given that B is true.
– P(B|A) is the probability of observing event B given that A is true.
28. What is the figure of merit of DSBSC system?

The figure of merit of DSBSC signal is unity.
29. Compare the noise performance of an AM and FM system? (Nov/Dec
2008)
The figure of merit of AM system is 1/3 when the modulation is 100
percent and that of FM is (3/2)mf 2. The use of FM offers improved noise
performance over AM when (3/2)mf 2 > 1/3. mf -modulation index in
FM.
30. What is capture effect? (May/June 2012) (or) Define capture effect in
FM. (May/June 2016)
When the interference signal and FM input are of equal strength, the
receiver fluctuates back and front between them. This phenomenon is
known as the capture effect.
31. What is threshold effect? (May/June 2012)
As the input noise power is increased the carrier to noise ratio is
decreased the receiver breaks and as the carrier to noise ratio is
reduced further crackling sound is heard and the output SNR cannot
be predicted by the equation. This phenomenon is known as threshold
effect.
32. How is threshold reduction achieved in FM system?
Threshold reduction is achieved in FM system by using an FM demod-
ulator with negative feedback or by using a phase locked loop demodu-
lator.
33. What is pre-emphasis? (Apr/May-2015)
The premodulation filtering in the transistor, to raise the power spectral
density of the base band signal in its upper-frequency range is called pre
emphasis (or pre distortion) pre emphasis is particularly effective in FM
systems which are used for transmission of audio signals.
34. Define de-emphasis
The filtering at the receiver to undo the signal pre-emphasis and to
suppress noise is called de-emphasis.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 224 — #32

i i
35. What are the characteristics of a receiver?

The characteristics of a receiver are sensitivity, selectivity, fidelity, signal
to noise ratio.
36. What is a post detection filter?
The post detection filter named as “base-band low pass filter” has
a bandwidth that is just large enough to accommodate the highest
frequency component of the message signal.
37. Define SNR.
It is defined as the ratio of signal power to the noise power.
38. How to achieve threshold reduction in FM receiver?
When the carrier to noise ratio reduces to certain value, the mes-
sage information is lost. The performance of the envelope detector
deteriorates rapidly and it has no proportion with carrier to noise ratio.
39. What is threshold effect with respect to noise?
(OR)
Define threshold effect in AM systems. (Nov/Dec 2015) (Apr/
May-2015)
The loss of message signal in the output of the envelope detector due to
low carrier to noise ratio is called as threshold effect.
40. What is FM threshold effect? (Nov/Dec 2010)
As the input noise power is increased the carrier to noise ratio is
decreased the receiver breaks and as the carrier to noise ratio is reduced
further crackling sound is heard and the output SNR cannot be pre-
dicted by the equation. This phenomenon is known as threshold effect.
41. Define Pre-emphasis and De-emphasis. (Nov/Dec 2010) (November/
December 2013).
Pre-emphasis: The premodulation filtering in the transistor, to raise
the power spectral density of the base band signal in its upper-frequency
range is called pre-emphasis (or pre distortion) pre emphasis is
particularly effective in FM systems which are used for transmission of
audio signals.
De-emphasis: The filtering at the receiver to undo the signal
pre-emphasis and to suppress noise is called de-emphasis.
42. What is coherent system (May/June 2013)
The local carrier generated at the receiver phase locked with the
transmitter carrier is known as coherent detection.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 225 — #33

i i
43. When carrier to noise ratio is high, how will you get the figure of
merit of FM systems? (Nov/Dec 2013).
FOM = 3KF2 P/W 2
44. Define noise figure. (May/June 2014), (May/June 2013). (Nov/Dec

2015)
S/N at the input

Nose figure F =
S/ N at the output
Signal power
S/N =
Noise power ratio
45. Define the term noise equivalent temperature (Nov/Dec 2014)

