Computer Engineering Department

Fourth Class
Digital Communications 4th IE & NE Computer Engineering

Digital Communications

Chapter (1) Introduction

1. Base band signal

The simplest digital data signal contains a sequence of signal element (units or pulses of a data
signal) where each element is binary coded. Having the choice of two possible shapes that
correspond to the element values 0 or 1, each signal element has the same duration of T seconds.
So that the signal element rate is 1/T elements per sec (or bauds). The digital signal above is
clearly a 2-level or binary signal. In a sequence of M-level signal elements, where and n
is the number of bits, the M-level data symbol that determines the element value of a signal
element can be represented by a sequence of n bits. For example, if bits then
levels and the corresponding sequence of 3 bits are 000, 001, 010… 111.

 2-level (binary) signal

, , 1

…bits per second

…symbol per second (baud)

 8-level signal

, 1 0 1

….bits per second

… symbol/sec

In general,

Exercise.1
For 2-level and 8-level systems, what is the difference between them (bit rate? noise?) for the
same .

1|P a ge
Digital Communications 4th IE & NE Computer Engineering

2. Model of Digital communication system

Message Source Channel Modulator

Tx Channel
encoder encoder

Codec Modem Noise

Rx User Source Channel Demodulator

decoder decoder

1. Information Source (Message): The source of information can be analog or digital, e.g.
analog: audio or video signal, digital: like teletype signal. In digital communication the
signal produced by this source is converted into digital signal which consists of 1′s and
0′s.
2. Source Encoder: In digital communication we convert the signal from source into digital
signal as mentioned above. The point to remember is we should like to use as few binary
digits as possible to represent the signal. In such a way this efficient representation of the
source output results in little or no redundancy. This sequence of binary digits is called
information sequence.
Source Encoding or Data Compression: the process of efficiently converting the
output of whether analog or digital source into a sequence of binary digits is known as
source encoding.
3. Channel Encoder: The information sequence is passed through the channel encoder. The
purpose of the channel encoder is to introduce, in controlled manner, some redundancy in
the binary information sequence that can be used at the receiver to overcome the effects
of noise and interference encountered in the transmission on the signal through the
channel.
4. Digital Modulator: The binary sequence is passed to digital modulator which in turns
converts the sequence into electric signals so that we can transmit them on channel (we
will see channel later). The digital modulator maps the binary sequences into signal wave
forms, for example if we represent 1 by sin x and 0 by cos x then we will transmit sin x
for 1 and cos x for 0. ( a case similar to BPSK).
5. Channel: The communication channel is the physical medium that is used for
transmitting signals from transmitter to receiver. In wireless system, this channel consists
of atmosphere , for traditional telephony, this channel is wired , there are optical
channels, under water acoustic channels etc.We further discriminate this channels on the
basis of their property and characteristics, like AWGN channel etc.

2|P a ge
Digital Communications 4th IE & NE Computer Engineering

6. Digital Demodulator: The digital demodulator processes the channel corrupted

transmitted waveform and reduces the waveform to the sequence of numbers that
represents estimates of the transmitted data symbols.
7. Channel Decoder: This sequence of numbers then passed through the channel decoder
which attempts to reconstruct the original information sequence from the knowledge of
the code used by the channel encoder and the redundancy contained in the received data.
8. Source Decoder: At the end, if an analog signal is desired then source decoder tries to
decode the sequence from the knowledge of the encoding algorithm. And which results in
the approximate replica of the input at the transmitter end.
9. Output Transducer: Finally we get the desired signal in desired format analog or
digital.
Exercise.2
Explain briefly the function of each component in the model
3. Measure of information
Consider M-level system with symbols .The information contents of a symbol ,
denoted by is defined by:

Where b is the radix of digits (2 for binary) and is the probability of symbol

The information measure (bits) of a message is equal to the minimum number of binary pulses
required to encode that message.
Example.1
How many bits per symbol to encode 32 different symbols?

Example.2
The four symbols occur with probability 1/2, 1/4, 1/8, 1/8 respectively. Find the
info content in the message ( )

Note:

3|P a ge
Digital Communications 4th IE & NE Computer Engineering

4. Bandwidth and Noise

There are two factors affecting the information transfer rate on a channel
 The bandwidth of the channel will determine how quickly the signaling states on the
channel can be changed.
 The level of noise in the channel will impose a limit on the number of different unique
states that can be correctly decoded at Rx. In addition, the degree of distortion introduced
by the channel will also limit the number and rate of change of symbol states.
So, if we had a channel with infinite BW (or no noise and distortion), it would be
possible to send, say, one Mbits at the speed of light.

5. Bandwidth efficiency
It is a measure of how well a particular format (and coding scheme) is making use of the
available BW. The units of BW efficiency is bits/second/Hz.

For example, if a system requires 4 KHz of BW to send 8000 bps of information, then

6. Multi-level signaling (M-ary)

In any communication system, there are two major resources that are used efficiently:
 Channel bandwidth
 Transmit power

01 2
3 1

1 0 10 00 4
0 0
0
5 7
11
6
2-level 4-level 8-level

4|P a ge
Digital Communications 4th IE & NE Computer Engineering

 BASK, BFSK, BPSK and DPSK are binary modulation techniques.

 They are represented using only two symbols consisting of single bit either “1” or
“0”.
 Two amplitudes, two frequencies or two phases are used to represent two symbols
respectively.

Binary levels = M=2

M-ary: levels = M=4

 M-level signal is used to modulate the carrier, and then this technique is known as
M-ary Digital Modulation Technique.
Example.3
A modem claims to operate with BW efficiency of 5 bits/sec/Hz when using 1024 symbol states.
a) How many bits are being encoded in each symbol, and what is the modem capacity if
the symbol rate is 4000 symbol/sec
b) How many symbol states should be employed if the user wishes to send his info in half
the time?
Solution
a)

5|P a ge
Digital Communications 4th IE & NE Computer Engineering

b) To send the info in half the time, it would be necessary to send data at bit rate = 80 Kbps
hence we need 20 bits in each symbol, so symbol states = 220 = 1048576

7. Channel capacity (for baseband signals)

The minimum BW required for error-free transmission

Since is the symbol rate ( ), then

The channel capacity is max but measured in bit/sec, so

8. Additive white Gaussian Noise channel (AWGN)

As M increases, the ability of the receiver to distinguish between symbols in the presence of
noise/interference/distortion decreases. Hence SNR (signal to noise ratio) will be an important
factor in determining how many symbol states can be utilized and still achieve (error-free)
communication.
Of each symbol is also key in determining the noise tolerance of a receiver system, with
longer symbols giving the receiver more time to average out the effects of noise than shorter
symbols.

The capacity of AWGN channel (Shannon capacity) defined as

Where B is the channel BW, SNR = , S is the signal power, N is the noise power =No B, and
No is PSD of the noise (watt/Hz)
Note

 The channel is error-free if R ≤ C

 For given C, the BW can be increased for decreased signal power

Example.4
Consider AWGN channel with B=4 KHz and noise PSD is 2x10 -12 W/Hz, the signal power
required at the modem receiver is 0.1 mW. Calculate the capacity of this channel.

