Digital Communications 3units

Sri Sai Aditya Institute of Science & Technology
ECE Department
JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY KAKINADA

III Year B. Tech. Electronics and Communication Engineering I Sem.
DIGITAL COMMUNICATIONS
UNIT I-PULSE DIGITAL MODULATION:
Elements of digital communication systems, advantages of digital communication systems,
Elements of PCM: Sampling, Quantization & Coding, Quantization error, Compading in PCM
systems. Differential PCM systems (DPCM).
UNIT II-DELTA MODULATION :
Delta modulation, its draw backs, adaptive delta modulation, comparison of PCM and DM
systems, noise in PCM and DM systems.
UNIT III-DIGITAL MODULATION TECHNIQUES :
Introduction, ASK, FSK, PSK, DPSK, DEPSK, QPSK, M-ary PSK, ASK, FSK, similarity of BFSK and
BPSK.
UNIT IV-DATA TRANSMISSION :
Base band signal receiver, probability of error, the optimum filter, matched filter, probability
of error using matched filter, coherent reception, non-coherent detection of FSK, calculation
of error probability of ASK, BPSK, BFSK,QPSK.
UNIT V-INFORMATION THEORY :
Discrete messages, concept of amount of information and its properties. Average information,
Entropy and its properties. Information rate, Mutual information and its properties,
UNIT VI-SOURCE CODING :
Introductions, Advantages, Shannons theorem, Shanon-Fano coding, Huffman coding,
efficiency calculations, channel capacity of discrete and analog Channels, capacity of a
Gaussian channel, bandwidth S/N trade off.
UNIT VII-LINEAR BLOCK CODES :
Introduction, Matrix description of Linear Block codes, Error detection and error correction
capabilities of Linear block codes, Hamming codes, Binary cyclic codes, Algebraic structure,
encoding, syndrome calculation, BCH Codes.
UNIT VIII-CONVOLUTION CODES :
Introduction, encoding of convolution codes, time domain approach, transform domain
approach. Graphical approach: state, tree and trellis diagram decoding using Viterbi
algorithm.
TEXT BOOKS :
1. Digital communications - Simon Haykin, John Wiley, 2005
2. Principles of Communication Systems H. Taub and D. Schilling, TMH, 2003
REFERENCES :
1. Digital and Analog Communication Systems - Sam Shanmugam, John Wiley, 2005.
2. Digital Communications John Proakis, TMH, 1983. Communication Systems Analog &
Digital Singh & Sapre, TMH, 2004.
3. Modern Analog and Digital Communication B.P.Lathi, Oxford reprint, 3rd edition,2004.
Prepared by Venkata Satish N
Page 1
ECE Department
UNIT I-PULSE DIGITAL MODULATION

ELEMENTS OF A DIGITAL COMMUNICATION SYSTEM
The analysis and design of digital communication systems Involves the
transmission of information in digital form from a source that generates the information
to one or more destinations.
The source output may be either an analog signal, such as an audio or video signal, or
a discrete signal, such as the output of a teletype machine, that is discrete in time and has
a finite number of output characters.
In a digital communication system, the messages produced by the source are
converted into a sequence of binary digits. The process of efficiently converting the
output of either an analog or discrete source into a sequence of binary digits is called
source encoding or data compression.
The sequence of binary digits from the source encoder, which we call the information
sequence, is passed to the channel encoder
FIGURE 1. Basic elements of a digital communication system.

. The purpose of the channel encoder is to introduce, in a controlled manner, some
redundancy in the binary information sequence that can be used at the receiver to
Page 2
ECE Department
overcome the effects of noise and interference encountered in the transmission of the
signal through the channel. This increases the reliability of the received data and
improves the fidelity of the received signal.
The binary sequence at the output of the channel encoder is passed to the digital
modulator, which serves as the interface to the communication channel. Since nearly all
the communication channels encountered in practice are capable of transmitting electrical
signals (waveforms), the primary purpose of the digital modulator is to map the binary
information sequence into signal waveforms.
To elaborate on this point, let us suppose that the coded information sequence is to be
transmitted one bit at a time at some uniform rate R bits per second (bits/s). The digital
modulator may simply map the binary digit 0 into a waveform So(t) and the binary digit 1
into a waveform S1(t). In this manner, each bit from the channel encoder is transmitted
separately. We call this binary modulation.
The communication channel is the physical medium that is used to send the signal
from the transmitter to the receiver. In wireless transmission, the channel may be the
atmosphere (free space). On the other hand, telephone channels usually employ a variety
of physical media, including wire lines, optical fiber cables, and wireless (microwave
radio).
Whatever the physical medium used for transmission of' the information, the essential
feature is that the transmitted signal is corrupted in a random manner by a variety of
possible mechanisms, such as additive thermal noise generated by electronic devices;
man-made noise, e.g., automobile ignition noise; and atmospheric noise,
e.g., electrical lightning discharges during thunderstorms.
At the receiving end of a digital communication system, the digital demodulator
processes the channel-corrupted transmitted waveform and reduces the waveforms to a
sequence of numbers that represent estimates of the transmitted data symbols. This
sequence of numbers is passed to the channel decoder, which attempts to reconstruct the
original information sequence from knowledge of the code used by the channel encoder
and the redundancy contained in the received data.
A measure of' how well the demodulator and decoder perform is the frequency with
which errors occur in the decoded sequence. More precisely, the average probability of a
Page 3
ECE Department
bit-error at the output of the decoder is a measure of the performance of the demodulator
decoder combination.
In general, the probability of error is a function of the code characteristics, the types of
waveforms used to transmit the information over the channel, the transmitter power, the
characteristics of the channel, and the method of' demodulation and decoding.
The source decoder accepts the output sequence from the channel decoder and, from
knowledge of the source encoding method used attempts to reconstruct the original
signal.
Because of channel decoding errors and possible distortion introduced by the source
encoder, and perhaps, the source decoder, the signal at the output of the source decoder is
an approximation to the original source output. The difference or some function of the
difference between the original signal and the reconstructed signal is a measure of the
distortion introduced by the digital communication system.
The points worth noting are:
The source coding algorithm plays important role in higher code rate
The channel encoder introduced redundancy in data
The modulation scheme plays important role in deciding the data rate and immunity of
signal towards the errors introduced by the channel
Channel introduced many types of errors like multi path, errors due to thermal noise etc.
Page 4
ECE Department
ADVANTAGES OF DIGITAL COMMUNICATION OVER ANALOG MODULATION:

There are many advantages of using Digital Communication over analog
communication. Some of them are listed as below:
1. The Digital communication has mostly common structure of encoding a signal so
devices used are mostly similar.
2. The Digital Communication's main advantage is that it provides us added security to
our information signal.
3. The Digital Communication system has more immunity to noise and external
interference.
4. Digital information can be saved and retrieved when necessary while it is not possible
in analog.
5. Digital Communication system is cheaper than Analog Communication.
6. The configuring process of digital communication system is simple as compared to
analog communication system.
7. In Digital Communication System, the error correction and detection techniques can be
implemented easily.
8. Digital hardware implementation is flexible & permits the use of microprocessors,
digital switching elements & layer scale.
9. Digital systems are relatively less expensive than analog systems
10. Transmission rate can be changed easily.
11. Easy for processing and applying multiplexing techniques.
Page 5
ECE Department
ANALOG-TO-DIGITAL CONVERSION
A digital signal is superior to an analog signal because it is more robust to noise
and can easily be recovered, corrected and amplified. For this reason, the tendency today
is to change an analog signal to digital data. In this section we describe two techniques,
pulse code modulation and delta modulation.
PULSE CODE MODULATION (PCM)
Definition: Pulse code modulation (PCM) is essentially analog-to-digital conversion of a

special type where the information contained in the instantaneous samples of an analog
signal is represented by digital words in a serial bit stream.
PCM consists of three steps to digitize an analog signal:
1. Sampling
2. Quantization
3. Binary encoding
Before we sample, we have to filter the signal to limit the maximum frequency of the
signal as it affects the sampling rate.
Filtering should ensure that we do not distort the signal, ie remove high frequency
components that affect the signal shape.
PCM Transmitter
Page 6
ECE Department
SAMPLING
Analog signal is sampled every TS secs. Ts is referred to as the sampling interval.

fs = 1/Ts is called the sampling rate or sampling frequency. According to the Nyquist
theorem, the sampling rate must be at least 2 times the highest frequency contained in the
signal. There are 3 sampling methods:
Ideal - an impulse at each sampling instant
Natural - a pulse of short width with varying amplitude
Flat top - a pulse of short width with constant amplitude
Usually Flat top sampled signal is generated by sampler.
Three different sampling methods for PCM
Nyquist sampling rate for low-pass and bandpass signals
Page 7
ECE Department
QUANTIZATION
Sampling results in a series of pulses of varying amplitude values ranging between

two limits: a min and a max. The amplitude values are infinite (or many) between the two
limits. We need to map the infinite amplitude values onto a finite set of known values.
This is achieved by dividing the distance between min and max into q zones, each of
height
rangeofinputsignal
noofQuantiztionlevels
vmax vmin
q
The midpoint of each zone is assigned a value from 0 to q-1 (resulting in q values). Each
sample falling in a zone is then approximated to the value of the midpoint. That is
quantization is a process of rounding-off each sampled value to the nearest value.
The reason for approximating to the mid point is that minimizes the maximum
quantization error.
Example:
Assume we have a voltage signal with amplitudes Vmin= -20V and Vmax=+20V.
We want to use q=8 quantization levels. Then zone width = (20 - -20)/8 = 5
The 8 zones are: -20 to -15, -15 to -10, -10 to -5, -5 to 0, 0 to +5, +5 to +10, +10 to +15,
+15 to +20
The midpoints are: -17.5, -12.5, -7.5, -2.5, 2.5, 7.5, 12.5, 17.5
Each zone is then assigned a binary code.
The number of bits required to encode the zones, or the number of bits per sample as it is
commonly referred to, is obtained as follows: v = log2 q
Hence no of bits required to represent each sample are v = 3
The 8 zone (or level) codes are therefore: 000, 001, 010, 011, 100, 101, 110, and 111
Assigning codes to zones: 000 will refer to zone -20 to -15; 001 to zone -15 to -10, etc
Page 8
ECE Department
Note: If suppose quantization levels are 16 (2v), no of bits required to represent each
sample are 4 (v bits).
If no of quantization levels are not in the power of 2, for example to distinguish 10 > 23 (
> 2v ) quantization levels, 4 bits (v+1) are required. Possible no of 4 bit code words are
16, use any 10 code words out of 16 for representing samples.
Quantization error is defined as the difference between actual sample and quantized
sample.
i.e x(nTs ) xq (nTs )
Page 9
ECE Department
TYPES OF QUANTIZERS
1. Uniform Quantizers
Types: a) Symmetrical type of mid rise quantizer
b) Symmetrical type of mid tread quantizer
2. Non uniform Quantizers
Uniform Quantization
Most ADCs use uniform quantizers.
The quantization levels of a uniform quantizer are equally spaced apart.
Uniform quantizers are optimal when the input distribution is uniform ie when all
values within the Dynamic Range of the quantizer are equally likely.
Symmetrical type of mid rise quantizer
a) Symmetrical type of mid rise Quantizer b) Quantization error
Page 10
ECE Department
Origin lies in the middle of a rising part of the staircase graph like. Note that in mid rise
type, any input value in between 0 to is mapped to an output value of /2, any input
value between to 2 is mapped to an output value of 3/2 and so on.
Mid rise characteristic is desirable because of symmetry and because it uses the 2 v levels
of a v bit coder efficiently. A disadvantage of this mid rise characteristic is that it cannot
represent a zero output level.
Symmetrical type of mid tread quantizer
a) Symmetrical type of mid tread Quantizer b) Quantization error

Origin lies in the middle of a tread of a staircase like graph. Note that in mid tread type,
any input value in between -/2 to + /2 is mapped to an output value of zero, any input
value between + /2 to 3 /2 is mapped to an output value of and so on.
Unfortunately, this characteristic has an odd number of levels (if it is symmetric) or it
must be non symmetric about zero. Therefore it does not use the 2 v possible levels of a v
bit coder efficiently.
Page 11
ECE Department
Illustration of Quantization process for an analog signal & discrete time signal and
error signal in the approximations
Fig.: (a) An analog signal and its quantized version (b) The error signal
Fig: (c) Equispaced samples of m(t ) (d) Quantized sample sequence

Page 12
ECE Department
NOISE IN PCM SYSTEMS

The performance of a PCM system is influenced by two major sources of noise.
1. TRANSMISSION NOISE: It is introduced anywhere between the transmitter output
and the receiver input. The effect of transmission noise is to introduce bit errors into
the received PCM wave, with the result that, in case of a binary system, a symbol 1
occasionally is mistaken for a symbol 0, or vice versa. Clearly, the more frequently
such errors occur, the more dissimilar the receiver output becomes compared with the
original message signal.
2. DISTORTION DUE TO QUANTIZING
There are two types of distortions associated with a quantizer:

1. Overload or clipping distortion: Overload distortion occurs when the input signal
exceeds the quantizer's input range, then output will remain at its maximum (or
minimum) value until the input falls within the quantizer's input range. Overload
distortion results in a clipped output signal. To avoid clipping, a quantizer is
matched to the input signal.
2. Quantization distortion: Figure below shows the error signal introduced by the
quantizer. From this figure, it can be seen that quantization error occurs when the
input signal is within the input range of the quantizer. It arises because of the
difference between the input amplitude and the quantized sampled amplitude and
because of the limited sampling rate. The quantization error signal produces
quantization noise or distortion in the reconstructed message signal. Its frequency
spectrum covers a large bandwidth. Low-pass filtering which is used to smooth the
waveform will remove most of the quantization error above its cutoff frequency.
However, some of the quantization error is in the signal band, and that cannot be
removed by the low-pass filter. This will produce a gritty sound at the output of a
PCM system called quantization noise.
Figure: Characteristic of quantization and overload errors.