(May/June 2016) (Nov/Dec 2017)
Defined as the temperature at which a noisy resistor has to be
maintained such that, by connecting the resistor to the input of a
noiseless version of the system, it produces the same available noise
power at the output of the system as that produced by all the source of
noise in the actual system.
46. List the external sources of noise. (Nov/Dec 2017)
Often external noise is called interfering signals. External noise sources
are either natural (such as solar noise, galactic noise, and atmospheric
noise) or man-made (which include industrial noise, electric motors,
arc welders, switches, broadcast communication systems, mobile
phones, etc.).
47. Defend the reason why, the SNR of the receiver should be high.
(April/May 2018)
SNR of the receiver should be high so that good sensitivity, selectivity
and fidelity can be obtained. Noise and interference can also be reduced
through high SNR.
48. How does Pre-emphasis and De-emphasis process provide overall
SNBR improvement in FM systems? (April/May 2018)
Pre-emphasis provides improvement of SNR as at increases the energy
content of high frequency signals and it will be stronger than high
frequency noise components. This improves signal to noise ratio and
increases intelligibility and fidelity. To return the frequency response
to normal level deemphasis is used in the receiver. This combination
increases FM reception quality.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 226 — #34

i i
Part B
1. Explain how threshold improvement is done through de-emphasis.
(Nov/Dec 2008)
2. Derive an expression for the output signal-to-noise ratio of an AM
DSB-FC system. (8) (May/June 2007).
3. Discuss the noise performance of AM system using envelope detection.
(16)
4. Compare the noise performance of AM and FM systems. (16) (Nov/Dec
2007)
5. Explain the significance of pre-emphasis and de-emphasis in FM
system? (8)
6. Derive an expression for the output SNR of an FM receiver and hence
obtain the figure of merit. (Nov/Dec 2006).
7. Derive the noise power spectral density of the FM demodulation and
explain its performance with diagram. (16).
8. Draw the block diagram of FM demodulator and explain the effect of
noise in detail.
9. Explain the FM threshold effect and capture effect in FM? (16)
10. Discuss the following:
(i) Noise equivalent bandwidth [4]
(ii) Narrow-band noise [4]
(iii) Noise temperature [4]
(iv) Noise power spectral density [4] (May/June 2007)
11. Explain the noise in AM receiver using its noisy model block diagram.
What are pre emphasis and de emphasis in FM? Draw suitable circuits
and Explain. (Nov/Dec 2006)
12. (i) Explain the capture effect and FM threshold effect. [8]
(ii) Draw the block diagram of FM demodulator and explain the effect
of noise in detail and compare the performance of AM and FM in
the presence of noise (Nov/Dec 2006).
13. Derive an expression for the output signal to noise ratio of an AM
DSB-FC system.
14. Derive the expression for figure of merit of AM receiver using envelope
detection. What do u infer from the expression. (Nov/Dec 2010)
(Nov/Dec 2017)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 227 — #35

i i
15. (i) Explain pre-emphasis and De-emphasis in detail. (ii) Compare the
performances of AM and FM systems. (April/May 2010)
16. Compare the performance of noise in AM and FM.
17. Define and explain FM threshold effect. With suitable diagram, explain
threshold reduction by FMFB demodulator. (Nov/Dec 2010)
18. (i) Sketch the block diagram of DSB-SC/AM system and derive the
figure of merit.
(ii) Using heterodyne principle, Draw the block diagram of AM radio
receiver and briefly explain it. (April/May 2011)
19. Write short notes on noise in SSB receivers.
20. Explain how threshold improvement is done by FMFB technique.
21. Discuss the effects of noise on the carrier in a FM receiver with suitable
mathematical derivations. (April/May 2012)
22. Draw the superheterodyne receiver and explain the operation of each
block (May/June 2013).
23. Derive the figure of merit for AM system for non coherent system, with
suitable assumptions. (May/June 2013)
24. Explain the functioning of a superhetrodyne receiver and list its charac-
teristics (November/December 2013)
25. Compare the performance of any two CW modulation scheme.
26. Explain the advantages in usage of superheterodyne receivers
(May/June 2014).
27. Explain envelope detection with suitable diagram. (May/June 2014)
28. Explain the method of coherent detection (May/June 2014)
29. Compare at least three important characteristics of various FM systems
(May/June 2014)
30. Derive the expression for the SNR at the output of the FM receiver.
Assume that the input is corrupted by AWGN noise. Discuss the perfor-
mance of the receiver based on the derived expression. (Nov/Dec-2014)
31. The three amplifiers 1, 2 and 3 have the following characteristics: F1 =
9db, G1 = 50db, F2 = 6db, G2 = 30db, F3 = 4db, G3 = 20db. The
amplifiers are connected in tandem. Determine which combination
gives the lowest noise figure.
32. Discuss on thermal noise. (Nov/Dec 2014)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 228 — #36

i i
33. What is meant by narrow band noise? Explain the characteristics of

narrow band noise.
34. An AWGN of power spectral density 1uw is fed through a filter with
frequency response