6|P a ge
Digital Communications 4th IE & NE Computer Engineering

SNR= = = (1.25)104
C= B log2 (1+SNR) = 4000 log2 [1+1.25(104)] ≈ 54.44 Kb/s

Example.5
The specification of two telephone links are
Link B SNR
Class 1 300-3400 Hz 40 dB
Class 2 600-2800 Hz 30 dB

A company has a requirement to send data over a telephone link at bit rate R= 20 Kbps without
error. Would you advise the company to rent the more expensive class 1 service, or the cheaper
class 2 service? Justify your decision.

Solution:

For class one line:

B= 3400-300 = 3100 Hz, SNR = 40 dB=10000

For class two line:

B = 2800-600 = 2200 Hz, SNR= 30 dB=1000
C = 2200 log2 (1+1000) = 21.9 Kbps

So both of links will meet the specification of R=20 Kbps error-free. However, the performance
of class 2 line is very close to Shannon bound, in practice, it is unlikely that a modem could be
realized that would give the desired result on the class 2 line.

Exercise.3
A signal with 256 symbols is transmitted by 104 symbol per second.
a) What is the information rate R?
b) Can the output be transmitted without error over AWGN channel with B= 10 KHz and SNR=
100
c) Find the SNR required for error-free transmission for part (b)
d) Find the B required for AWGN channel for error-free transmission if SNR= 100

7|P a ge
Digital Communications 4th IE & NE Computer Engineering

9. Power and bandwidth efficiency

For a system transmission at maximum capacity, the average signal power (S) at receiver input
can be written as S=C Eb, where Eb is the average received energy per bit.
The average noise power (N) is N=No B, then Shannon expression can be written as:

( )

The ratio [ ] represents the bandwidth efficiency of the system (bits/sec/Hz).

The ratio [ ] is the power efficiency. The smaller the ratio, the less energy used by each bit.

Choosing a power-efficient modem is particularly important in cellular handsets (say to

maximize battery lifetime).

Example.6
A digital cellular telephone system is required to work at a BW efficiency of 4 bits/sec/Hz. What
is the min Eb/No that must be planned for in order to ensure that users on the edge of the
coverage area receive error-free communication? If the mobile telephone company wishes to
double the number of users, how much more power must the base-station and handsets radiate in
order to maintain coverage and error-free communication?
Solution:
log2 (1+ ),
4=log2 (1+4 )
= (24-1)/4 = 3.75 = 5.74 dB
In order to double the number of users for the same B, then =8
= (28-1)/8 = 31.87 = 15.03 dB
Thus, the transmitted power must be increase by a factor 15.03 - 5.74 = 9.29 dB

Exercise.4
Find the BW efficiency for a wireless communication system having a bit rate of 9.6 Kbps and B
of 200 KHz with of 10 dB.
Exercise.5
Data has to be transmitted which has B=3 KHz. If SNR at the receiver is 12 dB, determine for
data rates: 2.4 Kbps and 4.8 Kbps. Also, determine the BW efficiency.

8|P a ge
Digital Communications 4th IE & NE Computer Engineering

Exercise.6
Sketch roughly the relationship between and , then
a) Find the value of when
b) Mark the region on the graph that been considered error-free transmission
c) What is the minimum (in dB) for error-free transmission.

10. Inter Symbol Interference (ISI)

With any particle channel, the filtering effect will cause a spreading of data symbols through the
channel. For consecutive symbols, this spreading causes part of the symbol energy to overlap
with neighboring symbols, causing ISI. This degrade the ability of the detector to differentiate a
current symbol from diffused energy of adjacent symbols. Even with no noise present in the
channel, this can lead to defection errors.
It is possible to control ISI such that it does not degrade the system performance by pulse
shaping or using Nyquist filtering ( ), where is the sampling frequency.

9|P a ge
Digital Communications 4th IE & NE Computer Engineering

Bit Rate & Baud Rate

 Bit rate (R): it is number of bits / second.
 Baud Rate(r): it is number of symbols / second. Or [elements / sec.]
 If n = number of bit / symbol
⁄


Total number of symbols(M): M=2n

Example: An analog signal carries 4 bit/sec. elements. If 1000 signal elements are sent / sec.
Find the bit rate.
 n= 4bit/elements
 r= 1000 baud [element / sec. or symbols /sec.]
 R = n * r = 4 * 1000 = 4000 bits/sec. = 4 Kb/sec
 M =2n = 24 = 16.

Example: An analog signal has a bit rate 8000 bps and a baud rate of 1000 baud. How many data
elements are carried by each signal elements? How many signal elements do we need?
 R = 8000 bits / sec.
 r = 1000 baud
 n = R/r = 8000 / 1000 = 8 bits / elements
 M = 2n = 28 = 256

10 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Chapter (2) Waveform coding techniques

Pules Code Modulation (PCM)

 Pulse modulation is of two types

 Analog Pulse Modulation
 Pulse Amplitude Modulation (PAM): In
this scheme high frequency carrier (pulse) is
varied in accordance with sampled value of
message signal.
 Pulse width Modulation (PWM): In this
width of carrier pulses are varied in
accordance with sampled values of message
signal. Example: Speed control of DC Motors.
 Pulse Position Modulation (PPM): In this
scheme position of high frequency carrier pulse is changed in accordance with the sampled
values of message signal.

11 | P a g e
Digital Communications 4th IE & NE Computer Engineering

 Digital Pulse Modulation

 Pulse code Modulation (PCM)
 Delta Modulation (DM)

Pulse code Modulation (PCM) (or A/D conversion process)

 The message signal is the signal which is being transmitted for communication and the
carrier signal is a high frequency signal which has no data, but is used for long distance
transmission.
 A signal is pulse code modulated to convert its analog information into a binary
sequence, i.e., 1s and 0s. The output of a PCM will resemble a binary sequence.

 Instead of a pulse train, PCM produces a series of numbers or digits, and hence this
process is called as digital. Each one of these digits, though in binary code, represents the
approximate amplitude of the signal sample at that instant.
 In Pulse Code Modulation, the message signal is represented by a sequence of coded
pulses. This message signal is achieved by representing the signal in discrete form in both
time and amplitude.

12 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Basic Elements of PCM

 We do sampling to convert analog signal into discrete signal.

 After that, we do quantization to convert discrete signal into digital signal.
 At last, we do encoding of that digital signal.

Note: Before sampling the signal is filtered to limit bandwidth.

 Low Pass Filter

This filter eliminates the high frequency components present in the input analog signal which is
greater than the highest frequency of the message signal, to avoid aliasing of the message signal.

13 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Sampling Theory
 It is a process to convert time signal into discrete signal.
 Sampling of input signal x(t) can be obtained by multiplying x(t) with an impulse train
δ(t) of period Ts. The output of multiplier is a discrete signal called sampled signal which
is represented with y(t) in the following diagrams:

Sampled signal y(t)=x(t).δ(t)......(1)

δ(t)=a0+∑ ......(2)
 A sufficient number of samples must be taken so that the original signal is reconstructed.
 Number of samples to be taken depends on the maximum signal frequency present in the
signal.

statement of the sampling theorem

a) A bund limited signal of finite energy, which has a number of the frequency component
higher than fm (Hz), is completely described by its sample values at uniform intervals less
than or equal to 1/2fm.