Page 13
ECE Department
Quantization noise is the result of the quantization process. Since the quantization
process adjusts the height of each sample, the original waveform cannot be exactly
reconstructed using a low-pass filter as is the case with PAM signals and the classical
sampling theorem. The sampling rate will also affect the quantization noise since the
quantization error will become larger as the sampling rate decreases.
Figure below shows an analog input signal and its quantized waveform. Shown
below this is the resulting quantization error signal. The maximum amplitude of this error
signal is half a quantization interval. The overall amplitude variation is from half a
quantization interval to minus half a quantization interval. During a period of small
intervals, the error signal appears to be a sawtooth wave.
Figure: Analog input signal, quantized waveform, and quantization error waveform.
Quantization error is another reason for using compressed encoding for digitizing
a voice signal. Compressed encoding allows a higher signal-to-quantization-noise ratio
(SNQR) than linear encoding. This ratio defined as where S is the voice signal level and
NQ is noise due to the quantization error. Clearly, keeping the quantization error small is
key to keeping a high SNQR. As signal amplitude gets smaller, NQ must get smaller to
keep SNQR from dropping. Compression accomplishes this by forcing quantization error
magnitude to decrease with lower amplitudes.
Page 14
ECE Department
Illustration of how quantization error is reduced by increasing quantization levels

When a signal is quantized, we introduce an error since the coded signal is an
approximation of the actual amplitude value. The difference between actual and coded
value (midpoint) is referred to as the quantization error.
The more zones, the smaller which results in smaller errors. But, the more zones, the
more bits required to encode the samples which leads higher bit rate.
Example:
In the above example, increasing the no of quantization levels from 5 to 10, decreases the
step size by 2, there by decreases the quantization error.
Page 15
ECE Department
NON UNIFORM QUANTIZING
In this step size of the quantizer is not fixed over entire input range and it varies
according to the input signal. i.e step size of the quantizer is reduced at low levels and
increased at high levels.
Non uniform Quantizer of 8 levels
Importance of Non uniform Quantization: Voice signals are more likely to have
amplitudes near zero than at extreme peaks.. Signals with lower amplitude values will
suffer more from quantization error as the error range: /2, is fixed for all signal levels.
Non linear quantization is used to alleviate this problem. The Goal is to keep SNQR fixed
for all sample values.
Two approaches for obtaining Non uniform Quantization:
Direct approach:
The quantization levels follow a logarithmic curve. Smaller s at lower amplitudes and
larger s at higher amplitudes. But this process of varying directly is very difficult.
Page 16
ECE Department
Indirect approach:
An Effect of non linear quantizing can be can be obtained by first passing the
sample values through a compressor at the sender, then through a uniform quantizer. This
technique increase amplitudes near zero. To compensate the effects happened at the
sender, pass the sample values through an expander at the receiver. The process of
compression, uniform quantization and expansion is called Companding.
A-law and -law Companding
These two are standard companding methods.
-Law is used in North America and Japan
A-Law is used elsewhere to compress digital telephone signals
Page 17
ECE Department
Two types according to compression filter
-law : used in US
ln(1 x )
ln(1 )
sgn( x)
A-law : used in Europe

Ax
sgn( x),
0 x 1
A
1 ln A
y
1 ln( A x ) sgn( x), 1 x 1
1 ln A
A
A-law & -law compression curve
Similarities between Alaw and law

Both are linear approximations of logarithmic input/output relationship.
Both are implemented using eightbit code words (256 levels, one for each
quantization interval).
Eightbit code words allow for a bit rate of 64 kilobits per second (kbps). This is
calculated by multiplying the sampling rate (twice the input frequency) by the size
of the code word (2 x 4 kHz x 8bits = 64 kbps).
Both break a dynamic range into a total of 16 segments
Page 18
ECE Department
Differences Between Alaw and law

Different linear approximations lead to different lengths and slopes.
Alaw provides a greater dynamic range than ulaw.
ulaw provides better signal/distortion performance for low level signals than
Alaw.
Alaw requires 13bits for a uniform PCM equivalent. ulaw requires 14bits for
a uniform PCM equivalent
SNR of Compander
Example: -law Companding
Page 19
ECE Department
ENCODING
The output of the quantizer is one of qpossible signal levels. If we want to use
a binary transmission system, then we need to map each quantized sample into a v bit
binary word.
Encoding is the process of representing each quantized sample by an bit code
word. The mapping is one-to-one so there is no distortion introduced by encoding. Some
mappings are better than others.
A Gray code gives the best end-to-end performance.
With gray codes adjacent samples differ only in one bit position.
The weakness of Gray codes is poor performance when the sign bit (MSB) is
received in error.
Example (3 bit quantization):

With this gray code, a single bit error will result in an amplitude error of only 2.
Unless the MSB is in error.
There are several ways by which binary symbols 1 and 0 can be represented by electrical
signals:
Unipolar NRZ (on-off signaling): Symbol 1 is represented by transmitting a pulse of
constant amplitude for the duration of symbol, and symbol 0 is represented by switching
off the pulse. This type of signal is referred to as an on-off signaling or Unipolar non
return to zero.
Polar NRZ: Symbols 1 and 0 are represented by pulses of equal positive and negative
amplitudes. This type of signal is referred to as a polar Non Return to Zero signal.
Page 20
ECE Department
Unipolar RZ: A rectangular pulse (half symbol wide) is used for a 1 and no pulse for a
0. This type of signal is called Unipolar Return to zero.
Bipolar RZ: Positive and negative pulses are used alternatively for symbol 1, and no
pulse for symbol 0. This type of signal is called a bipolar signal.
Manchester or Split-phase code: Symbol 1 is represented by a positive pulse followed
by a negative pulse, with both pulses being of equal amplitude and half-symbol wide; for
symbol 0, the polarities of these pulses are reversed. This type of signal is called a split
phase or Manchester code.
Electrical representations of binary data

Reasons and advantages of different encodings will be discussed in UNIT 4
Page 21
ECE Department
Bit rate of PCM

The bit rate of a PCM signal can be calculated form the number of bits per sample x the
sampling rate. i.e. Bit rate = v x fs
Bandwidth requirements of PCM
The bandwidth of (serial) binary PCM waveforms depends on the bit rate R and the
waveform pulse shape used to represent the data.
For no aliasing case (fs 2B), the MINIMUM Bandwidth of PCM is:
Bpcm(Min) = R/2 = vfs//2.
The Minimum Bandwidth of vfs//2 is obtained only when sin(x)/x pulse is used to
generate the PCM waveform.
For PCM waveform generated by rectangular pulses, the First-null Bandwidth is:
Bpcm = R = nfs
A digitized signal will always need more bandwidth than the original analog signal. Price
we pay for robustness and other features of digital transmission.
EXAMPLE: DESIGN OF A PCM SIGNAL FOR TELEPHONE SYSTEMS
Assume that an analog audio voice-frequency (VF) telephone signal occupies a band
from 300 to 3,400Hz. The signal is to be converted to a PCM signal for transmission over
a digital telephone system. The minimum sampling frequency is 2x3.4 = 6.8 ksample/sec.
To be able to use of a low-cost low-pass anti aliasing filter, the VF signal is oversampled
with a sampling frequency of 8ksamples/sec. This is the standard adopted by the Unites
States telephone industry. Assume that each sample values is represented by 8 bits; then
the bit rate of the binary PCM signal is Bit rate = v x fs = 8 x 8k = 64k bit/sec
This 64-kbit/s signal is called a DS-0 signal (digital signal, type zero).
The minimum absolute bandwidth of the binary PCM signal when sin(x)/x pulse is
used to generate is Bpcm(Min) = R/2 = vfs//2 = 32k bit/sec
If we use a rectangular pulse for sampling the first null bandwidth is given by
Bpcm(Min) = R = vfs = 64k bit/sec
We require a bandwidth of 64 kHz to transmit this digital voice PCM signal, whereas the
bandwidth of the original analog voice signal was, at most, 4 kHz.
Page 22
ECE Department
APPLICATIONS OF PCM
With the advent of fibre optic cables, PCM is used in telephony.
In space communication, space craft transmits signal to earth. Here the ransmitted
power is quite small and the distances are very large.Hence due to high noise
immunity, only pcm systems can be used in such applications.
ADVANTAGES OF PCM
Relatively inexpensive digital circuitry may be used extensively.
PCM signals derived from all types of analog sources may be merged with
data signals and transmitted over a common high-speed digital
communication system.
In long-distance digital telephone systems requiring repeaters, a clean

PCM waveform can be regenerated at the output of each repeater, where
the input consists of a noisy PCM waveform.
The noise performance of a digital system can be superior to that of an

analog system.
The probability of error for the system output can be reduced even further
by the use of appropriate coding techniques.
DRAW BACKS OF PCM
Encoding, Decoding and quantizing circuitry of PCM is complex

PCM requires a large bandwidth as compared to other systems
Page 23
ECE Department
QUANTIZATION ERROR/NOISE IN PCM

sample.
If the step size of the quantizer (mid rise or mid tread) is , then maximum quantization
error max is

and the range of quantization error is , .
2
2 2

As the error is equally likely in the range , , it is better to assume error as uniform
2 2
random variable. The probability density function of this error is given by

f ( )
1
1

2 2
Mean Square value of this quantization error (Noise power) is given by
1
2
E[ ] f ( )d
d
12
SIGNAL TO QUANTIZATION NOISE RATIO IN PCM

Case 1: input signal is sinusoidal signal x(t ) Am sin mt
Am2
SNR = Signal Power (rms) / Quantization noise power = 22
12
Where
2A
2A
rangeofinputsignal
m vm
q
2
SNR = Signal Power / Quantization noise power
Am2
= 22 =
12
Am2
3 2 BWpcm fm
3 2v
2
2
or
2
2
2
2
2 Am
v
2
12
3
SNR in decibels = 10log10 ( 22v ) 1.76 6v
2
Page 24
ECE Department
Case 2: Input signal is a DC signal ranging between Am to +Am

SNR= Signal Power / Quantization noise power =
Where
SNR =
Am2
2

12
2A
2A
rangeofinputsignal
m vm
q
2
Am2
Am2
=
3* 22 v
2
2
2 Am
v
12 2
12
SNR in decibels = 10log10 (3*22v ) 4.76 6v .

Note: Signal to noise ratio of PCM system improved by 6db for every one bit increase.
PCM TRANSMISSION PATH & REGENERATION

The path between the PCM transmitter and PCM receiver over which the PCM signal
travel, is called as PCM transmission path. The most important feature of PCM system
lies in its ability to control the effects of distortion and noise when the PCM signal travels
on the channel. This capability is accomplished by reconstructing the PCM wave by
means of a chain of regenerative repeaters located at sufficiently close spacing along the
transmission route.
There are three basic functions are performed by a regenerative repeater, namely
1. Equalization
2. Timing
3. Decision Making
The equalizer shapes the received pulses so as to compensate for the effects of amplitude
and phase distortions produced by the transmission characteristics of the channel.
The timing circuitry provides a periodic pulse train, derived from the received pulses, for
sampling the equalized pulses at the instants of time where the signal to noise ratio is a
maximum.
Page 25
ECE Department
The decision making device makes a decision in the favor of 1, if the equalized pulse plus
noise is above the threshold level and it makes a decision in the favor of 0 if the equalized
pulse plus noise is below the threshold level.
PCM RECEIVER
The first operation in the receiver is to regenerate the received pulses. These clean
pulses are then regrouped into code words and decoded into a quantized PAM signal. The
decoding process involves generating a pulse the amplitude of which is the linear sum of
all the pulses in the code word, with each pulse weighted by its place value in the code.
The final operation in the receiver is to recover the signal wave by passing the
decoder output through a low-pass reconstruction filter whose cutoff frequency is equal
to the message bandwidth W. Assuming that the transmission path is error free, the
recovered signal includes no noise with the exception of the initial distortion introduced
by the quantization process
Page 26
ECE Department
TIME DIVISION MULTIPLEXING PAM SYSTEM

Normally, in PAM system, the duration of the pulse is much less than the time
period of pulses Ts. Thus no information is being transmitted through the system for most
of the time. The time space Ts can be utilized to transmit information from other
signals. The signal numbers 2, 3 and 4 are transmitting information with the help of
samples numbered 2, 3 and 4 respectively. This is along with the samples numbered 1 of
the signal number 1.
The time period Ts is equally divided between the four signals, thus allocating a
time slot of
to each signal. Thus the duration of time slot is such that
there is a guard time
> . Thus,
- between all successive sampling pulses, ensuring that there is
less cross talk between signals. The arrangement by which the information from more
than one signal is transmitted in this manner is known as time division multiplexing.
A TDM PAM system is shown in figure below, which transmits information from
n signals. The switch 1 and switch 2 respectively known as commutator and
decommutator are synchronized electronic switches which rotate at the same speed of 2f M
rotations per second. The commutator samples and combines the samples, while the
decommutator seperates the samples belonging to individual signals.
Page 27
ECE Department
Synchronization is the most crucial in TDM system. Thus, for example, if the
commutator is at position 2, the decommutator must also be in position 2. To provide
synchronization, a synchronizing puse is transmitted in every frame (time interval
between two successive samples of the same signal, i.e Ts).
Thus to multiplex n channels, n+1 time slots are provided in a frame; n for
channels and 1 for the synchronizing pulse. The synchronizing pulse is chosen in such a
way that it is easily distinguishable. For this purpose, one of its properties is adjusted in
such a way that it is never attained by the other pulses. For example, in case of PAM, its
amplitude is made larger than the amplitudes of all the other pulses.
Page 28
ECE Department
DIFFERENTIAL PULSE CODE MODULATION

When a voice or video signal is sampled at a rate slightly higher than the nyquist
rate, the resulting sampled signal is found to exhibit a high correlation between adjacent
samples. The meaning of this high correlation is that, in an average sense, the signal does
not change rapidly from one sample to next with the result that the difference between
adjacent samples has a variance that is smaller than the variance of the signal itself.
When these highly correlated samples are encoded, as in standard PCM system,
the resulting encoded signal contains redundant information. By removing this
redundancy before encoding, we obtain a more efficient coded signal.
For example, we can observe that the samples taken at 4T s, 5Ts and 6Ts are
encoded to same value of 110. This information can be carried only by one sample. But
three samples are carrying the same information means that it is redundant. Cosider
another example of samples taken at 9T s and 10Ts. The difference between these samples
only due to last bit and first two bits are redundant, since they do nit change.
If this redundancy is reduced, then overall bit rate will decrease and number of
bits will decrease and number of bits required to transmit one sample will also be
reduced.
Redundant information in PCM