1
; |f | < 40 kHz
H(f ) = 2
0; elsewhere
Calculate the noise power at the output of the filter. (Nov/Dec 2014)
35. Define noise. Explain the various types of internal noise (8).
(Apr/May-2015)
36. Explain with derivation the effect of noise in cascaded amplifier circuit.
(8). (Apr/May-2015)
37. Derive the SNR performance of DSB system and the AM system. Also
prove that the output SNR in AM is at least 3 DB worse than that of DSB
system (Apr/May-2015)
38. Explain the noise in DSB-SC receiver using synchronous or coherent
detection and calculate the figure of merit for a DSB-SC system?
(May/June 2016)
39. Define narrow band noise and explain the re-presentation of narrow
band noise interms of in phase and quadrature components. (May/June
2016)
40. Explain Pre-emphasis and De-emphasis in FM. (May/June 2016)
(Nov/Dec 2017)
41. An amplifier has three stages with gain 5dB, 20dB and 12dB. The noise
figures of the stages are 7dB, 13dB and 12dB respectively. Determine
the overall noise figure and the noise equivalent temperature. (Nov/Dec
2017)
42. (i) Classify the different types of noise and also comment its cause and
effects. (April/May 2018)
(ii) Prove that the random band pass noise signal n(t) can be expressed
as n(t) = ne (t) cos ωc t + ns (t) sin ωc t, where ne (t) and ns (t) are low
frequency signal band limited to ωm radians/second. (April/May
2018)
43. Obtain an expression for figure of merit for an FM signal, with
assumption that the noise added in the channel is Additive White
Gaussian Noise. (April/May 2018)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 229 — #37

i i
Unit V - Sampling and Quantization
Part A
1. Define time limited and time unlimited signals. (Apr-May 2011)
A time limited signal is one that is non-zero only for a finite length time
interval and time unlimited signal is one which is non-zero for infinite
length time interval. A signal that is band limited is not time-limited
and vice-versa.
2. Give an advantage and disadvantage of digital communication.
(Apr-May 2011)
(i) Circuitry becomes simpler than analog communication.

(ii) Highly resistive to noise as binary signals are dealt.
(iii) Highly secure.
(iv) Can travel long distances and can store data for long time.
3. Which parameter is called figure of merit of a digital communication

system and why? (Nov-Dec 2010)
The ratio of bit energy to noise power spectral density Eb /N0 is called
as the figure of merit of a digital communication system. This is because
in digital communication system a symbol is transmitted and received
by using a transmission window within a symbol tree. Since power goes
to zero, symbol energy is a more useful parameter. So a energy related
parameter in terms of bit level is required to compare different systems.
4. Draw the basic block diagram of digital communication system.
(Nov-Dec 2011)
5. What is the function of formatter in digital communication system?

It converts electrical signals at the output of the transducer into a
sequence of digital signals.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 230 — #38

i i
6. What is meant by distortion less transmission? (Nov-Dec 2010)

If the overall system response has constant magnitude and if its phase
shift is linear with frequency then that is said to be distortion less
transmission.
7. How is the performance of digital communication system evaluated?
The performance of digital communication system is evaluated depend-
ing upon (i) spectral efficiency and (ii) bit error rate (BER).
8. Define Nyquist rate.
Let the signal be bandlimited to W Hz. Then Nyquist rate is given as,
Nyquist rate = 2W samples/sec
Aliasing will not take place if sampling rate is greater than Nyquist rate.
9. What is meant by aliasing effect? (May/June 2016) (Nov/Dec 2016)
Aliasing effect takes place when sampling frequency is less than Nyquist
rate. Under such condition, the spectrum of the sampled signal overlaps
with itself. Hence higher frequencies take the form of lower frequencies.
This interference of the frequency components is called as aliasing effect.
10. What is meant by temporal waveform coding? (Nov-Dec 2011)
Temporal waveform coding is a scheme in which time domain wave-
form is encoded. Bit allocation depends upon time domain features. Bit
rate is high compared to signal bandwidth and reconstruction is perfect.
11. Compare uniform and non-uniform quantization. (Nov-Dec 2011)
In uniform quantization, the step size or the difference between two
quantization levels remain constant over the complete amplitude range.
In non-uniform quantization step size varies.
12. What are the two types of quantization that occur in delta modula-
tion? (Apr-May 2010)
(i) Slope overload distortion and (ii) Granular noise.
13. Write an expression for bandwidth of binary PCM with N messages
each with a maximum frequency of fm Hz.
If v number of bits are used to code each input sample, then bandwidth
of PCM is given as,
BT ≥ N.v.fm
Here v.fm is the bandwidth required by one message.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 231 — #39