14 | P a g e
Digital Communications 4th IE & NE Computer Engineering

TS ≤ 1/2fm
b) A bund limited signal of finite energy, which has a number of the frequency component
higher than fm (Hz), may be completely recovered from the knowledge of its samples
taken at the rate of 2fm samples per second.

 A continuous time signal can be completely represented in its samples and recovered
back if the sampling frequency is twice of highest frequency content of the signal.
fs ≥ 2W

 Analog signal is sampled after every T S interval.

 Sampling frequency is fS=1/TS. Is called sampling rate or sampling frequency.

15 | P a g e
Digital Communications 4th IE & NE Computer Engineering

 There are three sampling method:

 Ideal – An impulse at each sampling instant: an impulse at each instant.
 Multiplication process
 low energy
 High noise interference
 Practically not possible

The Fourier transform of the ideally sampled

signal given by above equation may be expressed
as,

16 | P a g e
Digital Communications 4th IE & NE Computer Engineering

 Natural sampling: a pulse of short width with varying amplitude.

 Chopping process
 Have energy
 Practically
 low noise interference
g(t) = x(t) when c(t) = A
g(t) = 0 when c(t) = 0
Where A is the amplitude of c(t).

17 | P a g e
Digital Communications 4th IE & NE Computer Engineering

 Flat top sampling: a pulse of short width with fixed amplitude. [PAM]

 Sample and hold circuit.

 Practically like natural
sampling but it easier to
natural sampling.
 It has very high noise
interference.

18 | P a g e
Digital Communications 4th IE & NE Computer Engineering

 Quantization
Each sample is rounded to the nearest on of set of levels (n = 3 bits means there are only 8 levels,
n = 4 bits means 16 levels and so on) M = 2n, where M is the number of quantized levels.

Ts = 1/fs
fs ≥ 2fm ….. Nyquist rate

19 | P a g e
 Digital Communications 4th IE & NE Computer Engineering

11
Δ
10

01

00

No. of level == quantizer {no. of bits (n) 2, 3, 4...}

In our example: n = 2, M = 2n =22 = 4

Sample value X 0 0.707 1 0.707 0 -0.707 -1 -0.707 0

Quantized value Xq 0 0.5 0.5 0.5 0 -0.5 -1 -0.5 0
Binary value 10 11 11 11 10 01 00 01 10
Eq= X - Xq
Quantization error 0 0.207 0.5 0.207 0 -0.207 0 -0.207 0

Quantization - Equations

, M = 2n

Δ = Step size of quantizer , M = total number of quantization levels

n = number of bits use or bit depth

XMAX

M
XMIN

20 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Xq

Xq = XMIN + IΔ
X = sampled input sample
I = index corresponding to
binary value
Xq = quantized sample
X

From the figure above:

Δ = 1-(-1)/22 = 0.5
M = 22 = 4
I = Round (0.707-(-1)/0.5 = Round (3.414) = 3
Xq = -1+3*0.5 = 0.5
Sample value X X
Index I
Quantized value Xq Xq

Bit Depth: number of bits used to represent the quantized samples

Bit depth = 2 or 3 or 4 …..
Bit depth =3 M=23 =8 Δ = 1-(-1)/23 = 0.25

21 | P a g e
Digital Communications 4th IE & NE Computer Engineering

 Bit rate and Bandwidth of PCM

The bit rate of PCM is and

 Signal to Quantization noise ratio (SNR) for PCM

Each additional coding bit, which doubles the number of quantize levels (M), halves Δ, decrease
the quantization error, and increases SNR.

Example.1

For voice message, calculate: quantization levels, sampling

frequency, bit rate, and the bandwidth of PCM.
Solution:
Quantization levels =
The message bandwidth
Sampling frequency =
Bit rate of PCM =

Bandwidth of PCM =

Example.2
An audio signal has bandwidth of 5.8 MHz encoded by PCM. Given that the total number of bits
to represent a level are 10 bits. Determine: total number of levels, bit rate, bandwidth, and SNR
in dB.
Solution:
Quantization levels =
Bit rate of PCM =

22 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Dynamic Range of Quantization:

It is a ratio of the largest (loudest) to smallest (quietest) measurable amplitude.

2Δ

Δ/2

23 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Types of Quantization

There are two types of Quantization - Uniform Quantization and Non-uniform

Quantization.
The type of quantization in which the quantization levels are uniformly spaced is
termed as a Uniform Quantization. The type of quantization in which the
quantization levels are unequal and mostly the relation between them is
logarithmic, is termed as a Non-uniform Quantization.
 Uniform Quantization
There are two types of uniform quantization. They are Mid-Rise type and
Mid-Tread type. The following figures represent the two types of uniform
quantization.

Uniform
X Y
Quantizer

 The Mid-Rise type is so called because the origin lies in the middle of a
raising part of the stair-case like graph. The quantization levels in this type
are even in number.
 The Mid-Tread type is so called because the origin lies in the middle of a
tread of the stair-case like graph. The quantization levels in this type are
odd in number.
 Both the mid-rise and mid-tread type of uniform quantizers are symmetric
about the origin.

24 | P a g e
Digital Communications 4th IE & NE Computer Engineering

25 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Example 1: a) A sinusoidal signal, with an amplitude of 3.25 volt, is applied to

uniform quantizer of the Mid-Tread type whose output takes on values
0,±1,±2, ±3 volts. Sketch the waveform of resulting quantizer output for one
complete cycle of the input. b) Repeat this evaluation for the case when the
quantizer is the Mid-Rise type whose output takes on the values 0.5, ±1.5,
±2.5, ±3.5 volts.

26 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Example 2: the signal m(t) = 6 sin (2πt) volts is transmitted using 4 bit binary
PCM system. The quantizer is of the Mid-Rise type, with 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=±1/8, ±3/8,±5/8,….., seconds.

at the sampling instants we have:

27 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Example.3

A 12 bit ADC, with analog input voltage ranging from -2 to 2 V. Determine the
following:
1. No. of quantization levels.
2. Step size.
3. Quantization level, when the analog voltage is 1.33V.
4. Quantization error.
5. Dynamic range.
6. SNR of quantization.
Example 4
A sinusoidal signal is transmitted using PCM scheme. The target output SNR
should be greater than 13 dB. Find the min. a number of representation levels and
the min number of bits required to represent each sample to achieve the above
performance.
Example 5
Consider a sinusoidal signal given by s(t) = 3 sin (1000π t) find :-
1. The output SNR when the signal is quantized using a 9 bit PCM.
2. The min. number of bits needed to achieve the output SNR of at least 40 dB
Example 6
A TV signal with a max. frequency 42 MHz is transmitted using binary PCM. The
number of quantization level is 1024. Calculate:
1. Code word length (n). 3. Average output SNR.
2. Transmission bandwidth. 4. Bit-rate.
Example 7
An input signal applied to PCM has a max. Frequency of 4 KHz and the input
range verses from -4.8 volt to +4.8 volt. The average power of the input signal is
30 Mw the target output SNR is 20 dB Assume uniform quantization and PCM
produces binary output.