Page 29
ECE Department
DPCM TRANSMITTER
DPCM can be treated as a variation of PCM; it also involves the three basic steps
of PCM, namely, sampling, quantization and coding. But, in the case of DPCM, what is
quantized is the difference between the actual sample and its predicted value, as
explained below.
Let x(t) represent the analog signal that is to be DPCM coded, and let it be
sampled with a period Ts. The sampling frequency fs = 1/Ts is such that there is no
aliasing in the sampling process. Let x(nTs) = m(t) at t= nTs. Quite a few real world
signals such as speech signals, biomedical signals (ECG, EEG, etc.), telemetry signals
(temperature inside a space craft, atmospheric pressure, etc.) do exhibit sample-to-sample
correlation. This implies that x(n) and x(n + 1) (or x((n) and x (n 1)) do not differ
significantly. In fact, given a set of previous M samples, say x (n 1), x (n 2), x (n
M) , it may be possible for us to predict (or estimate) x (n) to within a small percentage
error.
DPCM transmitter
Let x (nTs ) denote the predicted value of x(nTs) and let e(nTs ) x(nTs ) x (nTs )
Which is the difference between the unquantized input sample m(nTs) and a prediction
of it, denoted by x (nTs ) . This predicted value is produced bu using a prediction filter
whose input, as we will see, consists of a quantized version of the input signal x (nTs).
Page 30
ECE Department
The difference signal e(nTs ) is called a prediction error, since it is the amount by which
the prediction filter fails to predict the input exactly.
In DPCM, error sequence is quantized, coded and obtained a variation of PCM,
which is known as differential pulse code modulation.
The quantizer output may be expressed as eq (nTs ) e( nTs ) qe ( nTs ) , where
qe (nTs ) is the quantization error.
According to fig, the quantizer output eq (nTs ) is added to the predicted value
x (nTs ) to produce the prediction filter input xq (nTs ) x (nTs ) eq (nTs ) .

xq (nTs ) x (nTs ) e(nTs ) qe (nTs )
xq (nTs ) x(nTs ) qe (nTs )
That is irrespective of the properties of the prediction filter, the quantized signal
xq (nTs ) at the prediction filter input differs from the original input signal x(nTs ) by the
quantizing error qe (nTs ) . Accordingly if prediction is good, the variance of the prediction
error eq (nTs ) will be smaller than the variance of x(nTs ) .
DPCM RECEIVER
The receiver for reconstructing the quantized version of the input is shown in figure. It
consists of a decoder to reconstruct the quantized error signal. The quantized version of
the original input is reconstructed from the decoder output using the same prediction
filter as used in the transmitter.
DPCM receiver
Page 31
ECE Department
Output Signal to Noise ration of the DPCM system:

By definition,
(SNR)o =Variance of the input signal/Variance of the quantized noise =
X2
Q2
X2 E2
*
E2 Q2
Where E2 is the variance of the prediction error.
X2 E2
(SNR)o = 2 * 2 = Gp * prediction error to quantization noise ratio.
E Q
Where Gp is the predictive gain. This prediction gain must be high as possible. This Gp is
maximized by minimizing the variance E2 of the prediction error.
THE PREDICTION FILTER
The predicted value x (nTs ) is modeled as a linear combination of past values of the
quantized input as shown below
p
x (nTs ) wk xq (nTs kTs )

k 1
Where the tapped delay line weights w1, w2, w3wp define the desired prediction
filter coefficients and p is order of the prediction filter.
Tapped delay line filter used as prediction filter

Page 32
ECE Department
The prediction error e(nTs ) x(nTs ) wk xq (nTs kTs )

k 1
The variance of the prediction error is therefore E2 = E[e2 (nTs )]

p
E[e2 (nTs )] E[{x(nTs ) wk xq (nTs kTs )}2 ]

k 1
In order to choose a set of weights that minimize the variance E2 , we must differentiate
E2 with respect to each weight and then put the resulting derivatives equal to zero.
ADVANTAGES OF DPCM
1. As the difference is being encoded and transmitted by the DPCM technique, a

small difference voltage is to be quantized and encoded.
2. This will require less number of quantization levels and hence less number of bits
to represent them
3. Thus signaling rate and bandwidth of a DPCM system will be less than that of
DPCM.
COMPARISON BETWEEN PCM AND DPCM
Parameter
of Pulse code modulation
Differential
comparison
Number of bits
Pulse
Code
Modulation
It can use 4, 8 or 16 bits per sample
Bits can be more than one but

are less than PCM
Quantization error
Quantization
error
depends
on
Quantization error is present
number of levels used.

Transmission
Highest bandwidth is required since Bandwidth requires is lower
Bandwidth
number of bits are high
Feed back
There is no feedback in transmitter Here, feedback exists
than PCM
and receiver
Complexity
of System Complex
Simple
implementation
Signal to noise ratio
Good
Fair
Applications
Audio and video telephony
Speech and video
Page 33
ECE Department
Important Questions
1. a) Draw the block diagram of PCM scheme. Explain each block?
b) The bandwidth of a TV radio plus audio signal is 4.5MHz. If this signal is converted to PCM
with 1024 quantizing levels. Determine the bit rate of the resulting PCM signal. Assume that the
signal is sampled at a rate 20% above the Nyquist rate.
2 .a) The signal m(t)=6sin(2t) volts is transmitted using a 4-bit binary PCM system. The
quantizer is of the midrise type, with a step size of 1 volt . Sketch the resulting PCM wave for one
complete cycle of the input. Assume a sampling rate of four samples per second, with samples
taken at t=1/8, t=3/8, t=5/8.seconds.
b)What is quantization error? How does it depend upon the step size? Suggest some methods to
overcome the difficulties encountered when the modulating signal amplitude swing is large.
3. a)A PCM system uses a uniform quantizer followed by a 7-bit binary encoder. The bit rate of
the system is equal to 50x106b/sec.
(i) What is the maximum message bandwidth for which the system operates satisfactorily?
(ii) Determine the output signal to quantization noise ratio when a full load sinusoidal
Modulating wave of frequency 1MHz is applied to the input.
(b) Explain the importance of prediction in DPCM & draw the structure of Prediction filter?
4. (a) Show that in a PCM system, the output signal power to quantization noise
3 BWpcm fm
power ratio can be expressed as S/NQ= 4
. Where BWpcm is the channel bandwidth and
2
fm is the message bandwidth.
(b) Draw and explain different ways of representing binary data by electrical signals?
5. (a) Explain - law and A - law companding technique?

(b)Draw the block diagram of DPCM system and explain each block.
or
Explain the process of differential quantizing scheme and compare differential quantizing
scheme with direct quantizing scheme.
(c) Discuss the advantages & Drawbacks of PCM system?
Page 34
ECE Department
Communication Systems- Simon haykin 4th edition Exercise Problems
Problem1: A speech signal has a total duration of 10 s. It is sampled at the rate of 8 kHz
and the encoded. The signal to Quantization noise ratio is required to be 40db. Calculate
the minimum storage capacity needed to accommodate this digitized speech signal.
Solution: The minimum number of bits per sample is 7 for a signal to quantization noise
ratio of 40 dB. Hence
The number of samples in a duration of 10 s = 8000*10 = 8*104 samples.
The minimum storage is therefore = 7*8*104 = 560 Kbits
Problem 2 : A PCM system uses a uniform quantizer followed by a 7 bit binary encoder.
The bit rate of the system is equal to 50 106 b / s .
(a) What is the maximum message bandwidth for which the system operates
satisfactorily?
(b) Determine the output signal to quantization noise ratio when a full- load
sinusoidal modulating wave of frequency 1 MHz is applied to the input.
Solution: (a)
Bit rate of the PCM system is given by R vf s
For the system to operate satisfactorily, sampling rate must be atleast equal to the
nyquist rate. Hence R vf s v 2 f max
50 106 b / s = 7 2 f max
f max = 3.57*106 Hz
(c) The output signal to Quantizing noise ratio is given by SNR in dB= 1.8 + 6v=
43.8 dB
Page 35
ECE Department
Problem 3: (a) A sinusoidal signal, with an amplitude of 3.25 volts, is applied to a

uniform quantizer of the mid rise type whose output takes on the values 0, 1, 2,3
volts. Sketch the waveform of the resulting quantizer output for one complete cycle of the
input.
(b) Repeat this evaluation for the case when the quantizer is of the midrise type
whose output takes on the values 0.5, 1.5, 2.5, 3.5 volts.
Page 36
ECE Department
Problem 4: The signal m (t) = 6sin (2t) volts is transmitted using a 4 bit binary PCM
system. The quantizer is of the midrise type, with a step size of 1 volt. Sketch the
resulting PCM wave for one complete cycle of the input. Assume a sampling per second,
with samples taken at t = 18, 3/8, 5/8,, seconds.
Solution:
Page 37
ECE Department
UNIT2- DELTA MODULATION

Delta modulation, like DPCM is a predictive waveform coding technique and can
be considered as a special case of DPCM. It uses the simplest possible quantizer, namely
a two level (one bit) quantizer. The price paid for achieving the simplicity of the
quantizer is the increased sampling rate (much higher than the Nyquist rate) and the
possibility of slope-overload distortion in the waveform reconstruction, as explained in
greater detail later on in this section. In DM, the analog signal is highly over-sampled in
order to increase the adjacent sample correlation. The implication of this is that there is
very little change in two adjacent samples, thereby enabling us to use a simple one bit
quantizer, which like in DPCM, acts on the difference (prediction error) signals. In its
original form, the DM coder approximates an input time function by a series of linear
segments of constant slope. Such a coder is therefore referred to as a Linear (or nonadaptive) Delta Modulator (LDM). Subsequent developments have resulted in delta
modulators where the slope of the approximating function is a variable. Such coders are
generally classified under Adaptive Delta Modulation (ADM) schemes. We use DM to
indicate either of the linear or adaptive variety.
LINEAR DELTA MODULATION
PRINCIPAL OF WORKING
The principle of operation of an LDM system can be explained with the help of
Fig 2.1 below. The signal x (t), band limited to W Hz is sampled at the rate f s 2W .
If x(nT s) denote the sample of x(t) at t= nTs. The staircase approximation to x(t), denote
by x (nTs ) is arrived as follows. One notes, at t=nTs, the polarity of the difference
between x(nTs) and the latest approximation to it; that is x (nTs ) at t= nTs.
The difference between the input and the previous approximation is quantized
into only two levels, namely, , corresponding to positive and negative differences,
respectively. Thus, if the approximation falls below the signal at any sampling epoch, it is
increased by . If on the other hand, the approximation lies above the signal, it is
diminished by . Provided that the signal does not change too rapidly from sample to
sample, we find that the staircase approximation remains within of the input signal.
Page 38
ECE Department
Figure 2.1 Illustration of Delta Modulation

If e(nTs ) 0
then eq (nTs ) & b(nTs ) 1
If e(nTs ) 0
then eq (nTs ) & b(nTs ) 0
DELTA MODULATION TRANSMITTER
The principal virtue of delta modulation is its simplicity. It may be generated by applying
the sampled version of the incoming baseband signal to a modulator that involves a
summer, quantizer and accumulator interconnected as shown in figure 2.2.
Fig.2.2 Delta Modulation Transmitter
Page 39
ECE Department
Denoting the input signal as x(t) and the staircase approximation as x q(t), the basic
principal of delta modulation may be formalized in the following set of discrete-time
relations.
e(nTs ) x(nTs ) xq (nTs Ts ) eq1
eq (nTs ) sgn(e(nTs )) eq 2
& xq (nTs ) x q (nTs Ts ) eq (nTs ) eq 3

where Ts is the sampling period; e(nT s) is an error signal representing the difference
between the present sample value x(nTs) of the input signal and the latest approximation
to it. Namely, x (nTs ) xq (nTs Ts ) ; and eq (nTs ) is the quantized version of e(nTS).
The quantized output eq (nTs ) is finally coded to produce the desired DM wave.
Figure 2.1 illustrates the way in which the staircase approximation xq (t ) follows
variations in the input signal x (t) in accordance with above equations and it also displays
the corresponding binary sequence at the delta modulator output
Working of Accumulator (Stair case wave form generator)

1. In particular, quantizer consists of a hard limiter with input and output relation
defined by eq2 which is depicted in fig 2.2.1. The quantizer output is applied to an
n
i 1
i 1
accumulator, producing the result xq (nTs ) sgn(e(iTs )) eq (iTs ) .
Fig 2.2.1: Input output characteristic of quantizer for DM system

2. Thus at the sampling instant nTs, the accumulator increments the approximation
by a step in a positive or negative direction, depending upon the algebraic sign
of error signal e(nTs).
Page 40
ECE Department
3. If the input signal x(nTs) is greater than the most recent approximation x (nTs ) , a
positive increment + is applied to the approximation.
4. If on other hand, the input signal is smaller, a negative increment - is applied to
the approximation.
5. In this way the accumulator does the best it can to track the input samples by one
step at a time.
DELTA MODULATION RECEIVER
In the receiver, the staircase approximation xq(t) is reconstructed by passing the sequence
of positive and negative pulses, produced at the decoder output, through an accumulator
in a manner similar to that used in the transmitter. Then pass this staircase waveform
through a low pass filter (with a bandwidth equal to Original signal bandwidth) to recover
the original signal.
Fig.2.3 Delta Modulation Receiver
In comparing the DPCM and DM networks, we note that they are basically similar,
except for two important differences, namely, the use of a one-bit quantizer in delta
modulator and the replacement of the prediction filter by a single delay element.
Page 41
ECE Department
QUANTIZING NOISE
Delta modulation systems are subject to two types of quantizing error.
(1) Slope overload distortion
(2) Granular Noise
SLOPE OVERLOAD DISTORTION:
This distortion arises because of large dynamic range of the input signal. The rate of
rise of input signal x(t) is so high that the staircase signal cannot approximate it. The
slope overload is said to occur when the step size is too small to follow steep
segment of the input waveform x(t). To reduce this error, the step size must be
increased when slope of the signal x(t) is high. Since the step size of delta modulator
remains fixed, its maximum or minimum slopes occur along straight lines. Therefore
this modulator is also known as Linear Delta Modulator.
Quantiztion errors in delta modulation for an arbitrary input

To reduce this slope overload distortion, the slope of the quantizer must be greater
than the maximum slope of the input signal.
Ie.
dm(t )
Ts
dt max imum
GRANULAR NOISE (IDLE NOISE):
Granularity, on other hand refers to a situation where the stair case function x (nTs )
hunts around a relatively flat segment of the input function, with a step size that is too
large relative to local slope characteristic of the input. This means that for very small
Page 42
ECE Department
variations in the input signal, the staircase signal is changed by large amount
because of large step size. The solution is to this problem is to make step size small.
Quantiztion errors in delta modulation for another input

BIT RATE (SIGNALING RATE) OF DELTA MODULATION
Delta Modulation bit rate (R) = Number of bits transmitted/seconds

= Number of samples/sec Number of bits/sample
= f s 1 = f s
Therefore, the delta modulation bit rate is (1/N) times the bit rate of a PCM system.
Where N is the number of bits per transmitted PCM codeword. Hence, we can say
that the channel bandwidth for the delta modulation system is reduced to a great
extent as compared to that for the PCM system.
ADVANTAGES OF DELTA MODULATION
1. Since the delta modulation transmits only one bit for one sample, therefore the
signaling rate and transmission channel bandwidth is quite small for delta
modulation compared to PCM.
2. The transmitter and receiver implementation is very much simple for delta
modulation. There is no analog to digital converter required in delta modulation.
Page 43
ECE Department
DRAWBACKS OF DELTA MODULATION
The Delta Modulation has two major drawbacks as under;

(i)
Slope overload distortion
(ii)
Granular Noise
SIMON HAYKIN Problem1: Given a sine wave of frequency fm and amplitude Am

applied to a delta modulator having step size . Show that the slope overload will
occur if Am
here Ts is the sampling period.