i i
14. What is meant by quantization? (May-June 2012)

While converting the signal value from analog to digital, quantization
is performed. The analog value is assigned to nearest digital value.
This is called quantization. The quantized value is then converted into
equivalent binary value. The quantization levels are fixed depending
upon the number of bits. Quantization is performed in every Analog
to Digital conversion.
15. The signal to quantization noise ratio in a PCM system depends on
what criteria?
The signal to quantisation noise ratio in PCM is given as,
(S/N)dble(4.8 + 6v)dB
Here v is the number of bits used to represent samples in PCM. Hence

signal to quantization noise ratio in PCM depends upon the number of
bits or quantization levels.
16. What should be the minimum bandwidth required to transmit a
PCM channel?
The minimum transmission bandwidth in PCM is given as,
BT = vW
Here v is the number of bits used to represent one pulse. W is the

maximum signal frequency.
17. What is the advantage of delta modulation over PCM?
Delta modulation uses one bit to encode on sample. Hence bit rate of
delta modulation is low compared to PCM.
18. Why is prefiltering done before sampling? (Apr-May 2011)
Prefiltering is done so that the new maximum frequency is reduced to
fs /2 or less. Thus there will be no aliased components.
19. State the advantages and disadvantages of digital communication
systems over analog communication systems. (May-June2013).
(May/June 2014)
Advantages:
• In digital communication, the speech, video and other data may be
merged and transmitted over a common channel using multiplexing
technique.
• The digital communication system are simpler and cheaper because
of advancement made in the IC technologies.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 232 — #40

i i
• Channel coding is used in digital communication that why it reduces

the amount of errors in the detector and correct them in the receivers.
• As the transmitted signals are digital in nature thus the amount of
interference is controlled in this form of communication.
Disadvantages:
• Due to Analog to Digital conversion the data rate become high.

Therefore more transmission bandwidth is required for digital com-
munication. This is the major disadvantage of digital communication.
• Synchronization is required in digital communication during the
process of synchronous modulation.
20. State sampling theorem for band limited signals and the filter to
avoid aliasing. (Nov/Dec 2015) (May-June 2012)
A band limited signal of finite energy, which has no frequency com-
ponents higher than W Hz, may be completely recovered from the
knowledge of its samples taken at the rate of 2W samples per second.
Low pass anti-aliasing filter is used to avoid aliasing.
21. What is natural sampling? (May-June2013)
Natural sampling is a practical method of sampling in which pulse (rect-
angular) have finite width equal to τ . Sampling is done in accordance
with the carrier signal which is digital in nature.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 233 — #41

i i
22. What is difference between natural and flat top sampling? (Nov-Dec
2014)
Flat top sampling is like natural sampling i.e., practical in nature. In
comparison to natural sampling flat top sampling can be easily obtained.
In this sampling techniques, the top of the samples remains constant and
is equal to the instantaneous value of the message signal x(t) at the start
of sampling process. Sample and hold circuit are used in this type of
sampling.
23. What is difference between uniform and non-uniform quantization.