28 | P a g e
Digital Communications 4th IE & NE Computer Engineering

1. Calculate the number of bits required represent each sample.

2. Identify the transmission bandwidth.
Exercise.3
The bandwidth of an input signal to the PCM is restricted to 4 kHz. The required
SNR is 20 dB.
(1). Calculate number of bits required per sample.
(2). Calculate total transmission bandwidth for 30 PCM coder that has been time
multiplexed.
Exercise.4
PCM system uses a uniform Quantizer followed by 7-bit encoder. The bit rate of
the system is 5 Mb/s. Calculate the maximum message bandwidth. [ Check
answer: 360 kHz]

1|P a ge
Digital Communications 4th IE & NE Computer Engineering

 Log PCM (Non-uniform Quantization)

One way to control the noise in PCM is by distribute the levels as log form. This
can overcome the range problem by making SNR less dependent on signal level. If
step size (Δ) is not constant for all input amplitudes but rather is proportional to
input magnitude, then weaker signals use a smaller Δ than more intense signals do.
The quantization levels should be logarithmically spaced using non-uniform
quantizer:

 Non-uniform quantization is generally used for speech and music signals.

 Crest factor =

𝑥 Non-Uniform 𝑦 𝑥
Quantizer

An effect of non-liner quantizing can be obtained by first passing the sample value
through a compressor at the sender, then through a uniform quantizer. This
technique increase amplitudes near zero. To compensate the effects happened at

2|P a ge
Digital Communications 4th IE & NE Computer Engineering

the sender, pass the sample values through an expander at the receiver. The process
of compression, uniform quantization and expansion is called Companding.
 it amplify low signal  it attenuate weak signal
 attenuated for strong signal  amplify strong signal

𝑐 𝑥 Uniform 𝑐 𝑥
𝑥 Compress Expand 𝑥
Quantizer

𝑐 𝑥 𝑥

𝑐 𝑥
𝑥 𝑐 𝑥

3|P a ge
Digital Communications 4th IE & NE Computer Engineering

 Quantization
The quantizer characteristic is linear for input values up to a certain threshold
(for )

Where A is the compression parameter.

And logarithmic beyond that

Where

 Quantization
The quantizer characteristic is defined with one smooth function as:

Where μ is the compression parameter.

For small input values ( ) the output is linear

For large values of x, the output y varies directly with the logarithm of |x|.
The values of specifies the relative input value near which the quantizer evolves
from a linear to a logarithmic characteristic.
Increasing makes the quantizer logarithmic over a wider range of input values
and makes SNR decreases slightly with large values of .

1|P a ge
Digital Communications 4th IE & NE Computer Engineering

Encoding
The output of quantizer is one of “q” possible signal levels. If we want to use a
binary transmission system, then we need to map each quantized sample into “v”
bit binary word.
Encoding is the process of representing each quantized sample by an v bit
code word. The mapping is one-to-one so there is no distortion introduced by
encoding. Some mapping is better than others.
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 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.
Unipolar RZ: A rectangular pulse (half symbol wide) is used for a 1 and no pulse
for 0.
Bipolar RZ: Positive and negative pulse are used alternatively for symbol 1 and
no pulse for symbol 0.
Manchester or Split phase code: Symbol 1 is represented by positive pulse
followed by negative pulse, with both pulses being of equal amplitude and half-
symbol wide; for symbol 0, the polarities of these pulses are reversed.

2|P a ge
Digital Communications 4th IE & NE Computer Engineering

3|P a ge
Digital Communications 4th IE & NE Computer Engineering

Differential Pulse Code Modulation (DPCM)

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
 DPCM is technique of
analog to digital signal
conversion (Difference of
two signals are sent instead
of actual value of the signal
at any sampling instant).
 This technique samples the analog signal and then quantizes the difference
between the sampled value and predicted value
 Then encodes the signal from a digital value.

4|P a ge
Digital Communications 4th IE & NE Computer Engineering

DPCM Receiver

Example
Consider the input samples X(n)={1.1, 1.2, 1.3, 1.6, 1.7}. Explain how encoding
and decoding is done in DPCM. Assume first order prediction filter X/(n)= Xq (n-
1)
X(nTS) X’(nTS)= Xq(nTS) e(nTS)= X(nTS)- X’(nTS) eq(nTS) Xq(nTS)= X’(nTS)+ eq(nTS)
1.1 0 1.1 1 0+1=1
1.2 1 0.2 0 1+0=1
1.3 1 0.3 0 1+0=1
1.6 1 0.6 1 1+1=2
1.7 2 0.3 0 0+2=2
Transmitted bits (1 0 0 1 0 ….)10 Transmitted bits (001 000 000 001 ….)2
Decoder:
eq(nTS) X’(nTS) Xq(nTS)= X’(nTS)+ eq(nTS)
1 0 1+0=1
0 1 0+1=1
0 1 1+0=1
1 1 1+1=2
0 2 2+0=2

Recovered samples 1 1 1 2 2 ….

5|P a ge
Digital Communications 4th IE & NE Computer Engineering

Example
Consider the input samples X(n)={2.1, 2.2, 2.3, 2.6, 2.7, 2.8}. Explain how
encoding and decoding is done in DPCM. Assume first order prediction filter
X/(n)= Xq (n-1)
X(nTS) X’(nTS)= Xq(nTS) e(nTS)= X(nTS)- X’(nTS) eq(nTS) Xq(nTS)= X’(nTS)+ eq(nTS)
2.1 0 2.1 2 0+2=2
2.2 2 0.2 0 2+0=2
2.3 2 0.3 0 2+0=2
2.6 2 0.6 1 2+1=3
2.7 3 -0.3 0 3+0=3
2.8 3 -0.2 0 3+0=3
Transmitted sequence {2 0 0 1 0 0}
DPCM ={ 010 000 000 001 000 000 }
Decoder:
eq(nTS) X’(nTS) Xq(nTS)= X’(nTS)+ eq(nTS)
2 0[initially] 2+0=2
0 2 0+2=2
0 2 0+2=2
1 2 1+2=3
0 3 0+3=3
0 0 0+3=3
Recovered signal = {2 2 2 3 3 3}

6|P a ge
Digital Communications 4th IE & NE Computer Engineering

Example
Using mid-riser quantizer, find DCPM output for the given input sequence {0, 0.3,
1.5, 0.7, 1 and 2.3}

3
2.5
2
1.5
1
0.5
0
-3 -2 -1 -0.5 0 1 2 3
-1
-1.5
-2
-2.5
-3

Encoder//

X(nTS) X’(nTS)= Xq(nTS) e(nTS)= X(nTS)- X’(nTS) eq(nTS) Xq(nTS)= X’(nTS)+ eq(nTS)
0 0[initially] 0 0.5 0.5+0=0.5
0.3 0.5 -0.2 -0.5 0
1.5 0 1.5 1.5 1.5
0.7 1.5 -0.8 -0.5 1
1 1 0 0.5 1.5
2.3 1.5 0.8 0.5 2
Transmitted sequence ={0.5, -0.5, 1.5, -0.5, 0.5, 0.5}
Decoder: eq(nTS) X’(nTS) Xq(nTS)= X’(nTS)+ eq(nTS)
0.5 0[initially] 0.5+0=0.5
-0.5 0.5 0
1.5 0 1.5
-0.5 1.5 1
0.5 1 1.5
0.5 1.5 2

Received sequence= {0.5, 0, 1.5, 1, 1.5, 2}

7|P a ge
Digital Communications 4th IE & NE Computer Engineering

Application of DPCM:
1. This technique mainly used for speech, image and audio signal compression.
2. In images, there is a correlation between the neighboring pixels, in video
signals; the correlation is between the same pixels in consecutive frams and
inside frames.
3. This method is suitable for Real-Time applications
Advantages of DPCM:
1. Bandwidth requirement of DCPM is less compared to PCM.
2. Quantization error is reduced because of prediction filter.
3. Number of bits used to represent one sample value is also redused compared
to PCM.
Signal-to-Quantize Noise Ratio (SQNR) of DPCM

δx2 δE2 δQ 2

δx2 = variance of original i/p

δE2 = variance of prediction error
δQ2 = variance of quantization error

Gp prediction gain produced by differential quantization = > 1 == δx2> δE2

(SNR)p prediction error to quantization noise ratio =

 Gp is maximized by minimizing δE2 of prediction error.