2 f mTs
Solution: Let us consider that the sine wave is represented as x(t ) Am sin(2 f mt )
Maximum slope of delta modulator is given as
.
Ts
We know that, the slope overload distortion will take place if slope of the sine wave
is greater than slope of delta modulator i.e., max
max
dx(t )
>
Ts
dt
dAm sin(2 f mt )
dt
Ts
max 2 f m Am cos(2 f mt )
2 f m Am >
Am
Ts
Ts
2 f mTs
Note:
To avoid slope overload distortion, the condition that must be satisfied is Am
2 f mTs
Page 44
ECE Department
QUANTIZATION ERROR/NOISE IN DELTA MODULATION
sample.
If the step size of the quantizer is , then maximum quantization error max is and the
range of quantization error is , .
As the error is equally likely in the range , , it is better to assume error as uniform
random variable. The probability density function of this error is given by
f ( )
1
1
Mean Square value of this quantization error (Noise power) is given by

E[ 2 ]
2
f ( )d
1
2
d
SIGNAL TO QUANTIZATION NOISE RATIO IN DELTA MODULATION
Case 1: input signal is sinusoidal signal x(t ) Am sin mt

Am2
SNR = Signal Power (rms) / Quantization noise power = 22
No slope overload distortion occurs, for Am
, then substituting into the above

2 f mTs
2
equation gives
Am2
2
2
f
T
m s
SNR = 2 =
2
2
3
3
This noise power
2
is uniformly distributed over the frequency band upto f s (which is
3
more than f m ). Then the output quantization power within the bandwidth f BWLPF is given
by Nq' f BWLPF
2
fs
Page 45
ECE Department
In the receiver, at the output of Low pass filter of Bandwidth

SNR =
2 2 f mTs
f
2
BWLPF
3
fs
f BWLPF
3 f s3
f BWLPF f
2
m
ADAPTIVE DELTA MODULATION

To reduce slope overload distortion, a large step size is required to accommodate
wide dynamic range of the input signal and small steps are required to reduce granular
noise. In fact, adaptive delta modulation is the modification to overcome these errors.
Finally, we should mention that a delta modulator may also be made adaptive,
wherein the variable step size increases during a steep segment of the input signal and
decreases when the modulator is quantizing an input signal with a slowly varying
segment. In this way the step size is adapted to the level of the input signal. The resulting
system is called an adaptive delta modulator.
The problem in adaptive delta modulation, of course, is to specify suitable rules
for step size variation. Figure below illustrates the operation of ADM.
Waveforms illustrative of ADM operation
Page 46
ECE Department
ADAPTIVE DELTA MODULATION TRANSMITTER:
The logic for step size control is added in the diagram. The step size increases or
decreases according to a specified rule depending on one bit quantizer output. As an
example, if one bit quantizer output is high (i.e. 1), then the step size may be doubled for
next sample. If one bit quantizer output is low, then step size may be reduced by one step.
ADAPTIVE DELTA MODULATION RECEIVER:
In the receiver of Adaptive delta modulator shown in figure, there are two portions. The
first portion produces the step size from each incoming bit. Exactly the same process is
followed as that in transmitter. The previous input and present input decides the step size.
Page 47
ECE Department
It is then applied to an accumulator which builds up staircase waveform. The low pass
filter then smoothens out the staircase waveform to reconstruct the original signal.
ADVANTAGES OF ADAPTIVE DELTA MODULATION
1. Signal to Noise ratio becomes better than ordinary delta modulation because of
the reduction in slope overload distortion and idle noise.
2. Because of the variable step size, the dynamic range of ADm is wider than simple
DM.
3. Bandwidth required for the transmission through channel is also less.
Results have been reported in the literature which compares the (SNR)o performance of
-law PCM and the ADM scheme discussed above. One such result is shown in Fig.
below for the case of band pass filtered (200-3200 Hz) speech. For PCM telephony, the
sampling frequency used is 8 kHz. As can be seen from the figure, the SNR comparison
between ADM and PCM is dependent on the bit rate. An interesting consequence of this
is, below 50 kbps, ADM which was originally conceived for its simplicity, out-performs
the logarithmic PCM, which is now well established commercially all over the world. A
60 channel ADM (continuous adaptation) requiring a bandwidth of 2.048 MHz (the same
as used by the 30 channel PCM system) was in commercial use in France for sometime.
French authorities have also used DM equipment for airborne radio communication and
air traffic control over Atlantic via satellite. However, DM has not found wide-spread
commercial usage simply because PCM was already there first!
Performance of PCM and ADM versus bit rate
Page 48
ECE Department
Communication Systems SIMON HAYKIN Exercise Problems:

Problem2: A linear delta modulator is designed to operate on speech signals limited to
3.4 kHz. The specifications of the modulator are as follows; Sampling rate = 10 fNyquist,
where is the Nyquist rate of the speech signal, Step size = 100 mV.
The modulator is tested with a 1-kHz sinusoidal signal. Determine the maximum
amplitude of this test signal required to avoid slope overload.
Solution: fs = 10 fNyquist =10 (2*3.4K) =68k Hz
dx(t )
<
dt
Ts
To avoid slope overload distortion max
max
dAm sin(2 f mt )
dt
Ts
max 2 f m Am cos(2 f mt )
2 f m Am <
Am
Therefore Am
Ts
Ts
2 f mTs
100m 68k
=
= 1.08 v
2 1k
2 f mTs
Problem3: Consider a test signal m(t) defined by a hyberbolic tangent function m(t)=
Atanh(t) where A and are constants. Determine the minimum step size for delta
modulation of the signal, which is required to avoid slope overload.
Solution: m(t)= Atanh(t)
To avoid slope overload distortion max
Or
dx(t )
<
Ts
dt
dx(t )
> max
Ts
dt
Page 49
ECE Department
> A sech2(t)
Ts
> AT s since the maximum value of sech(t) is 1 at t=0
Problem4: Consider a DM system designed to accommodate analog message signals

limited to bandwidth W= 5kHz. A sinusoidal test signal of amplitude A=1 volt and
frequency fm = 1 kHz is applied to the system. The sampling rate of the system is 50 kHz
(a) Calculate the step size required to minimize slope overload distortion.
(b) Calculate signal to Quantization noise ratio of the system for the specified
sinusoidal test signal.
Solution: (a) To avoid slope overload distortion > 2 f m Am
Ts
Therefore = 2 f m AmTs = 2 f m Am =
fs
2 1k 1v
=0.126 v
2 50k
(c) Signal to Quantization noise ratio

SNR
3 f s3
8 2 f BWLPF f
SNR in db = 10
2
m
3 (50k )3
= 475
8 2 5k (1k ) 2
475
log10
= 26.8 db
Page 50
ECE Department
Problem 5: Consider a low pass signal with a bandwidth of 3 kHz. A linear delta
modulation system with step size =0.1 v is used to process this signal at a sampling
rate ten times the nyquist rate.
(a) For linear delta modulation, the maximum amplitude of a sinusoidal test signal of
frequency 1 kHz which can be processed by the system without slope-overload
distortion.
(b)For the specifications given in part a, evaluate the output signal to noise ratio
under (i) prefilterd and (ii) postfiltered conditions
Solution: (a) For linear delta modulation, the maximum amplitude of a sinusoidal test
signal that can be used without slope overload distortion is
Am
f s
=
=
2 f mTs 2 f m
0.110 2 3k
=0.95 v
2 1k
(b) (i) Under the pre-filtered condition, it is reasonable to assume that the granular
quantization noise is uniformly distributed between and +. Hence the
variance of the quantization noise is
1
2
E[ ] f ( )d
d
When input signal is sinusoidal signal x(t) Am sin mt
Am2 0.952
2
2
SNR = 2 = 0.12 = 135= 21.3 db
3
3
(iii)
The signal to noise ratio under the post filtered condition is
3 f s3
SNR
8 2 f BWLPF f
2
m
3 (60k )3
=1367== 31.3 db
8 2 3k (1k ) 2
The filtering gain in signal to noise ratio due to the use of a reconstruction
filter at the demodulator output is therefore 31.3 db- 21.3 db= 10db.
Page 51
ECE Department
Problem 6: A linear delta modulator has a step size of 100 mV and the minimum output
amplitude is + 50 mV. A signal s(t) = 0.5 u(t) is applied to the input of the delta
modulator. Show how the modulator tracks the input indicating the distortions in the
waveform. Sketch the waveform for 12 clock cycles, beginning at least 2 clock cycles
before t = 0. Also, sketch the output waveform in NRZ format.
Solution:
Figure (a) below shows the sketch of the delta modulator input and the tracking
distortions.
The input is a step signal of amplitude 0.5 volts beginning at t = 0 as shown by the heavy
line. The input for t < 0 is 0 volts. Initial amplitude of the DM predictor, at clock instant
1, is assumed to be + 50 mV. The clock instants are shown in (b).
At the clock instant 2 the predictor output is higher than the input (0 V) and hence, a
negative step (-100mV) is added to the predictor output. At clock instant 3 the predictor
output is lower (- 50 mV) than the input (0.5 V) and hence, a positive step is added to the
Page 52
ECE Department
predictor. At clock instant 4 the predictor output (+50 mV) is still lower than the input.
Hence, a 100 mV step is added. At clock instants 4, 5, 6, 7, and 8 the predictor output is
lower than input and at each instant a 100 mV step is added to the previous predictor out.
At clock instant 9 the predictor output (550 mV) is found higher than the input. Hence, a
100 mV step is subtracted from the predictor output. At clock instant 10 a 100 mV step is
added. The DM output waveform is shown in figure (c)
Problem 7: A segment of a delta modulated data stream is a sequence given below.

01010111111000110001
This sequence is applied to a linear delta demodulator having a step size of 100 mV.
Assuming initial output of the demodulator is 0 V, show the output sample voltages at
each input bit and sketch the waveform.
Solution:
The output of demodulator is obtained by adding the step voltage for input 1 and
subtracting the step voltage for input 0. The output is shown in the table below for the
input bits given above.
Input
1 0
1 0
1 1
Output -0.1 0 -0.1 0 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.4 0.3 0.2 0.3 0.4
0.3
0.2
0.1
0.2
The waveform of the output is shown in figure below.
Page 53
ECE Department
Problem 8:
An instantaneously companded delta modulator employs the following step

size adaptation algorithm.
where Sk and Sk-1 are the current and previous step sizes, Bk and Bk-1 are the
current and previous output bits, Bk and Bk 1 have opposite polarity. The
minimum step size is 100 mV, so the amplitude of the steps when the input
is zero is 50 mV. If a step input x(t) = 1.2 V is applied to the modulator at
t=0 show how the predictor output tracks the input by sketching the
waveform. Sketch the binary output waveform of the delta modulator.
Solution:
We can show the step size Sk, predictor output Pk and modulator outputs for some clock
cycles in a tabular form as below. The input of 1.2 V is applied to the modulator at t = 0
and we start at t = -2.
The waveforms of the predictor output and the modulator output are plotted in the figure
below.
Page 54
ECE Department
Problem 9: A linear delta modulator is used to digitize speech signal band limited to 3.4
kHz. An output filter with 4 kHz cutoff frequency is used. Find the sampling frequency
required to get a performance equivalent to that of a 6-bit linear PCM coder. Compare the
information rates for PC and DM outputs.
Solution:
The S/N obtained from a linear PCM coder is
S
6v 1.8 6 6 1.8 37.8db
N max
3 f s3
The S/N obtained from a linear DM coder is SNR
8 2 f BWLPF f
2
m
3 f s3
In db we can write the above equation as SNR dB = 10 log
8 2 f BWLPF f
2
m
To get DM performance equivalent to PCM technique SNR of DM must be equal to SNR

of PCM.
Page 55
ECE Department
3
37.8 = 10log 2 30log f s 10log f BWLPF 20log f m
8
3
37.8 = 10log 2 30log f s 10log 4 103 20log 3.4 103
8
fs = 191 kb/s
Assuming 8kHz sampling the information rate for PCM data is R PCM=8*6=48kb/s
For DM the information rate is same as the sampling rate, hence R DM = 191 kb/s
Thus, DM requires approximately four times the data rate compared to 6-bit PCM for
similar performance.
Problem 10: A stereo music signal is sampled at 44.1 kHz and digitized with 16 bits for
recording on CD. If the CD stores 80 minutes of music find the total capacity of the CD
in bytes. What is the quality of the music if it has an RMS value 15 dB below the peak
value of the quantizer?
Solution:
We have fs = 44.1 kHz, n = 16 and number of channels is 2.
Hence, the bit rate
R = 2.n.fs = 2 x 16 x 44.1 x 103 = 1411.2 kb/s
The capacity of the CD is
C = 80 x 60 x 1411.2 kbits
= 6773760 kbits
= 846.72 Mbytes
The signal to noise ratio is
S/Nq = 6.n +1.8 -15 = 6 x16 +1.8 -15 = 82.8 dB
Page 56
ECE Department
Important Questions
1. a) Consider a test signal m(t)= A tanh(t) defined by a hyperbolic tangent function.
Where A and are constants. Determine the minimum step size for delta modulation of
this signal, which is required to avoid slope overload.
b) Comparison between PCM and DM.
2. a) A signal to transmitted is of the form S(t)=10COS 1000t+5COS1500t.

i) Choose an appropriate f and step size for delta modulator.
ii) Find the SNR for your design.
b) Draw the block diagram for Adaptive delta modulation system and explain each
block?
3. (a) Explain the noise effects in delta modulation

(b) A DM system is designed to operate at 3 times the Nyquist rate for a signal with a
3 KHz BW. The quantization step size is 250mv
(i) Determine the maximum amplitude of a 1 KHz sinusoid for which delta modulator
does not show slope overload.
(ii) Determine post filtered output SNR for the signal at part (i)
4. (a) Derive the condition for step size of the quantizer in a DM system to avoid slope
over load distortion for the message signal x(t) = A cos(wt)
(b) Explain the major drawback of the DM system with relevant waveforms.
5. A sinusoidal modulating signal is represented by m(t)= A cos (w mt) where wm=2fm.