Uniform Quantization Non Quantization
The quantization step size The quantization step size varies
remains same throughout the with the amplitude of the input
dynamic range of the signal. signal.
SNR ratio varies with input signal SNR ratio can be maintained
amplitude. constant.
24. What is the need for non-uniform quantization? (May-June2014)
In uniform quantization, step or difference between two quantization
levels remains constant over the complete amplitude range. So, the
maximum quantization error also remains same, which causes problems
at some amplitude levels. In such cases, non uniform quantization is
preferred.
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 234 — #42

i i
25. Write the two fold effects of quantization process. (Nov/Dec 2015)
1. The peak-to-peak range of input sample values subdivided into a
finite set of decision levels or decision thresholds.
2. The output is assigned a discrete value selected from a finite set of
representation levels are reconstruction values that are aligned with
the treads of the staircase.
26. Define companding. Sketch the input-output characteristics of a
compressor and an expander (May/June 2016) (Nov/Dec 2016)
(Apr/May 2017)
Companding refers to a technique for compressing and then expanding
(or decompressing) an analog or digital signal. It is a combination of
the words “compressing” and “expanding.” Companding is the process
of compressing and expanding. With companded systems, the higher
amplitude analog signals are compressed prior to transmission and then
expanded at thereceiver.
27. A certain lowpass bandlimited signal x(t) is sampled and the spec-
trum of the sampled version has the first guard band from 1500 Hz
to 1900Hz. What is the sampling frequency? What is the maximum
frequency of the signal? (Apr/May 2017)
Sampling frequency = fs = 2fm (max)

= 2 × 1900 = 3800 Hz
Maximum frequency = fm (max) = 1900 Hz
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 235 — #43

i i
Part B
1. Draw a neat block diagram of a typical digital communication system
and explain the function of each blocks. (Nov-Dec 2010, May-June
2012)
2. Discuss the advantages and disadvantages of digital communication
and give a functional description of a digital communication system.
(Nov-Dec 2014)
3. Draw a neat block diagram of a typical digital communication system
and explain the function of the key signal processing blocks (May-June
2013)
4. (i) Draw the block diagram of a digital communication system. Explain
each block. (ii) Discuss the advantages of digital communication over
analog communication. (May-June 2014)
5. Define basis set. (i) How they are useful in representing a signal? (ii)
With an example. Explain how the basis set is determined by gram
schemidt procedure. (May-June 2014)
6. Explain non-uniform quantization process. (Apr-May 2011)
7. State and prove Nyquist sampling theorem. (Nov-Dec 2010)
8. Explain non-uniform quantization. (Apr-May 2010)
9. Explain the operation of PCM and delta modulation systems. (Apr-May
2010)
10. Explain what is natural sampling and flat-top sampling. (May-June
2012)
11. With neat block diagram, explain pulse code modulation and demodu-
lation system. (May-June 2012) (May/June 2016)
12. Explain the working of differential PCM and hence derive the expres-
sion of signal to noise ratio. (Apr-May 2010)
13. Explain a PCM system. Derive the expression for quantization noise of
a PCM system with uniform quqntizer. (May-June 2013)
14. Compare any two speech encoding techniques. (May-June 2013)
15. (i) Explain Nyquist sampling theorem and how the message can be
reconstructed from its samples with an example. (ii) Explain the
practical limitations in sampling and reconstruction. (May-June 2013)
16. (i) Explain non-uniform quantization techniques. (ii) Explain temporal
waveform encoding technique. (May-June 2014)
i i
i i
i i
“QA” — 2018/12/17 — 15:02 — page 236 — #44

i i
17. (i) Explain the principle of quantization and obtain the expression for
the signal to quantization noise for the case of a uniform quantizer. (ii)
Explain the spectral waveform encoding methods. (May-June 2013)
18. Write notes on temporal waveform coding. (Apr-May 2011)
19. Describe the process of sampling and how the message signal is
reconstructed from it’s samples. Also illustrate the effect of aliasing with
neat sketch. (Nov/Dec 2015)
20. Describe PCM waveform coder and decoder with neat sketch and list
the merits compared with analog coders. (Nov/Dec 2015)
21. Illustrate and describe the types of quantizer? Describe the midtread
and midrise type characteristics of uniform quantizer with a suitable
diagram. (Nov/Dec 2016)
22. Draw and explain the TDM with its applications. Explain the difference
between analog TDM and digital TDM. (Nov/Dec 2016) (May/June
2016)
23. State the low pass sampling theorem and explain reconstruction of the
signal from its samples. (May/June 2016)
24. The signal
x(t) = 4 cos 400π t + 12 cos 360π t
is ideally sampled at a frequency of 300 samples per second. The
sampled signal is passed through a unit gain low pass filter with a cut
off frequency of 220 Hz. List the frequency components present at the
output of the low pass filter? (May/June 2016)
25. What is mean by quantization? Derive the expression for
signal-to-quantization noise ratio in PCM system. (Apr/May 2017)
26. The information in an analog signal with maximum frequency of 3
kHz is required to be transmitted using 16 quantization levels in PCM
system. Determine (1) The maximum number of bits/sample that
should be used (2) the minimum sampling rate required and (3) the
resulting transmission data rate. (Apr/May 2017)
27. (1) Explain the following terms with respect to sampling: (i) Alising and
(ii) Aperture effect distortation.
(2) Explain time division multiplexing system for N number of chan-
nels. (Apr/May 2017)
i i
i i