 Our objective should by to design prediction filter so as minimize δE2

8|P a ge
Digital Communications 4th IE & NE Computer Engineering

Delta Modulation:
It is special type of DPCM, where we use error function with only 2 levels in
quantization. M= 2n ==== n=1== 2=21 === 1 bit encoded in delta modulation
 When n B.W. where B.W = nfs === B.W = fs
 PCM all samples are encoding
 DPCM difference in samples are encoding (saving in BW)
 DM encode error signal
 X(t)> X\(t) ==Δ
 X(t)< X\(t) ==-Δ
[1 bit required]
e(nTs) =X(nTs) – X\(nTs)
eq(nTs) = +Δ
-Δ

9|P a ge
Digital Communications 4th IE & NE Computer Engineering

Quantizing Noise:
Delta modulation systems are subject to two types of quantizing error:
1. Slope overloads distortion.
2. Granular Noise.
Slope overloads distortion

This distortion arises because of large dynamic range of input signal. The rate of
rise of input signal x (t) is so high that staircase signal cannot approximate it. The
slope overload is said to occur when the step size “Δ” is too small to follow steep
segment of the input waveform 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.
To reduce this slope overload distortion, the slope of the quantizer must be greater
than maximum slope of the input signal.

10 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Granular Noise

Granularity, on the hand refers to situation where the stair case function x ^(nTS)
hunts around a relatively flat segment of the input function, with step size that too
large relative to local slope characteristic of the input. This means that very small
variations in the input signal, the staircase signal is changed by large amount Δ
because of large step size. The solution is to this problem is make step size small.
Bit Rate (Signaling Rate) of Delta Modulation
Delta Modulation Bit Rate (R) = Number of bit transmitted / seconds
= Number of samples/sec * Number of bits/samples
= fs * 1 = fs
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 a PCM codeword. Hence, we can
say that the channel bandwidth for delta modulation system is reduced to great
extent as compared to that for PCM system.
Example: Given sine wave of frequency f m and amplitude Am applied to a
delta modulation having step size Δ. Show that the slope overload will occur if

Here TS is the sampling period.

Solution: let us consider that the sine wave is represented as x(t) = Am sin (2πfmt)

11 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Maximum slope of delta modulator is given as .

We know that, the slope overload distortion will take place if slope the sine wave
is greater than slope of delta modulator i.e.

| |

| |

 To avoid slope overload distortion, the condition that must be satisfied is :

Quantization Error/ Noise in Delta Modulation

Quantization Error
If step size of the quantizer is Δ, the maximum quantization error εmax is Δ and the
range of quantization error is (-Δ, Δ).
The probability density function of this error is given by :

Mean Square Value of this quantization error (Noise Power) is given by:

12 | P a g e
Digital Communications 4th IE & NE Computer Engineering

E[ε2] = ∫ ∫

Signal to Quantization Noise Ratio in Delta Modulation:

Input signal is sinusoidal signal x(t) = Am sin wmt.

SNR =

( )
SNR = =

‫ ؞‬the output quantization power within the bandwidth f BWLPF is given by:

In the receiver, at the output of LPF of BW fBWLPF

( )
SNR = =

13 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Example 1: A DM system is designed to operate at sampling frequency of 6 KHz

and step size of 350 mV. Determine:
1. The Max. Amp. Of a 1 KHz input sinusoidal signal for which DM does not
show the slope overload.
2. Output SNR.
Example 2: Consider an analog i/p signal x(t) = 0.1 sin (2π 104 t) for DM system.
The signal is sampled at a rate to 2 * 104Hz. Find whether the slope overload
distortion occurs for the following step-size (4 mV and 60mV).
Example 3: A linear delta modulator is designed to operate on speech signals
limited to 3.4 kHz. The specification of modulator are follows; sampling rate = 10
fNyquist, where the Nyquist rate of speech signal, step size Δ = 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 * 3.4 k) = 68 kHz.
To avoid slope overload distortion:

Therefor

Example 4: Consider a DM system designed to accommodate analog message

signals limited BW=5 kHz. A sinusoidal test signal of amplitude A=1v and
frequency fm=1kHz is applied to the system. The sampling rate of the system is
50 kHz.
1. Calculate the step size required to minimize slop overload distortion.
2. Calculate the signal to quantization noise ratio of the system for
specified sinusoidal test signal.
Solution:

1. To avoid slope overload distortion: ==

14 | P a g e
Digital Communications 4th IE & NE Computer Engineering

2. SNR=

Example 5 Consider a low pass signal with BW= 3 kHz. A linear delta modulation
system with step size Δ = 0.1 v is used to process this signal at sampling rate ten
times the Nyquist rate.
1. For linear delta modulation, the maximum amplitude of sinusoidal test signal
of frequency 1 kHz which can be processed by system without slope-
overload distortion.
2. For the specifications given in part 1, evaluate the output signal to noise
ratio under (i) prefiltered and (ii) postfiltered conditions.
Solution:

1.

2. (i) SNR =

(ii) SNR=

Example 6 Find the Min. sampling freq. (fs)min to avoid slope overload when step-
size Δ=0.1 and x(t) = cos (2π 800 t).
Example 7 Let a message signal x(t) be the input to a DM when
x(t) = 6 sin (2π 103t) + 4 sin (4π 103t) volt. Determine the Min. sampling rate that
will prevent slope overload, if the step size is 0.314 volt.
Example 8 The input to a linear DM is sinusoidal signal whose freq. can vary from
200 to 4000 Hz. The input is sampled at 8 times the Nyquist rate. The peak amp.
Of the signal is 1 v. Determine:
1. The value of the step size in order to avoid slope overload when the input
signal freq. is 800 Hz.
2. What is the peak amp. of the input signal to just the overload the
modulator, when the input signal freq. is 200 Hz.
3. Is the modulator overloaded when the input signal freq. is 4 KHz.

15 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Example 9: A linear delta modulator has a step size 100mV and the minimum
output amplitude is +50 mV . A signal s(t) = 0.5 u(t) is applied to the input of delta
modulator. Show how the modulator tracks the input indicating the distortion in the
waveform. Sketch the waveform for 12 clock cycle, beginning at least 2 clock
cycles before t=0. Also, sketch the output waveform in NRZ format.
Solution:
a. Figure a shows the sketch of delta modulator input and the tracking
distortions. The input is a step signal of amplitude 0.5 volts beginning at t=0.
The input t 0 is 0 volt. Initial amplitude of DM predictor, at clock instant 1
is assumed +50mV .
b. The clock instants are shown in b. at the clock instant 2 the predictor output
is higher than the input (0V) and hence, a negative step(-100Mv ) is added
to predictor output. At clock instant 3 the predictor out is lower ( -50mV )
than the input (0.5V) and hence, a positive step is added to the predictor. At
clock instant 4 the predictor out is lower ( +50mV ) is still lower than the
input. Hence, a 100mv step is added. At clock instants 4,5,6,7 and 8 the
predictor out is lower than input and each instant a 100mv step is added to
the previous predictor out. At clock instant 9 the predictor output ( 550mV )
is found higher than the input. Hence, a 100mv step is subtracted from the
predictor output. At clock instant 10 a 100mv step is added. The DM output
waveform is shown in c.