Derive the expression for the maximum output signal to quantization noise ratio in the
DM system system with no slope overload distortion and also determine maximum
output signal to quantization noise ratio at the post receiver?
Page 57
ECE Department
UNIT 3- DIGITAL MODULATION TECHNIQUES

Modulation is defined as the process by which some characteristics of a carrier is
varied in accordance with a modulating signal. In digital communications, the modulating
signal consists of binary data or an M-ary encoded version of it. The data is used to
modulate a carrier wave (usually sinusoidal) with fixed frequency.
The modulation process involves switching or keying the amplitude, frequency or
phase of the carrier in accordance with the input data.
Thus there are three basic modulation techniques for the transmission of digital
data. They are known as amplitude-shift keying (ASK), frequency shift keying (FSK) and
phase shift keying (PSK).
If the amplitude of the carrier is switched depending on the input digital signal,
then it is called amplitude shift keying (ASK). This process is quite similar to analog
amplitude modulation.
If the frequency of the sinusoidal carrier is switched depending upon the input
digital signal, then it is known as the frequency shift keying. This is very much similar to
the analog frequency modulation.
If the phase of the carrier is switched depending upon the input digital signal, then
it is called phase shift keying. This is similar to phase modulation.
ASK, PSK & FSK waveforms (with sine as carrier signal)

Page 58
ECE Department
Since the phase and frequency modulation has constant amplitude envelope,
therefore FSK and PSK, the effect of non-linearities, noise interference is minimum on
signal detection. However, these effects are more pronounced on ASK. Therefore FSK
and PSK are preferred over ASK.
Figure shows the waveforms for amplitude-shift keying, phase shift keying and
frequency shift keying. In these waveforms, a single feature of the carrier (i.e. amplitude,
phase or frequency) undergoes modulation.
In digital modulations, instead of transmitting one bit at a time, we transmit two
or more bits simultaneously. This is known as M-ary transmission. This type of
transmission results in reduced channel bandwidth.
However, sometimes, we use two quadrature carriers for modulation. This process is
known as Quadrature modulation.
Thus we see that there are a number of modulation schemes available to the designer of a
digital communication system required for data transmission over a bandpass channel.
Every scheme offers system trade-offs of its own. In particular choice is made in favour
of a scheme which possesses as many of the following design characteristics as possible:
(i)
Maximum data rate
(ii)
Minimum probability of symbol error
(iii)
Minimum transmitted power
(iv)
Minimum channel bandwidth
(v)
Maximum resistance to interfering signals.
(vi)
Minimum circuit complexity.
DEFINITIONS AND TERMINOLOGY
There are basically two types of transmission of digital signals

i)
BASEBAND DATA TRANSMISSION:
The digital data is transmitted over the channel directly. There is no carrier or any
modulation. This is suitable for transmission over short distances.
A signal whose frequency content (i.e. its spectrum) is in the vicinity of zero (i.e.,
f = 0 or dc) is said to be a baseband signal.
Original source signal are sometimes said to be baseband. Baseband systems

transmit baseband signals.
Page 59
ECE Department
This is usually not an effective means of communication.
ii) PASSBAND (BAND PASS OR NARROW BAND) DATA TRANSMISSION: The digital data
modulates high frequency sinusoidal carrier. Hence it is also called digital CW
modulation. It is suitable for transmission over long distances.
Types of passband Modulation are ASK, PSK, FSK and etc.
Bandpass signal spectrum is nonzero in some band of frequency with BW = 2B

centered about f = fc, where fc >> 0.
Effective transmission of signal usually requires bandpass signal.
Bandpass transmission involves some translation of the baseband signal to some

band of frequency centered around fc.
Types of Reception for Passband transmission

There are two types of methods for detection of passband signals.
i) COHERENT (SYNCHRONOUS) DETECTION: In this method, the local carrier generated at
the receiver is phase locked with the carrier at the transmitter. Hence it is also called
synchronous detection.
ii) NON COHERENT (ENVELOPE) DETECTION: In this method, the receiver carrier no need
to be phase locked with transmitter carrier. Hence it is also called envelope detection.
Noncoherent detection is simple but it has higher probability of error.
Page 60
ECE Department
Types of Digital Modulation techniques (Classification based on envelope)
Types of Digital Modulation techniques (Classification based on coherence)

Page 61
ECE Department
DIGITAL MODULATION TECHNIQUES
As mentioned earlier, the binary (i.e. Digital) modulation has three basic forms
amplitude-shift keying(ASK), phase-shift keying(PSK) and frequency shift keying
(FSK).
BINARY AMPLITUDE SHIFT KEYING (ON-OFF KEYING)
Definition:
Amplitude shift keying (ASK) or ON-OFF keying (OOK) is the simplest digital
modulation technique. In this method, there is only one unit energy carrier and it is
switched on or off depending upon the input binary sequence.
a) Binary Modulating signal and b) BASK signal
a) Modulating signal b) spectrum of a c) Spectrum of BASK signal

Page 62
ECE Department
EXPRESSION:
s(t ) 2Ps cos(2 f ct )

s(t ) 0
To transmit symbol 1
To transmit symbol 0 i.e. no signal is transmitted.
Signal s(t) contains some complete cycles of carrier frequency fc.

Hence ASK waveform looks like an ON-OFF of the signal. Therefore it is also known as
the ON-OFF keying (OOK).
SIGNAL SPACE DIAGRAM (CONSTELLATION DIAGRAM) OF BASK
Study of signal spaces provides us with a geometrical method of conceptualizing the

modulation process.
The ASK waveform of equation for symbol 1 can be represented as,
s(t ) PT
s b
2
cos(2 f ct ) PT
s b 1 (t )
Tb
This means that there is only one carrier function 1 (t ) which is a unit energy signal over
(0, Tb). The signal space diagram will have two points on 1 (t ) . One will be at zero and
other will be at PT
s b . The collection of all possible signal points is called the signal
constellation.
Thus, the distance between the two signal points is d= PT
s b = Eb
The decision boundary is determined by the threshold value . If x lies in the region Z1,
then a decision of a 1 is made. If x lies in the region Z2, then a decision of a 0 is
made.
One advantage in using the signal space representation is that it is much easier to identify
the distance between signal points. The distance between two signal points will be
increased which makes the received signal point less probable be located in the wrong
region.
Page 63
ECE Department
GENERATION OF BASK SIGNAL
ASK signal may be generated by simply applying the incoming binary data and the
sinusoidal carrier to the two inputs of a product modulator. The resulting output will be
the ASK waveform.
Generation of BASK signal
BASK RECEPTION:
COHERENT DETECTION OR DEMODULATION OF BINARY ASK SIGNAL
The demodulation of BASK waveform can be achieved with the help of coherent detector
as shown in figure.
It consists of a product modulator which is followed by an integrator and Decision
making device. The incoming ASK signal is applied to one input of the product
modulator. The other input of the product modulator is supplied with a sinusoidal carrier
which is generated with the help of a local oscillator.
The output of the product modulator goes to input of the integrator. The integrator
operates on the output of the multiplier for successive bit intervals and essentially
performs a low-pass filtering action. The output of the integrator goes to the input of a
decision making device.
.
Coherent detection of Binary ASK signal

Page 64
ECE Department
Now, the decision making device compares the output of the integrator with a
preset threshold. It makes a decision in favour of symbol 1 when the threshold is
exceeded and in favour of symbol 0 otherwise.
In this method we assumed that the local carrier is in perfect synchronization with the
carriers used in the transmitter. This means that the frequency and phase of the locally
generated carrier is same as those of the carriers used in the transmitter
NON-COHERENT DETECTION OR DEMODULATION OF BINARY ASK SIGNAL
Figure shows the block diagram of noncoherent ASK receiver. In this figure the
received ASK signal is given to the band pass filter. This band pass filter passes only
carrier frequency, fo or fc.
The envelope detector generates high output voltage when carrier is present.
When carrier is absent, there is only noise at the input of envelope detector. Hence it
produces low output.
Non-Coherent Detection or Demodulation of Binary ASK Signal
The decision device is basically a regenerator. It generates the binary sequence

b(t). Threshold is provided to the decision device to overcome effects due to noise. When
y(t) is greater than threshold, b(t)= 1 and when y(t) is less than threshold, b(t)=0. Non
coherent reception of ASK does not need any carrier synchronization.
Page 65
ECE Department
Power Spectral Density

The ASK signal, which is basically the product of the binary sequence and the carrier
signal, has a power spectral density same as that of the baseband on-off signal but shifted
in the frequency domain by fc.
Power Spectral density of ASK Signal (One Sided)
Bandwidth of BASK
The Spectrum of the ASK signal shows that it has an infinite bandwidth. However for
practical purpose, the bandwidth is often defined as the range of frequency over which
ASK contains about 95% of the total average power content.
If suppose the center lobe of PSD contains 95% of the total power, then Bandwidth (Null
to Null) is given as 2Rb.
B= 2Rb =2/Tb
Advantages:
The advantage of using BASK is its simplicity. It is easy to generate and detect.
Drawback:
It is very sensitive to noise. Therefore it finds limited application in data transmission. It
is used at low bit rates, upto 100 bits per sec.
Page 66
ECE Department
BINARY PHASE SHIFT KEYING (BPSK)
Binary Phase Shift Keying is the most efficient of the three digital modulation
schemes. In Binary PSK, phase of the sinusoidal carrier is changed according to the data
bit to be transmitted. Also, a polar NRZ signal is used to represent the digital data coming
from the digital source.
Mathematically BPSK signal is expressed as
s(t ) b(t ) Ac cos(2 f ct ) Where b (t) is represented in bipolar NRZ format.
For Binary 1, b (t) = +1v then s(t ) Ac cos(2 f ct )

For Binary 0, b (t) = -1v
then s(t ) Ac cos(2 f ct ) = Ac cos(2 fct 1800 )
BPSK Waveforms (cosine as carrier)

Page 67
ECE Department
Spectrum of BPSK signals

Input b(t) is a NRZ binary waveform. In this waveform, there are rectangular pulses of
amplitude Ab. If we assume that each pulse is
Tb
around its centre, then it becomes
2
easy to find the fourier transform of such pulse. The fourier transform of this pulse is
given as X ( f ) AbTb
sin( fTb )
fTb
a) Modulating signal b) spectrum of a c) Spectrum of BPSK signal
For a large number of such positive and negative pulses, the power spectral density is
expressed as X avg ( f )
X(f )
Ts
Ab 2Tb
sin 2 ( fTb )
sin 2 ( fTb )
PT
=
b
( fTb )2
( fTb )2
Where T s is symbol duration. In this case T s = Tb. The above equation gives the power
spectral density of baseband signal b (t). Due to modulation of the carrier of frequency fc,
the spectral components are translated from f to fc+f and fc-f. The magnitude of these
components is divided by half.
Therefore the power spectral density of BPSK signal is given by
PTb
2
sin( ( f f c )Tb ) PTb sin( ( f f c )Tb )
2 ( f f c )Tb
( f f c )Tb
Page 68
ECE Department
Generation of BPSK signal

BPSK signal may be generated by applying carrier signal to a product modulator.
The binary data signal b (t) is converted into a NRZ bipolar signal by an encoder. Here
the bipolar signal b (t) is applied as modulating signal to the Product modulator. The
other input to the product modulator is supplied with a carrier signal.
RECEPTION OF BPSK SIGNAL-COHERENT DETECTION
The signal undergoes the phase change depending upon the time delay from
transmitter end to receiver end. This phase change is, usually, a fixed phase shift in the
transmitted signal.
Let us consider that this phase shift is . Because of this, the signal at the input of
the receiver can be written as s(t ) b(t ) 2Ps cos(2 f ct ) . Now, from this received
signal, a carrier is separated because this is coherent detection. The received signal is
allowed to pass through a square law device followed by a band pass filter and frequency
divider. Thus at the output of frequency divider, we get a carrier signal whose frequency
is fc i.e. cos(2 fct ) .
The synchronous demodulator multiplies the input signal and the recovered
carrier. Hence at the output of multiplier,
we get b(t ) cos2 (2 fct ) =
b(t )
{1 cos(4 f ct 2 )}
2
The output of multiplier contains a dc term
b(t )
and a high frequency signal of
2
frequency 2fc
Page 69
ECE Department
Block Diagram of BPSK Receiver
The signal is then applied to the bit synchronizer and integrator. The integrator
integrates the signal over one bit period and it essentially performs low pass filtering and
it produces a voltage proportional to b(t).
Tb
i.e.
b(t )
b(t )
Tb
{1 cos(4 f ct 2 )}dt =
2
2
The output of integrator goes through the decision making device. The decision
making device compares the output of integrator with the preset threshold. It makes a
decision in the favor of symbol 1 when the threshold is exceeded and in favor of symbol
0 otherwise.
The bit synchronizer takes care of starting and ending times of the bit. At the end
of the bit duration Tb, the bit synchronizer closes switch S2 temporarily. This connects the
output of an integrator to the decision making device. The synchronizer then opens the
switch S2 and switch S1 is closed temporarily. This resets the integrator voltage to zero.
The integrator then integrates next bit.
Page 70
ECE Department
SIGNAL SPACE DIAGRAM OF BPSK SIGNAL
We know that BPSK signal carries the information about two symbols. These symbols
are symbol 1 and symbol 0.
BPSK signal is given by s(t ) b(t ) Ac cos(wct )
b(t ) 2Ps cos(wct )

b(t ) PT
s b
2
cos( wct )
Tb
b(t ) PT
s b 1 (t )
Where 1 (t ) =
2
cos( wc t ) represents a unit energy signal over (0, Tb).
Tb
If b (t) is binary 1 (+1 v), then s(t ) PT

s b 1 (t )
If b (t) is binary 0 (-1 v), then s(t ) PT
s b 1 (t )
Thus on the single axis of 1 (t ) , there will be two points. One point will be located at
PT
s b and other will be located at PT
s b . Thus it has been shown in figure below.
Geometrical representation of BPSK signal
At the receiver end, the point at PT

s b on 1 (t ) represents symbol 1 and the point at
PT
s b on 1 (t ) represents symbol 0. The separation between these two points represents
the isolation in symbols 1 and 0 in BPSK signal. This separation is generally called
Page 71
ECE Department
Euclidean distance d. As the distance increases, the isolation between the symbols in
BPSk signal is more. Thus probability of error reduces.
Bandwidth for BPSK signal
BW= Highest frequency Lowest frequency in the main lobe

f c fb ( f c fb )
2 fb
Hence the minimum bandwidth of BPSK signal is equal to twice the highest frequency
contained in baseband signal.
Salient Features of BPSK