Comm Theory

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Comm Theory

Hochgeladen von

Copyright:

Verfügbare Formate

i i

“FM” — 2018/12/17 — 16:26 — page v — #1

World has come very close due to rapid advancements in communication

“FM” — 2018/12/17 — 16:26 — page vi — #2

“FM” — 2018/12/17 — 16:26 — page vii — #3

Dr. M.P. Chitra

“FM” — 2018/12/17 — 16:26 — page viii — #4

“FM” — 2018/12/17 — 16:26 — page ix — #5

“FM” — 2018/12/17 — 16:26 — page x — #6

1.2 Double Sideband-Suppressed Carrier System 30

“FM” — 2018/12/17 — 16:26 — page xi — #7

2.1.4 Time-Domain Representation of FM 62

“FM” — 2018/12/17 — 16:26 — page xii — #8

3.2 Stationary and Non-stationary Process 95

“FM” — 2018/12/17 — 16:26 — page xiii — #9

4.2.5 Shot Noise 119

“FM” — 2018/12/17 — 16:26 — page xiv — #10

5.3 Sampling (Discretization in Time) 151

“FM” — 2018/12/17 — 16:26 — page xv — #11

5.8.3 Pulse Modulator 189

University Questions 193

“FM” — 2018/12/17 — 16:26 — page xvi — #12

“Unit-1” — 2018/12/15 — 16:11 — page 1 — #1

BASIC COMMUNICATION SYSTEM

“Unit-1” — 2018/12/15 — 16:11 — page 2 — #2

Modulation is the process of changing the characteristics of high frequency

CLASSIFICATION OF MODULATION PROCESS

Analog Modulation Digital Modulation

Continuous Wave Pulse Modulation

PAM PWM PPM

Digital pulse Keying

PCM DM DPCM ASK FSK PSK

NEED FOR MODULATION/ADVANTAGES OF

Modulation is extremely necessary in communication system due to the

“Unit-1” — 2018/12/15 — 16:11 — page 3 — #3

(iii) Multiplexing is possible.

(i) Line (or) Wired Communication:

FREQUENCY RANGE IN COMMUNICATION SYSTEMS

Types of Signal Frequency Range Application

“Unit-1” — 2018/12/15 — 16:11 — page 4 — #4

“Unit-1” — 2018/12/15 — 16:11 — page 5 — #5

1.1.2 Mathematical Expression

em (t) = Em sin 2π fm t (1.1)

ec (t) be the carrier signal,

ec (t) = Ec sin 2π fc t (1.2)

“Unit-1” — 2018/12/15 — 16:11 — page 6 — #6

6 . Unit 1: Amplitude Modulation

EAM = Ec + Em sin 2π fm t (1.3)

where ma = EEmc is modulation index (or) depth of modulate. Therefore the

eAM (t) = EAM sin 2π fc t (1.5)

Substitute eqn. (1.4) in (1.5),

eAM (t) = Ec [1 + ma sin 2π fm t] · sin 2π fc t

Figure 1.1: AM waveform.

“Unit-1” — 2018/12/15 — 16:11 — page 7 — #7

AM wave has a time-varying amplitude called as the envelope of the AM

1.1.4 Modulation Index (ma )

1.1.5 Percentage of Modulation

1.1.6 Calculation of Modulation Index from an AM Wave

2Em = Emax − Emin

Substituting the value of Em in above eqn.

“Unit-1” — 2018/12/15 — 16:11 — page 8 — #8

8 . Unit 1: Amplitude Modulation