16 | P a g e
Digital Communications 4th IE & NE Computer Engineering

17 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Example 10: A segment of a delta modulated data stream is sequence given below
(0 1 0 1 0 1 1 1 1 1 1 0 0 0 1 1 0 0 0 1) this sequence is applied liner modulator
having a step size of 100mv. Assuming initial output, of the modulator is 0v, show
the output sample voltages at each bit and sketch the waveform.
Solution:

Example:
0 2 3 7 8 ……… 153 156 160 154 153

18 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Adaptive Delta Modulation (ADM)

In Adaptive Delta Modulation, the step
size of the staircase signal is not fixed
and changes depending upon the input
signal. Here first the difference between
the present sample value and previous
approximation is calculated. This error is
quantized i.e. if the present sample is
smaller than the previous approximation,
quantized value is high or else it is low.
The output of the one-bit quantizer is
given to the Logic step size control
circuit where the step size is decided.

At the logic step size control circuit, the

output is decided based on the quantizer
output. If the quantizer output is high,
then the step size is doubled for the next
sample. If the quantizer output is low, the
step size is reduced by one step for the
next sample.

Advantages

Some of the advantages of this modulation method are:

 Adaptive delta modulation decreases slope error present in delta modulation.

 During demodulation, it uses a low pass filter which removes the quantized noise.
 The slope overload error and granular error present in delta modulation are solved using this
modulation. Because of this, the signal to noise ratio of this modulation is better than delta
modulation.
 In the presence of bit errors, this modulation provides robust performance. This reduces the
need for error detection and correction circuits in radio design.
 The dynamic range of Adaptive delta modulation is large as the variable step size covers large
range of values.

19 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Differences between Delta Modulation and Adaptive Delta Modulation

The differences between adaptive delta modulation and delta modulation are:

 In Delta Modulation step size is fixed for the whole signal. Whereas in Adaptive delta
modulation, the step size varies depending upon the input signal.
 The slope overload and granular noise errors which are present in delta modulation are not
seen in this modulation.
 The dynamic range of Adaptive delta modulation is wider than delta modulation.
 This modulation utilizes bandwidth more effectively than delta modulation.

Applications

Some of the applications of this modulation method are:

 This modulation is used for a system which requires improved wireless voice quality as well
as speed transfer of bits.
 In television signal transmission this modulation process is used.
 This modulation method is used in voice coding.
 This modulation is also used as a standard by NASA for all communications between mission
control and spacecraft.
 Motorola’s SECURENET line of digital radio products uses 12kbits/sec Adaptive Delta
Modulation.
 To provide voice detection quality audio at deployed areas, military uses 16 to 32 kbit/sec
modulation system in TRI-TAC digital telephones.
 US army forces use 16kbit/sec rates to conserve bandwidth over tactical links.
 For improved voice quality US Air Forces uses 32kbits/sec rates.
 In Bluetooth-services to encode voice signals, this modulation is used with 32bits/sec rates.
 HC55516 decoder is used in various arcade games such as sinistar and smash tv and pinball
machines such as gorgor or space shuttle, to play pre-recorded sounds.
 Adaptive delta modulation is also known as continuously variable slope delta modulation.

This modulation encodes at 1-bit per sample. Here the encoder maintains a reference sample and
a step size. Before deciding the step size of the input signal it is compared with the reference
sample. This modulation method compromise between simplicity, low bitrate, and quality.

20 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Chapter (3) Digital Modulation Techniques

 The carrier is commonly written as .

 The choice of modulation method affects the ease of implementation; the noise tolerance
and occupied channel bandwidth of the resulting band pass data modem.

21 | P a g e
Digital Communications 4th IE & NE Computer Engineering

1. Amplitude Shift keying (ASK)

The simplest form of band-pass data modulation is ASK. Here, the symbols are represented as
various discrete amplitudes of a fixed carrier frequency .

 BASK
In binary ASK (2ASK), where only two symbol states are needed, the carrier is simply turned on
or off, and this process is called ON-OFF keying (OOK)

The spectrum of ASK signal can easily be determined if the spectrum of the baseband data
symbol is known, by viewing ASK modulation process as a mixing or multiplication of the
baseband symbol with carrier

BASK Modulator

Where A is a constant, is the carrier frequency, and T is the bit duration. It has
a power , so that √

22 | P a g e
Digital Communications 4th IE & NE Computer Engineering

√ √ √ √ √

Where is the energy contained in a bit duration.

 M-ary ASK

If more than two levels are used, then an M-ary ASK is adopted for high bit rate (4ASK for
2bits, 8ASK for 3bits and so on). 4ASK is shown here:

An M-ary amplitude-shift keying (M-ASK) signal can be defined by

Where

√ √ √ √ √

MASK Modulator

23 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Here is a 4ASK signal constellation diagram:

(MASK coherent demodulator)

 Bandwidth efficiency and Capacity of ASK

BW α r

‫ ؞‬BW = (1+d)*r = (1+d)*

 r == baud rate.
 R == bit (data) rate.
 n == number of bits/sample.
 d == factor for modulation and filtering ( 0 ≤ d ≤ 1 )
 0 is ideal modulation. == BW = r.
 1 is worse modulation.== BW = 2r.

24 | P a g e
 Digital Communications 4th IE & NE Computer Engineering

Demodulation of ASK

Synchronous (Coherent) Non-Synchronous

(Non-Coherent)
m(t) cos2 wct
m(t) cos wct m(t) cos wct m(t)
LPF m(t) Envelope detector

 Adv.: cost is low

 Disadv.: performance is poor with
lower SNR received signal
 Adv.: it is efficient
 Disadv.: it is costly

 Applications:
 Broadcasting in signal.
 In optical fiber communications for laser intensity modulation.
Example.1
We have an available BW of 100 KHz which spans from 200 to 300 KHz. What are the
carrier frequency and bit rate, if we modulated our data by using BASK with d = 1?
Example.2
ASK is used for transmitted data at Kbps over a telephone channel with
bandwidth
a. How many symbol states are required in order to achieve this level of performance?
b. What would be the equivalent number of symbol states needed if the channel pass
band extended from 0 Hz to 3100 Hz and baseband M-ary was used?
c. What is the maximum capacity for the ASK if the SNR on the telephone link is 33
dB.
Solution:
a. The capacity of band pass ASK is

25 | P a g e
Digital Communications 4th IE & NE Computer Engineering

b. The capacity of baseband M-ary system is

c. Shannon capacity is

 Probability of Symbol error

There are two types of waveform used in digital communications: unipolar and bipolar
waveforms

1 1 0 1 1 1 0 1
A A

0 0

-A
Unipolar Bipolar

For unipolar waveform, the energy per symbol is different depending on whether a logic 0 or
1 is sent, having a zero value for logic 0 case. The Probability of symbol error for unipolar is:

[√ ]

Where is noise power density, and is the complementary error function:

∫
√

And the Probability of symbol error for bipolar waveform is

26 | P a g e
Digital Communications 4th IE & NE Computer Engineering

[√ ]

It is often to draw which is called SNR rather than where and n is the number of

bits per symbol.