1. BPSK has a bandwidth which is lower than that of a BFSK signal.
2. BPSK has the best performance of all the three digital modulation techniques in
presence of noise. It yields the minimum value of probability of error.
3. Binary PSK has very good noise immunity.
Drawbacks
1. Providing synchronization in the reception of BPSK is very difficult.
2. Discontinuity of PSK signal creates high frequencies as side effects.
Page 72
ECE Department
BINARY FREQUENCY SHIFT KEYING
In BFSK, the frequency of the carrier is shifted according to the binary symbol. In
other words, the frequency of a sinusoidal carrier is shifted between two discrete
values. This means that we have two different frequency signals according to binary
symbols. However the phase of the carrier is unaffected.
If b (t) =1 then sH(t) =
2 Ps cos(2 f H t ) 2 Ps cos(2 ( f c
)t )
2
If b (t) =0 then sL(t) =
2 Ps cos(2 f Lt ) 2 Ps cos(2 ( f c
)t )
2
These equations combinely may be written as s(t ) 2 Ps cos(2 ( f c d (t)
) t) .
2
Where d(t) = + 1 for binary 1 and d(t)=-1 for binary 0.
GENERATION OF BFSK
It may be observed from table that PH(t) is same as b(t) and also PL(t) is inverted
version of b(t).
Input b(t)
d(t)
PH(t)
PL(t)
+1 v
+1 v
0v
-1 v
0 v
+1v
Conversion table used in the generation of BFSK

Page 73
ECE Department
The block diagram of BFSK generation is shown in figure below.

We know that input sequence b(t) is same as PH(t). An inverter is added after b(t) to
get PL(t). Then these PH(t) & PL(t) signals are applied to the product modulators
separately and carriers of frequencies fH & fL are also applied to product modulators
respectively. The outputs of multipliers are then applied to the adder. The adder
produces the BFSK signal.
BFSK Modulator
From the above block diagram, the expression for BFSK signal is given by
S (t ) 2Ps PH (t ) cos(2 f H t ) 2Ps PL (t ) cos(2 f Lt )

When b(t) =1; PH(t)=1 & PL(t)=0; S (t ) 2Ps cos(2 f H t )
When b(t)=0; ; PL(t)=1 & PH(t)=0; S (t ) 2Ps cos(2 f Lt )
BFSK signal
Page 74
ECE Department
COHERENT DETECTION OF BINARY FSK
The expression of BFSK is given by
S (t ) 2Ps PH (t ) cos(2 f H t ) 2Ps PL (t ) cos(2 f Lt )

The detectors consists of two correlators that are individually tuned to two
different carrier frequencies to represent symbols 1 and 0. A correlator consists of a
multiplier followed by an integrator. Then the received binary FSK signal is applied
to the multipliers of both the correlators. To the other input of the multipliers, carriers
with frequency fH & fL are applied as shown in figure below. The multiplied output of
each multiplier is subsequently passed through integrators generating outputs X 1 and
X2 in the two paths.
Coherent Detection of BFSK

The outputs of the two integrators are then fed to the decision making device. The
decision making device is essentially a comparator which compares the output X1 and
output X2.
If X1 > X2 then decision making device makes a decision in the favour of symbol 1.
If X1 < X2, then decision making device makes a decision in the favour of symbol 0.
Note: In this coherent detection process, it is required to generate synchronized
(coherent) carriers of frequencies fL & fH.
Note: Upper correlator recovers PH (t ) and lower correlator recovers PL (t ) . When upper
correlator recovers PH (t ) , lower correlator produces null output & vice versa.
Page 75
ECE Department
NON-COHERENT DETECTION OF BINARY FSK
Binary FSK waves may be demodulated non-coherently using envelope detector.

The received FSK signal is applied to a bank of two band pass filters, one tuned to
frequency fH & other tuned to fL. Each filter is followed by an envelope detector. The
resulting outputs of the two envelope detectors are sampled and the compared with
each other.
Non-Coherent Detection of Binary FSK
A decision is made in favour of symbol 1 if the envelope detector output derived from
the filter tuned to frequency fH is larger than that derived from the second filter.
Otherwise, a decision is made in favour of the symbol 0.
SIGNAL SPACE DIAGRAM OF BFSK SIGNAL
The BFSK signal is given S (t ) 2Ps PH (t ) cos(2 f H t ) 2Ps PL (t ) cos(2 f Lt )
S (t ) PT
s b PH (t )
2
2
cos(2 f H t ) PT
cos(2 f Lt )
s b PL (t )
Tb
Tb
S (t ) PT
s b PH (t )1 (t ) PT
s b PL (t )2 (t )
Page 76
ECE Department
Where 1 (t ) & 2 (t ) are orthogonal carriers (also unit energy carriers) over the period Tb.
Because in one bit interval of the input signal, 1 (t ) & 2 (t ) have integral number of
cycles.
i.e f H mfb & f L nfb .
If the carriers are orthogonal then the distance dmin is maximum. As probability of error
depends on dmin, maximizing the distance dmin decreases the error rate in BFSK
modulation scheme.
When symbol 1 is transmitted, modulated carrier BFSK has level of
symbol 0 is transmitted, modulated carrier BFSK has level of
PT
and when
s b
PT
s b .
Note that there are two signal points in the signal space. The distance between these two
points may be evaluated as under:
2
2
d 2 ( PT
s b ) ( PT
s b)
d 2 PT
s b 1.414 PT
s b
Signal Space diagram (Constellation) of BFSK signal
Note: dmin =
dmin =2
PT
s b in Binary ASK
PT
s b in Binary PSK
dmin = 1.414 PT
s b in Binary FSK
d min ( BPSK ) d min ( BFSK ) d min ( BASK )

Pe ( BPSK ) Pe ( BFSK ) Pe ( BASK )
Page 77
ECE Department
SPECTRUM OF BFSK SIGNALS
The BFSK signal is given S (t ) 2Ps PH (t ) cos(2 f H t ) 2Ps PL (t ) cos(2 f Lt ) . Let

us compare this equation with BPSK equation which is written below:
SBPSK=b(t) cos (2fct)
It may be noted that this equation is identical to BFSK equation. In BPSK equation,
b(t) is a polar signal where in BFSK, the similar coefficients P H(t) or PL(t) are
unipolar. Hence let us convert these coefficients in bipolar form as under:
1 1
PH(t) = PH 1 (t )
2 2
and PL (t )
1 1 1
PL (t )
2 2
Where PH 1 (t ) & PL1 (t ) will be bipolar (i.e., +1 or -1).

Substituting these values in the above equation, we obtain
1 1
1 1
S (t ) 2 Ps [ P I H (t )]cos(2 f H t ) 2 Ps [ P I L (t )]cos(2 f Lt )
2 2
2 2
S (t )
Ps
P
P
P
cos(2 f H t ) s P I H (t ) cos(2 f H t ) s cos(2 f Lt ) s P I L (t ) cos(2 f Lt )
2
2
2
2
Term
Fourier Transform s( f )
Power Spectral density
s( f )
cos(2 f H t )
( f fH )
( f fH )
P I H (t ) cos(2 f H t )
Sa( ( f f H )Tb )
Sa 2 ( ( f f H )Tb )
cos(2 f Lt )
( f fL )
( f fL )
P I L (t ) cos(2 f Lt )
Sa( ( f f L )Tb )
Sa 2 ( ( f f L )Tb )
Power spectral density of BFSK signal consists of impulses located at fL & fH and two
main lobes (sampling functions) located symmetrically about fL & fH of widths 2fb.
Page 78
ECE Department
Power Spectral Density of BFSK signal

Here the values fL & fH were chosen such that the bandwidth required by BFSK is
minimum. i.e f H f L 2 fb
BANDWIDTH OF BFSK SIGNAL
Width of one lobe is 2fb. the two main lobes due to fH & fL are placed such that the
total width due to both main lobes is 4 fb. Therefore, we have
Bandwidth of BFSK= 2 fb + 2 fb = 4 fb
if we compare this bandwidth with that of BPSK, we note that
BW (BFSK) = 2*BW (BPSK)
SALIENT FEATURES OF BFSK
i.
BFSK is relatively easy to implement
ii.
It has better noise immunity than ASK. Hence the probability of error free
reception of data is high.
DRAWBACK OF BFSK
The major drawback is its high bandwidth requirement. Therefore, FSK is

extensively used in low speed modems having bit rates below 1200 bits/sec.
Note: As there are no abrupt phase changes in BFSK, side lobe levels are minimum
in the spectrum of BFSK signal. Hence bandpass filtering is not required to remove
side lobes.
Page 79
ECE Department
COMPARISON BETWEEN BASK, BPSK & BFSK
Parameter of
Binary ASK
Binary FSK
Binary PSK
Comparison
Variable Characteristic
Amplitude
Frequency
Phase
Bandwidth
2 fb
4 fb
2 fb
Noise immunity
Low
Medium
High
Probability of error
High
Medium
Low
Performance in
Poor
Better than ASK
Best of three
System Complexity
Simple
Moderately Complex
Very Complex
Bit rate
Suitable upto
Suitable upto about
Suitable for
100 bits/sec
1200 bits/sec
high bit rates
Envelope (Non
Envelope(Non
Coherent
coherent )Detection
coherent )detection
detection
1.414 PT
s b
2 PT
s b
presence of noise
Demodulation method
Distance b/w signal
PT
s b
space points dmin
From the comparison table we can conclude that BPSK offers more advantages than
other modulation schemes. One difficulty lies in BPSK scheme is requirement of
coherent detector. Due to this, system complexity increases.
Page 80
ECE Department
DIFFERENTIAL PHASE SHIFT KEYING

DPSK does not need a synchronous carrier at the demodulator. The input sequence of
binary bits is modified such that the next bit depends upon the previous bit.
Therefore, in the receiver, the previous received bits are used to detect the present bit.
GENERATION OF DPSK
Thus in order to eliminate the need for phase synchronization of coherent receiver
with PSK, a differential encoding system can be used with PSK. The digital
information content of the binary data is encoded in terms of signal transitions.
A symbol 0 is used to represent transition in a given binary sequence and a
symbol 1 is used to indicate no transition.
This new signaling scheme which combines differential encoding with phase shift
keying (PSK) is known as differential phase shift keying.
Schematic Diagram
Block Diagram of DPSK Transmitter
The data stream b (t) is applied to the input of the encoder. The output of the encoder
(differential data) is applied to one input of the product modulator. To other input of this
product modulator, a sinusoidal carrier of fixed amplitude and frequency is applied.
The relation between the binary sequence and its differential encoded version is
illustrated in the following table for a assumed data sequence 0 0 1 0 0 1 0 0 1 1.
Page 81
ECE Department
Encoding has been done in such a way that transition in the given binary sequence
with respect to the previous encoded bit is represented by symbol 0 and no transition by
symbol 1. i.e., the logic function performed by encoder is XNOR logic function.
dk bk dk 1
It may be noted that an extra bit has been arbitrarily added as an initial bit. We
can choose 0 or 1 as an initial bit for the encoded sequence. The phase of the generated
DPSK signal has been shown in the third row of tables below.
Binary data bk
Differential encoded data d k
1*
Phase of DPSK
Phase of shifted DPSK signal
Phase comparison output
Detected binary sequence
Arbitrary starting reference bit 1
Binary data bk
Differential encoded data d k
0*
Phase of DPSK
Phase of shifted DPSK signal
Phase comparison output
Detected binary sequence
Arbitrary starting reference bit 0
Page 82
ECE Department
DETECTION OF DPSK
For the detection of the differentially encoded PSK, we can use the receiver
arrangement as shown in figure below.
Block Diagram of DPSK Receiver
The received DPSK signal is applied to one input of the multiplier. To other input
of the multiplier, a delayed version of the received DPSK signal by the time interval T b is
applied (it has been shown in 4th row of the table).
The output of the difference is proportional to cos , here is the difference
between the carrier phase angle of the received DPSK signal and its delayed version,
measured in the same bit interval. The phase angles of the DPSK signal and its delayed
version have been shown in 3rd and 4th rows respectively.
When =0, the integrator output is positive whereas when =, the integrator
output is negative. By comparing the integrator output with a decision level of zero volt.
The decision can reconstruct the binary sequence by assigning a symbol 0 for negative
output and symbol 1 for positive output. The reconstructed data is shown in the last row
of the table.
Reconstruction is invariant with the choice of the initial bit in the encoded data.
Bandwidth of DPSK signal
Bandwidth calculation is made by using power spectral density of it. The power spectral
density of DPSK signal is same as that of BPSK signal. The only difference b/w them is
the symbol duration. The symbol duration of BPSK scheme is T b where as for DPSK
scheme symbol duration is given by 2T b. As the generation of each symbol in differential
data depends on the present bit and previous encoded bit, the symbol duration is
equivalent to Ts = 2Tb.
Page 83
BW=
ECE Department
2
2
=
fb
Ts 2Tb
Hence minimum bandwidth in DPSK is equal to fb. i.e. maximum baseband signal
frequency.
ADVANTAGES
i.
DPSK does not need carrier at the receiver end. This means that the
complicated circuitry for generation of local carrier is not required.
ii.
The bandwidth requirement of DPSK is reduced as compared to that of BPSK.
DRAWBACKS
Because DPSK uses two successive bits for its reception, error in the first bit creates
error in the second bit. Therefore, error propagation in DPSK is more. On other hand, in
BPSK single bit can go in error since detection of each bit is independent.
i.
The probability of error of DPSK is higher than that of BPSK.
ii.
Noise interference in DPSK is more.
Note: In DPSK, previous bit is used to detect next bit. Hence, if error is present in
previous bit, detection of next bit can also go wrong. Hence error is created in next bit
also. Therefore, the tendency of appearing errors in pairs in DPSK.
Note: Geometrical representation of DPSK signal is same as that of BPSK.
Page 84
ECE Department
QUADRATURE PHASE SHIFT KEYING
In communication systems, we have two main resources. There are the

transmission power and the channel bandwidth. The channel bandwidth depends upon the
bit rate or signaling rate fb. if two or more bits are combined in some symbols, then the
signaling rate will be reduced. This reduces the transmission channel bandwidth. Hence
because of grouping of bits in symbols, the transmission channel bandwidth can be
reduced. In quadrature phase shift keying (QPSK), two successive bits in the data
sequence are grouped together. This reduces the bit rate or signaling rate and thus reduces
the bandwidth of the channel.
In case of BPSK, we know that when symbol changes the level, the phase of the
carrier is changed by 1800. Because, there were only two symbols in BPSK, the phase
shift occurs in two levels only.
However, in QPSK, two successive bits are combined. Infact, this combination of
two bits forms four distinct symbols. When symbol is changed to next symbol, then the
phase of the carrier is changed by 900.
GENERATION OF QPSK (OFFSET QPSK)
Figure below shows the block diagram of QPSK transmitter.
Input binary sequence is first converted to a bipolar NRZ type of signal. This
signal is denoted by b (t). It represents binary 1 by +1 V and binary 0 by -1 V.
In QPSK, we parallelize the bit stream so that every two incoming bits are split
up. The demultiplexer divides b (t) into two separate bit streams of the odd
numbered bits bo (t) and even numbered bits be (t) so that two successive bits of b
(t) can be applied to the modulator simultaneously. The symbol duration of both
of these odd and even numbered sequences is 2T b.
It may be observed that first even bit occurs after the first odd bit. Hence even
numbered bit sequence be (t) starts with the delay of one bit period due to the first
odd bit. This delay of Tb is known as offset. This shows that the change in levels
of be (t) and bo(t) cannot occur at the same time due to offset.
Page 85
ECE Department
Block Diagram of QPSK Transmitter
The bit stream bo (t) modulates the carrier Ps cos(2 f ct ) and be (t) modulates the
carrier Ps sin(2 f ct ) . These carriers are also known as quadrature carriers.
The two modulated signals can be written as,

Se (t) = be (t ) Ps sin(2 f ct )
So (t) = bo (t ) Ps cos(2 f ct )
Hence Se (t) & So (t) are basically BPSK signals. The only difference is that
T=2Tb here.
The output of the adder is QPSK signal and it is given by

S (t) =So (t) +Se (t)
s(t ) bo (t ) Ps cos(2 f ct ) be (t ) Ps sin(2 f ct )
Case 1: if bo (t ) 1(1V ) & be (t ) 0( 1V )

s(t ) 1 Ps cos(2 f ct ) (1) Ps sin(2 f ct )
s (t ) 2 Ps cos(2 f ct )
1
1
2 Ps sin(2 f ct )
2
2
s (t ) 2 Ps cos(2 f ct ) cos(450 ) 2 Ps sin(2 f ct ) sin(450 )

s (t ) 2 Ps cos(2 f ct 450 )
Page 86
ECE Department
Case 2: if bo (t ) 0(1V ) & be (t ) 0( 1V )

1
1
2 Ps sin(2 f ct )
2
2
3
3
s (t ) 2 Ps cos(2 f ct ) cos( ) 2 Ps sin(2 f ct ) sin( )
4
4
3
4
Page 87
ECE Department
Case 3: if bo (t ) 0(1V ) & be (t ) 1( 1V )

1
1
2 Ps sin(2 f ct )
2
2
5
5
4
4
5
4
Case 4: if bo (t ) 1(1V ) & be (t ) 1( 1V )

s(t ) 1 Ps cos(2 f ct ) 1 Ps sin(2 f ct )
1
1
2 Ps sin(2 f ct )
2
2
7
7
4
4
7
4
Symbol
Phase of the carrier
10
450
00
1350
01
2250
11
3150
Thus the phase shift in QPSK signal for symbol to symbol change is 900.
DETECTION OF QPSK
This is synchronous reception. Hence the coherent carrier is to be recovered from

the received signal S (t). The received signal S (t) is first raised to its 4 th power,
i.e., S4(t). After that, it is allowed to pass through a bandpass filter (BPF) which is
centered around 4fc. The output of the bandpass filter is a coherent carrier of
frequency 4fc. This is divided by 4 and it provides two coherent quadrature
carriers. i.e., cos(2 fct ) & sin(2 f ct) .
These coherent carriers are applied to two synchronous demodulators. These

synchronous demodulators consist of multiplier and an integrator.
Page 88
ECE Department
The incoming signal is applied to both the multipliers. Here, the integrator
integrates the product signal over two bit interval. The upper correlator recovers
the even ordered sequence and lower correlator recovers the odd ordered
sequence as follows.
Now, let us consider the product signal at the output of the upper multiplier, i.e.,
s(t )sin(2 fct ) {bo (t ) Ps cos(2 f ct ) be (t ) Ps sin(2 f ct )}sin(2 f ct )

This signal is integrated by the upper integrator over 2T b interval.
Therefore, we have
(2 k 2)Tb
s(t )sin(2 f ct )dt
2 kTb
(2 k 2)Tb
{bo (t ) Ps cos(2 f ct ) be (t ) Ps sin(2 f ct )}sin(2 f ct )dt
2 kTb
(2 k 2)Tb
bo (t ) Ps cos(2 f ct )sin(2 f ct )dt
2 kTb
2 kTb
be (t ) Ps sin 2 (2 f ct )dt
2 kTb
(2 k 2)Tb
(2 k 2)Tb
1
bo (t ) Ps sin(4 f ct )dt
2
(2 k 2)Tb
2 kTb
(2 k 2)Tb
2 kTb
1
be (t ) Ps {1 cos(4 f ct )}dt
2
1
be (t ) Ps dt be (t ) Ps Tb
2
Similarly odd sequence is recovered from lower multiplier and integrator combination.
These odd and even sequences are combined by multiplexer to generate binary sequence.
Page 89
ECE Department
SIGNAL SPACE DIAGRAM ( Phasor representation) OF QPSK SIGNAL
Symbol
If
QPSK signal
QPSK signal interms of unit energy signal
10
s(t ) 2Ps cos(2 f ct 450 )
00
s(t ) 2 Ps cos(2 f ct
01
11
s(t ) PT
s s{
2
cos(2 f ct 450 )}
Ts
3
)
4
s(t ) PT
s s{
2
cos(2 f ct 1350 )}
Ts
5
)
4
s(t ) PT
s s{
2
cos(2 f ct 2250 )}
Ts
7
)
4
s(t ) PT
s s{
2
cos(2 f ct 3150 )}
Ts
2
cos(2 f c t ) (unit energy carrier) is represented by phasor (t ) , then QPSK signal
Ts
can be geometrically represented as shown below.
Page 90
ECE Department
Minimum distance b/w signal space points dmin
2
2
( PT
s s ) ( PT
s s)
From the above signal space diagram d min =
2 PT
s s
=
For QPSK, Ts = 2Tb
dmin =
2 Ps 2Tb = 2
PT
s b
Thus dmin of QPSK is same as that of dmin of BPSK. It shows that noise immunities of
BPSK and QPSK are same.
SPECTRUM OF QPSK SIGNAL
In BPSK, the input sequence (Bipolar NRZ ) is of bit duration T b. Power spectral density
of such a waveform can be given as s(f)= S (t ) Vb 2Tb [
PT
s b[
sin( fTb ) 2
]
fTb
sin( fTb ) 2
]
fTb
Where Ps is the power of the signal over (0, T b)

In QPSK, the input sequences (Bipolar NRZ) of be (t) & bo (t) are of bit duration 2Tb.
Then spectrum of QPSK (shape) is similar to that of BPSK except in the width of
baseband signal spectrum (that is due to symbol duration).
Page 91
ECE Department
The power spectral densities of be (t) & bo (t) are given by

Se (t ) PT
s s[
sin( fTs ) 2
]
fTs
So (t ) PT
s s[
sin( fTs ) 2
]
fTs
The baseband power spectral density of QPSK signal equals the sum of the individual
power spectral densities of be (t) & bo (t) i.e.,
S H (t ) 2 PT
s s[
sin( fTs ) 2
]
fTs
Bandwidth for QPSK signal
Base band Power spectral density of QPSK signal

fc
1
1
( fc )
Ts
Ts
2
2
fb
Ts 2Tb
Hence the bandwidth of QPSK signal is equal to bit rate fb, where as in BPSK, BW is 2fb.
Advantages of QPSK
i.
For the same bit error rate, the bandwidth required by QPSK is reduced to half
as compared to BPSK.
ii.
Noise immunity is more
Page 92
ECE Department
iii.
Probability of error is less than BASK & BFSK signaling schemes
iv.
Noise interference is less
Disadvantages of QPSK
i.
Providing synchronization in the reception of QPSK is very difficult.
ii. Discontinuity of PSK signal creates high frequencies as side effects.

iii. An abrupt phase change in QPSK signal creates significant side lobes. Hence
band pass filtering is required to avoid inter channel interference. The solution to
this problem is MINIMUM SHIFT KEYING.
M-ary Transmission
In digital modulations, instead of transmitting one bit at a time, we transmit two
or more bits simultaneously. This is known as M-ary transmission. This type of
transmission results in reduced channel bandwidth.
COMPARISON of BPSK, QPSK & M-ary PSK
Parameter of
BPSK
QPSK
M-ary PSK
No of symbols
2 (symbol 0 &1 )
4 (Symbols 00, 01, 10,11)
M-Symbols
No of
1 bit
2 bits
N bits
the carrier
4
2
2
(M=2N)
M
Symbol
Ts =Tb
Ts =2Tb
Ts=NTb
Comparison
bits/symbol
Phase shift in
Duration
(2m 1)
)
4
Expression
s(t ) b(t ) Ac cos(2 f ct )
Bandwidth
2fb
fb
2fb/N
dmin
2 psTb
2 psTb
2 PT
s s sin(
Page 93
(2m 1)
)
M
ECE Department
M-ary PSK Transmitter
The serial to parallel converter forms a symbol of N successive bits. That is the
output of serial to parallel converter is Nbit word.
The digital to analog converter output remains unchanged till last N th bit received.
Then depending upon the input N bits, the output of D/A converter is defined.
This output of D/A converter remains unchanged till last bit is received.
The voltage m (t) is applied to modulator. This modulator modulates the phase of
sinusoidal carrier depending upon the amplitude of symbol m (t).
M-ary PSK Receiver
M-ary PSK signal given to coherent detectors. Each Coherent detector consists of
multiplier followed by a integrator. According to the phase of M-ary PSK signal
in each symbol duration, phase detector recovers the analog signal.
This analog signal is converted to digital signal by applying through A/D

converter. The analog to digital converter output is the N-bit symbol in parallel.
These bits are then converted to serial bit stream by parallel to serial converter.
Page 94
ECE Department
Power Spectral density of M-ary PSK

The power spectral density of M-ary PSK is given by S ( f ) 2 PT
s s[
sin( fTs ) 2
]
fTs
Where Ts = NTb.

=
2f
2
2
b
Ts NTb
N
Signal Space diagram or geometrical representation of M-ary PSK signals

Figure shows the signal space diagram of M-ary PSK. All signal space points lies on a
circle of radius
P sTb and any two successive signal space points phase differed by
2
.
M
Page 95
ECE Department
For M=8, the signal space diagram for M-ary PSK is shown below. In the figure, the
distance between signal point S1 and signal point S2 can be obtained by considering the
triangle followed by S1OA. The distance between S1 and S2 is denoted by d12.
Signal Space representation of 8-PSK scheme

Distance S1A = Distance S2A =
d min
2
By standard relation of right angle triangle sin(
Dis tan ceS1 A

Dis tan ceOS1
d min
sin( ) 2
M
PT
s s
i.e. d min 2 PT
s s sin(
) where Ts = NTb & M =2N
Let us verify the result for M =4, i.e 4-PSK scheme. d min 2 PT
s s sin(
d min 2 Ps 2Tb sin( )

4
= 2 PT
s b
Let us find the result for M =8, i.e 8-PSK scheme d min 2 PT
s s sin(
) 2 Ps 3Tb sin( )
M
8
Page 96
ECE Department
Comparison b/w BFSK and M-ary FSK

Parameter of Comparison
BFSK
M-ary FSK
No of symbols
2 (symbol 0 &1 )
M-Symbols
No of bits/symbol
1 bit
N bits
Symbol Duration
Ts =Tb
Ts=NTb
Output frequencies
fL & fH
f0, f1,f2,fM-1
Expression
S (t ) 2Ps cos(2 f mt )
S (t ) 2Ps cos(2 f mt )
For m=0,1
m=0, 1,..M-1
4fb if fH-fL=2fb
2 fs M 2
Bandwidth
1 N
1 N
2 2
2
Ts
NTb
M-ary FSK Transmitter
M-ary FSK transmitter
The N successive bits are presented in parallel to digital to analog converter.