When the number of levels is increased (M 2), the ability of the receiver to distinguish between
symbols in the presence of noise will decrease. The Ps for M-ary bipolar baseband signaling is:

[√ √ ]

Example.3

A company wishes to increase the through put of a telephone modem product by changing
from 2-level signaling to 8-level signaling and has set a design target of maintaining a
performance of no worse than one symbol error in every 10 000 symbols sent. By using the
plot of symbol error vs. Eb/No for M-ary, determine the reduction in noise tolerance for the
modem because of this change. What is the theoretical minimum Eb/No required
supporting the bandwidth efficiency achievable by the 8-level modem?

Solution:

From the plot of for M-ary signaling, at =10-4, it can be seen that an increase of about 8 dB
is required to maintain the same error rate. Therefore, the new modem will be approximately 8
dB less tolerant to noise.

For 8-level modem, maximum bandwidth efficiency is 6 bits/sec/Hz, so

[ ]

[ ]

Therefore the minimum for error-free transmission is

= = 10.5 or 10.2 dB

27 | P a g e
Digital Communications 4th IE & NE Computer Engineering

 Bit-error rate (BER) performances of ASK

The performance of digital communication systems is presented at the simplest level as a

probability of bit error or probability of symbol error , as a function of the received E b/No
ratio. Binary ASK effectively uses a unipolar baseband modulation source.

2. Frequency Shift keying (FSK)

FSK has until recent years been the most widely used form of digital modulation, being simple
both to generate and to detect, and being insensitive to amplitude fluctuations in the channel.
FSK conveys the data using distinct carrier frequencies to represent symbol states.  An
important property of FSK is that the amplitude of the modulated wave is constant.

Consider the case of unfiltered 2FSK. This waveform can be viewed as two separate ASK
symbol streams summed prior to transmission.

28 | P a g e
Digital Communications 4th IE & NE Computer Engineering

 BFSK generation
FSK can be generated by switching between distinct frequency sources; however, it is likely that
there will be discrete phase jumps between the symbol states at the switching time. Any phase
discontinuity at the symbol boundary will result in much greater prominence of high frequency
terms in the spectrum, implying a wider bandwidth for transmission.

Alternatively, FSK can be realized by applying the data signal as a control voltage to a voltage-
controlled oscillator (VCO). Here the phase transition between consecutive symbols states is
guaranteed to be smooth (continuous). FSK with no phase discontinuity between symbols is
known as a continuous phase (CPFSK).

 The vector modulator

29 | P a g e
Digital Communications 4th IE & NE Computer Engineering

The arrangement of mixer and a combiner forms an extremely useful building block in digital
communication systems. It achieves a linear frequency translation of all components in the input
signal (represented by its in-phase and quadrature components) by a carrier frequency
component (also represented by its in-phase and quadrature components). This block is often
referred to as a vector modulator or quadrature modulator, and can be used for both frequency
up-conversion and down-conversion. The output of the two mixing processes is given by

When the above terms are summed, the result gives a down-converted component:

In addition, when subtracted from each other result in a signal up-converted component:

Exercise.1
A vector modulator is fed with a perfect quadrature sine wave at the input, but there is a
small phase error of 5o between the notional quadrature inputs of the carrier signal. What
will be the ratio in dB between the sum and difference outputs of the vector modulator
(ratio of the amplitude of the wanted to unwanted output signal)? [Hint: Sin for
small ], [Check answer: 27 dB]

Let us write the input to the vector modulator as:

cos(w0t), and sin(w0t)
and the carrier inputs as:
cos(wct), and sin(wct + )
Where is the phase error. Now:

30 | P a g e
Digital Communications 4th IE & NE Computer Engineering

sin(wct + ) = sinwct cos – cos wct sin

and for small phase errors this can be approximated to:
sin(wct + ) = sinwct.cos
The mixer outputs then become:
cos(w0t).cos(wct) = 0.5cos(wc + w0)t + 0.5cos(wc –w0)t
and sin(w0t).sin(wct + ) = – 0.5cos(wc + w0)t.cos + 0.5cos(wc –w0)t.cos
At the output of the summing device we get a wanted term at the difference frequency
and an unwanted term (usually referred to as the image) at the sum frequency as follows:
Difference term:
0.5{1 + cos }cos(wc –w0)t
Sum term:
0.5{1 –cos }cos(wc + w0)t
The ratio of the amplitude of the wanted to unwanted terms is thus:
Amplitude ratio (image supression) = {1 + cos } / {1 –cos }
For a phase error of 5o, the amplitude ratio of wanted to unwanted signals is thus 525:1,
or a relative power level of approximately 27 dB

 BFSK Vector modulator

BFSK requires the generation of two symbols, one at a frequency and one at a
frequency . So to generate a shift of , I and Q inputs need to be fed with
and respectively. Generating a shift of requires inputs of
and .

This approach is now frequency used to generate filtered CPFSK particularly in cellular
handsets.

31 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Exercise.2
Draw the block diagram of vector modulator to generate 2ASK signal.

 Spectrum of BFSK
Approximations of BFSK spectrum can be obtained by plotting the spectra for two ASK streams
centered on the respective carrier frequencies.

Clearly, the overall bandwidth occupied by FSK signal depends on the separation between the
frequencies representing the symbol states. CPFSK system will have much lower side-lobe
energy than the discontinuous case.

 Coherent BFSK detection

This method is very similar to that for ASK but in this case there are two detectors tuned to the
two carrier frequencies.

32 | P a g e
Digital Communications 4th IE & NE Computer Engineering

 M-ary FSK system

M-ary FSK (multi-level) is very much of interest for increasing the noise immunity of the
modulation format compared with BFSK, allowing a designer to achieve reliable data
transmission in the presence of high levels of noise. This is only possible by using a set of
“orthogonal symbols”. Two symbol states are said to be orthogonal over the
symbol period if:

∫ →

If the frequencies of M-FSK symbols are chosen to be of the form:

[ ]

Then these frequencies are orthogonal over a symbol period.

Example.3

For 8-FSK and Rs=1200, the required frequencies are 1000,1600,2200,2800,3400,4000,4600 and
5200 Hz.

 Orthogonal system gives better SNR at detector output, improving the probability of correct
symbol detection but required high bandwidth.

33 | P a g e
Digital Communications 4th IE & NE Computer Engineering

 M-ary FSK detection

A typical M-ary FSK detector consists of a bank of “correlators” (mixers with coherent carrier
reference), followed by a decision circuit at the output determining which correlator has the
largest output and hence which symbol was sent.

 Advantage of FSK
 FSK is constant envelope modulation and hence insensitive to amplitude variations in
the channel.
 The detection of FSK is based on relative frequency changes between symbol states
and thus does not required absolute frequency accuracy in the channel.
 In deep space missions where the path loss is so great, M-ary FSK is very effective
modulation.
 Disadvantage of FSK
 FSK is less bandwidth efficient than ASK or PSK
 The bit/symbol error rate performance of FSK is worse than for PSK.