These N bits form a symbol at the output of digital to analog converter. There
will be a total 2N =M possible symbols. The symbol is presented every T s =NTb
period.
Page 97
ECE Department
The output of digital to analog converter is given to a frequency modulator. Thus

depending upon the value of symbol, the frequency modulator generates the
output frequency.
For every symbol, the frequency modulator produces different frequency output.
This particular frequency signal remains at the output for one symbol duration.
Thus for M symbols, there are M frequency signals at the output of modulator.
Thus the transmitted frequencies are f0, f1, f2,.& fM-1.
M-ary FSK Receiver
The M-ary FSK signal is given to the set of M bandpass filters. The center
frequencies of those filters are f0, f1,f2, .fM-1.
These filters pass their
particular frequency and alternate others.
The envelope detector outputs are applied to a decision device. The decision
device produces its outputs are applied to a decision device. The decision device
produces its output depending upon the particular symbol; only one envelope
detector will have higher output. The outputs of other detectors will be very low.
The output of the decision device is given to N bit analog to digital converter.
The analog to digital converter output is the N bit symbol in parallel. These bits
are then converted to serial bit stream by parallel to serial converter.
M-ary FSK Receiver
Page 98
ECE Department
POWER SPECTRAL DENSITY OF M-ARY FSK
We know that for M symbol f0, f1,f2, .fM-1 frequencies are used for
transmission. The probability of error is minimized by selecting those frequencies such
that transmitted signals are mutually orthogonal. If those frequencies are selected as
successive even harmonics of symbol frequency fs, then transmitted signals will be
orthogonal.
Lets say that the lowest carrier frequency f0 is the kth harmonic of symbol frequency i.e.,
f0 =kfs, then the other frequencies will be, f1=(k+2)fs, f2=(k+4)fs .etc. Thus every
frequency is separated by 2fs from its nearest carriers. Figure shows the power spectral
density of M-ary FSK.
Band width required for the transmission of M-ary FSK is 2 f s M 2
1 N
1 N
2 2
2
Ts
NTb
2 N 1
fb
N
Signal Space Diagram of M-ary FSK

Here S0, S1,SM-1 are mutually orthogonal signals for M-symbols.
S0 PT
s b 0 (t )
S1 PT
s b 1 (t )
S 2 PT
s b 2 (t )
SM 1 PT
s b M 1 (t )
The orthogonal carriers
0 (t ), 1 (t ), 2 (t )..............., M 1 (t )
can be represented as
follows.
Page 99
ECE Department
0 (t )
2
cos(2 f 0t )
Ts
1 (t )
2
cos(2 f1t )
Ts
:
M 1 (t )
2
cos(2 f M 1t )
Ts
The figure below shows signal space diagram for M-ary FSK for M=3
In
the
signal
space
diagram, 0 (t ), 1 (t ), 2 (t )..............., M 1 (t ) form
mutually
perpendicular axes. For simplicity of understanding we will consider M=3.

Then the three carriers
0 (t ), 1 (t ), 2 (t )
will form three axes. Then the signals S0(t), S1 (t)
and S2(t) will be represented by vectors of length
PT
s s
along those axes.
Distance between signal points:

Distance between the signal space points is
dmin =
=
2
2
( PT
s s ) ( PT
s s)
2 PT
s s
This equation gives the minimum distance between any two signal points. This relation
holds for M signal points since all axes are perpendicular to each other.
Page 100
ECE Department
MINIMUM SHIFT KEYING

In QPSK, the phase changes by 900 (OQPSK). This creates abrupt
amplitude variations in the waveform. Because of the abrupt amplitude variations, side
lobes are having significant levels. These side lobes may interfere with the main lobe of
other adjacent channels causing inter channel interference. Hence band pass filtering is
required to avoid side lobes but these filter have other side effects for example say, it
alters the amplitude of the waveform.
MSK overcomes these problems. In MSK, the output waveform
is continuous in phase hence there are no abrupt changes in amplitude. These side lobes
of MSK are very small hence band pass filtering is not required to avoid inter channel
interference.
Figure shows the waveform of MSK. The binary bit sequence at the top.
Fig.a shows the corresponding NRZ waveform b(t). From b(t), two waveforms are
generated for odd and even bits. be (t) represent even bits (Fig.c) and bo (t) represent odd
bits (Fig.b).
The duration of each bit in bo (t) or be (t) is 2Tb, where as it is Tb in b(t).The waveforms
b0 (t) and be (t) have an offset of Tb. this offset is essential in MSK.
Two waveforms
The waveform of
cos(2
sin(2
fb
f
t)
cos(2 b t )
4 and
4 are generated as shown in fig.(d).
sin(2
fb
t)
4 passes through zero at the end of symbol time in be (t) and
fb
t)
4 passes through zero at the end of symbol in b0 (t).
be (t) is multiplied by
sin(2
fb
f
t)
cos(2 b t )
4 and bo (t) is multiplied by
4 .
Those product waveforms are shown in fig.(e) and (f).

The transmitted MSK signal is represented as,
s(t ) 2 Ps be (t )sin(2
fb
f
t ) cos(2 f ct ) 2 Ps bo (t ) cos(2 b t )sin(2 f ct )
4
4
Page 101

That is the product signals be (t )sin(2
ECE Department
fb
f
t ) and bo (t ) cos(2 b t ) modulate the
4
4
quadrature carriers of frequency fc.

After the rearrangements
S (t ) 2 Ps {
bo (t ) be (t )
f
b (t ) be (t )
f
}sin 2 ( f c b )t 2 Ps { o
}sin 2 ( f c b )t
2
4
2
4
S (t ) 2Ps CH (t )sin 2 f H t 2Ps CL (t )sin 2 f Lt

Where CH (t )
and f H f c
bo (t ) be (t )
b (t ) be (t )
& CL (t ) o
2
2
fb
f
& f L fc b
4
4
Case 1: when bo (t ) 1(1v) & be (t ) 1(1v) ; then CH(t) =1 &CL(t)=0

MSK signal is S (t ) 2Ps sin 2 f H t
Case 2: when bo (t ) 0(1v) & be (t ) 0(1v) ; then CH(t) =-1 &CL(t)=0
MSK signal is S (t ) 2Ps sin 2 f H t 2Ps sin(2 f H t 1800 )
Case 3: when bo (t ) 1(1v) & be (t ) 0(1v) ; then CL(t) =1 &CH(t)=0
MSK signal is S (t ) 2Ps sin 2 f Lt
Case 4: when bo (t ) 0(1v) & be (t ) 1(1v) ; then CL(t) =-1 &CH(t)=0
MSK signal is S (t ) 2Ps sin 2 f Lt 2 Ps sin(2 f Lt 1800 )
The frequencies fH & fL are chosen such that they are orthogonal over the interval T b.
For orthogonality following relation must be satisfied i.e.,
Page 102

Tb
sin(2 f
ECE Department
t )sin(2 f Lt )dt 0
This relation will be satisfied if we have integers m and n such that

2 ( f H f L )Tb n
and
2 ( f H f L )Tb m
Let us substitute values of fH and fL in the above relations.
2 ( f c
fb
f
f c b )Tb n
4
4
fbTb n
n 1
Similarly from equation above
2 ( f c
fb
f
f c b )Tb m
4
4
4fcTb =m
fc
m
fb
4
fb
Here fc must be integer multiple of 4 .
With n=1, this equation
2 ( f H f L )Tb n
becomes
fH fL
fb
2 .
Here n=1, means the difference between fH and fL is minimum and at the same time,
(MSK) they are orthogonal. Therefore this technique is called minimum shift keying
(MSK).
Substituting, this value of fc in equations f H f c
fH = fc +fb/4
=> mfb/4 + fb/4 = (m+1)fb/4
fL =fc fb/4
=> mfb/4- fb/4 =(m-1) fb/4
fb
f
& f L fc b
4
4
for these waveforms fH & fL are calculated with m=5.

We know that if bo(t)=be(t), then the transmitted waveform is of frequency fH.
And if bo(t)=-be(t), then the transmitted waveform is of frequency fL.
This shows that MSK is basically FSK with reduced bandwidth and continuous phase.
Page 103
ECE Department
MSK generation
Page 104
ECE Department
MSK TRANSMITTER
Block diagram of MSK transmitter
The two sinusoidal signals sin(2fct) and cos(2fb/4t) are mixed.
The bandpass filters then pass only sum and difference components fc+fb/4 and
fc-fb/4.
The outputs of bandpass filters are then added and subtracted such that two
signals x(t) and y(t) are generated.
Signal x(t) is multiplied by 2Ps b0(t) and y(t) is multiplied by 2Ps be(t).
The outputs of the multipliers are then added to give final MSK signal. Thus the
block diagram of above figure is the step to step implementation.
RECEPTION OF MSK SIGNAL
Figure shows the block diagram of MSK receiver. MSK uses synchronous detection.
The signals x(t) and y(t) are multiplied with the received MSK signal. Here x(t)
and y(t) have same values as shown in transmitter block diagram.
The outputs of the multipliers are b0(t) and be(t). The integrators integrate over
the period of 2Tb.
For the upper correlator, the sampling switch samples output of integrator at t =
(2k+1)Tb. Then the decision device recovers b0(t).
Page 105
ECE Department
MSK receiver block diagram
Similarly, lower correlator output is be (t). The outputs of two decision devices are
staggered by Tb
The switch S3 operates at t=kTb and simply multiplexes the two correlator
outputs.
Signal Space diagram of MSK signal
S (t ) 2Ps CH (t )sin 2 f H t 2Ps CL (t )sin 2 f Lt

S (t ) PT
s s CH (t )
2
2
sin 2 f H t PT
sin 2 f Lt
s s CL (t )
Ts
Ts
S (t ) PT
s s CH (t )H (t ) PT
s s CL (t )L (t )
The carriers H(t) and L(t) are in quadrature. Depending on the values of CH (t) and
CL (t), there will be four signal points in HL plane.
Page 106
ECE Department
Power Spectral Density and Bandwidth of MSK

The power spectral density of MSK signal is given by
s( f )
8Eb cos(2 ( f f c )Tb ) 2 8Eb cos(2 ( f f c )Tb ) 2

{
} 2 {
}
2 1 [4( f f c )Tb ]
1 [4( f f c )Tb ]
Power Spectral densities of MSK and QPSK
Bandwidth calculation of MSK

From the PSD spectrum of MSK, BW=0.75fb-(0.75fb) =1.5fb
BW of MSK is higher than that of QPSK.
Advantages & Drawbacks of MSK as compared to QPSK

Advantages
1. The MSK baseband waveforms are smoother compared to QPSK.
2. MSK signal have continuous phase in all the cases, whereas QPSK has abrupt
phase shift of /2
3. MSK waveform does not have amplitude variations, whereas QPSK signals have
abrupt amplitude variations.
4. The main lobe of MSK is wider than that of QPSK. Main lobe of MSK contains
around 99% of signal energy whereas QPSK main lobe contains around 90% of
signal energy.
5. Side lobes of MSK are smaller compared to that of QPSK. Hence inter channel
interference is absent in MSK
Page 107
ECE Department
6. To avoid inter channel interference due to side lobes, QPSK needs the band pass
filtering, whereas it is not required in MSK.
Drawbacks
1. The bandwidth requirement of MSK is 1.5fb, whereas it is fb in QPSK. Actually,
this cannot be said serious drawback of MSK. Because power to bandwidth ratio
of MSK is more. In fact 99% of signal power can be transmitted within the
bandwidth of 1.2fb in MSK. While QPSK needs around 8fb to transmit the same
power.
2. The generation and detection of MSK is slightly complex. Because of incorrect
synchronization, phase jitter can be present in MSK. This degrades the
performance of MSK.
Important Questions
1. Comparison of BASK, BPSK & BFSK modulation schemes
2. Explain the working of BPSK modulator & Demodulator with block
diagrams?
3. Explain the coherent & NON coherent detection process of BFSK signals?
4. Explain with neat block diagram the generation and recovery of DPSK
signals.
5. Explain the working offset QPSK transmitter and receiver with neat block
diagrams?
6. Draw the constellation diagrams for the modulation schemes BASK, BPSK,
BFSK, & QPSK.
7. Explain M-ary PSK signaling scheme. Draw the signal space representation
of M-ary PSK for M=8.
8. Explain M-ary FSK signaling scheme.
9. (a) Draw the signal space representation of MSK.
(b) Show that in a MSK signaling scheme, the carrier frequency in integral
multiple of fb/4 where fb is the bit rate.
(c) Bring out the comparisons between MSK and QPSK.
Page 108

Digital Communications 3units

Enviado por

Dados do documento

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Digital Communications 3units

Enviado por

Direitos autorais:

Formatos disponíveis

Sri Sai Aditya Institute of Science & Technology

JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY KAKINADA

Prepared by Venkata Satish N

Sri Sai Aditya Institute of Science & Technology

UNIT I-PULSE DIGITAL MODULATION

FIGURE 1. Basic elements of a digital communication system.

Sri Sai Aditya Institute of Science & Technology

Sri Sai Aditya Institute of Science & Technology

Prepared by Venkata Satish N

Sri Sai Aditya Institute of Science & Technology

ADVANTAGES OF DIGITAL COMMUNICATION OVER ANALOG MODULATION:

Prepared by Venkata Satish N

Sri Sai Aditya Institute of Science & Technology

Definition: Pulse code modulation (PCM) is essentially analog-to-digital conversion of a

Sri Sai Aditya Institute of Science & Technology

Analog signal is sampled every TS secs. Ts is referred to as the sampling interval.

Nyquist sampling rate for low-pass and bandpass signals

Prepared by Venkata Satish N

Sri Sai Aditya Institute of Science & Technology

Sampling results in a series of pulses of varying amplitude values ranging between

Prepared by Venkata Satish N

Sri Sai Aditya Institute of Science & Technology

i.e x(nTs ) xq (nTs )

Prepared by Venkata Satish N

Sri Sai Aditya Institute of Science & Technology

Most ADCs use uniform quantizers.

The quantization levels of a uniform quantizer are equally spaced apart.

Symmetrical type of mid rise quantizer

a) Symmetrical type of mid rise Quantizer b) Quantization error

Prepared by Venkata Satish N

Sri Sai Aditya Institute of Science & Technology

a) Symmetrical type of mid tread Quantizer b) Quantization error

Sri Sai Aditya Institute of Science & Technology

Fig: (c) Equispaced samples of m(t ) (d) Quantized sample sequence

Sri Sai Aditya Institute of Science & Technology

NOISE IN PCM SYSTEMS

There are two types of distortions associated with a quantizer:

Figure: Characteristic of quantization and overload errors.

Sri Sai Aditya Institute of Science & Technology

Sri Sai Aditya Institute of Science & Technology

Illustration of how quantization error is reduced by increasing quantization levels

Prepared by Venkata Satish N

Sri Sai Aditya Institute of Science & Technology

NON UNIFORM QUANTIZING

Non uniform Quantizer of 8 levels

Prepared by Venkata Satish N

Sri Sai Aditya Institute of Science & Technology

A-law and -law Companding

These two are standard companding methods.

-Law is used in North America and Japan

A-Law is used elsewhere to compress digital telephone signals

Prepared by Venkata Satish N

Sri Sai Aditya Institute of Science & Technology

Two types according to compression filter

A-law : used in Europe

A-law & -law compression curve

Similarities between Alaw and law

Prepared by Venkata Satish N

Sri Sai Aditya Institute of Science & Technology

Differences Between Alaw and law

Example: -law Companding

Prepared by Venkata Satish N

Sri Sai Aditya Institute of Science & Technology

A Gray code gives the best end-to-end performance.