34 | P a g e
Digital Communications 4th IE & NE Computer Engineering

3. Phase shift keying (PSK)

With PSK, the information is contained in the instantaneous phase of the modulated carrier.
Usually this phase is imposed and measured with respect to fixed carrier of known phase-
coherent PSK. For binary PSK (2PSK), phase states of 0 o and 180o are used. It is also possible to
transmit data encoded as the phase change (phase difference) between consecutive symbols
(Differentially coherent PSK). There is no non-coherent detection for PSK.

For BPSK:

Where Es is energy per symbol, T is symbol time and √ is the amplitude (A) of the signal.

In general, for MPSK:

√ ( )

Where = is the modulation angle.

The constellation mapping for MPSK can be shown below

35 | P a g e
Digital Communications 4th IE & NE Computer Engineering

BPSK QPSK QPSK 8PSK

 QPSK is quadrature PSK (4PSK).

 BPSK spectrum

The bandwidth of BPSK signal is identical to that of BASK. In fact, BPSK can be viewed as
ASK signal with the carrier amplitudes as + A and –A (rather than +A and 0 for ASK).

 PSK generation

The simplest means of realizing unfiltered BPSK is to switch the sign of the carrier using the
data signal, causing 0o or 180o phase shift.

36 | P a g e
Digital Communications 4th IE & NE Computer Engineering

 The square pulses for data signal are not practical to send. They are hard to create and
required a lot of bandwidth. The solution here is to send shaped pulses that convey the same
information but use smaller bandwidth and have other good properties such as ISI rejection.

There are some common pulse shaping methods that control the shape and the bandwidth of the
signal:
 Root raised cosine (used with QPSK)
 Half sinusoid (used with MSK (minimum shift keying))
 Gaussian (used with GMSK. This system is used in several mobile systems around the
world such as in GSM (global special mobile)
 Quadrature partial response (QPR)

 Detection of BPSK
There is no non-coherent detection for PSK, and various forms of coherent detection must be
employed. The ideal detector thus requires perfect knowledge of the unmodulated carrier phase
at the receiver (carrier recovery). As with ASK, any phase error of the locally generated carrier
reference reduces the signal level at the output of the detector by cos . This in turn degrades the
Es/No performance.

Thus, we need zero phase error for optimum detection. Note that if the phase error reaches 90o,
the output falls to zero.

37 | P a g e
Digital Communications 4th IE & NE Computer Engineering

 Quadrature phase-shift keying (QPSK)

QPSK uses the orthogonality between cosine and sine carrier. This would imply that if we send
BPSK on the cosine of a carrier, and simultaneously send a second BPSK using the sine of a
carrier, then it would be possible to detect each one independently of the other.  Orthogonality
property of QPSK means that it can be used to send information at twice the speed of BPSK in
the same bandwidth. The block diagram of QPSK modulator is simply two BPSK using
quadrature carriers summed in parallel. The source data is first split into two data streams, with
each data stream running at half the rate of input data.

 Performance of MPSK

Increasing M allows further improvements in bandwidth efficiency but requires more for

same Ps.

38 | P a g e
Digital Communications 4th IE & NE Computer Engineering

4. Quadrature Amplitude Modulation (QAM)

So far, we have considered only signal property modulators using amplitude, frequency, or
phase symbols to conveying the data. We can combine two or more symbols types, which gives
improved performance (trade-off between bandwidth efficiency and noise performance).

16-QAM constellation
The simplest form of QAM is in fact the QPSK symbol set, which Can be viewed as two
quadrature amplitude modulated carriers, with amplitude levels of +A and -A .
Increasing the number of amplitude levels on each carrier to 4 (for example ±A, ±3A) gives 16
possible combinations of symbols at the output, each equally spaced on the constellation
diagram, and each represented by a unique amplitude and phase.

 QAM generation
The modulator is making use of orthogonality of the sine and cosine carriers to allow
independent detection of the two ASK data.

39 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Pulse shaping is performed by filtering the multi-level baseband input symbol streams as in
ASK.

 QAM detection

QAM can be decoded using coherent detection just as for PSK (requires carrier recovery). The
output of each demodulator is a baseband multi-level symbol set; this should undergo matched
filtering for optimum performance in noise. The aim of comparator is to determine the level at
the sampling instant, and hence decode the corresponding bit pattern.

 M-ary QAM vs. M-ary PSK

Comparing the constellation diagrams of 16 QAM with 16 PSK, we can see that the spacing
between symbol states for QAM is greater than that for PSK, which means that the detection
process in QAM should be less susceptible to noise. However, the power for QAM is greater

40 | P a g e
Digital Communications 4th IE & NE Computer Engineering

than that for PSK and this must be taken into account if the transmission process is power
limited.

Example.4
A digital TV has a source analogue video signal with BW from 0 Hz to 2 MHz. This signal is
sampled at four times the highest frequency using 16-bit ADC. The resulting data signal is sent
over the air using 16QAM modulation. Assume ideal pulse-shaping filter, what is the bandwidth
occupied by the transmitted digital video signal?
Solution:
Sampling rate at ADC =
Bit rate at ADC output = 16 bits (8M) = 128 Mbps

16 QAM uses , so = 4 bit/sec/Hz

Hence

41 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Example.5
A transmitter for digital radio system is peak power limited to 150 W with 50 antenna.
Determine the average power that can be supported for both 16 PSK and 16 QAM transmission
if each point in the constellation has an equal probability of transmission.
Solution:

With reference to one quadrant of the 16 QAM constellation, the average power developed by
each of the vectors A, B, C, D is as follows:
A2 = (3a)2+(3a)2=18 a2 , B2 = (3a)2 + (a)2 = 10 a2
C2 = (a)2 + (a)2 = 2 a2 , D2 = (3a)2 + (a)2 =10 a2

Average power = =

The maximum vector power is 150 w, so =150 = = a =√ = 20.4

The average power for all symbol states is: Pav (QAM) = = 83.33 w

Pav for 16 PSK is the same for all symbol states Pav (PSK) = = 150 w

Exercise.3
If the maximum vector length in 16 QAM is 100 v rms, determine the average power that would
be delivered into R=50 antenna load if each point in the constellation has an equal probability
of transmission. [ Check answer: Pav = 111 w]

Exercise.4

If the peak symbol power for 16QAM is 200 w, measured in R=50 antenna load. What are the
amplitudes of the different symbol vectors in the transmitted waveforms?

42 | P a g e
Digital Communications 4th IE & NE Computer Engineering

Exercise.5

Orthogonal 4FSK modem has . If the lowest symbol frequency is 8

kHz, what will be the other three symbol frequencies?

Exercise.6

64QAM data link operates at 256 kbps. What is the symbol rate on the channel, and what is the
occupied bandwidth?

Exercise.7

What is the minimum bandwidth required to support 256 kbps data stream using BPSK, QPAK,
and 64QAM?

Exercise.8

A customer requires a microwave radio link to provide a bit rate of 2 Mbps in a bandwidth of
400 kHz. The minimum SNR on the channel is 30 dB. Can the channel support the required
capacity? Moreover, how many symbol states would be required?

43 | P a g e