Digital

Communication
Unit 3 -Digital Modulation-II

7/23/2021
 Conten
1. Generation, ts Space Representation
Reception, Signal and
Probability of Error Calculation for

• Quadrature Amplitude Shift Keying (QASK)

• M-ary FSK (MFSK)
• Minimum Shift Keying (MSK)
2. Pulse Shaping to reduce Interchannel and Intersymbol Interference
3. some Issues in transmission and reception
4. Orthogonal Frequency Division Multiplexing (OFDM),
Comparison of

digital modulation systems

7/23/2021
Course Objective

• To familiarize students with variousdigital modulation

techniques used in digital communication systems.
• To equip students the students with tools required for
performance analysis of digital communication systems.

7/23/2021
Course
Outcome
CO2: Understand and explain various digital modulation techniques used
in digital communication systems and analyze their performance in
presence of AWGN noise.

7/23/2021
 Quadrature Amplitude Shift Keying (QASK)/
Quadrature Amplitude Modulation (QAM)
q It is possible to combine ASK and PSK.

q Advantage is to increase the number of symbols available.

q It is rare for all three methods to be combined

q Noise immunity can be improved

q ASK and PSK to be combined to create QASK / Quadrature
amplitude modulation (QAM)

7/23/2021
 Quadrature Amplitude
q Modulation(QAM)
Uses “two-dimensional” signalling
q Original information stream is split into two sequences
that consist of odd and even symbols, e.g. B k and A k
1 0 1 1 0 1 …
1 -1 1 1 -1 1 …
B1 A1 B2 A2 B3 A3 …

∙ Ak sequence (in-phase comp.) is modulated by cos(2πfc t)

and Bk sequence (quadrature-phase comp.) is modulated by
sin(2πfc t)
∙ Composite signalAkcos(2πfc t) + B k sin(2πfc t ) is sent through
the channel A k x Y (t) = A cos(2πf t)i k c

cos(2πfct) + Y(t) = A k cos(2πfct) + B k sin(2πfct)

x Transmitted Signal
Bk Y q (t) = B k sin(2πfct)

sin(2πfct)

7/23/2021
 QAM
QAM can also be seen as a combination of ASK &
PSK

Bk
(-A,A) (A, A)

4-level QAM Ak

(-A,-A) (A,-A)

7/23/2021
 QAM…

–
16- level QAM

7/23/2021
 QAM
Transmitter

7/23/2021
 QAM
Receiver

7/23/2021
7/23/2021
 Probability of Error of M-ary
QAM

7/23/2021
 Fig: QAM for M = 16

7/23/2021
7/23/2021
 Multi-level (M-ary) Phase and Amplitude
Modulation

16 16 PSK 16 APSK
QAM
q Amplitude and phase shift keying can be combined to transmit several bits per symbol.
q Often referred to as linear as they require linear amplification.
q More bandwidth-efficient, but more susceptible to noise.
q For M=4, 16QAM has the largest distance between points, but requires very linear
amplification. 16PSK has less stringent linearity requirements, but has less spacing
between constellation points, and is therefore more affected by noise.

7/23/2021
 M-ary Quadrature
Amplitude Modulation
q It’s(QAM)
a Hybrid modulation

q As we allow the amplitude to also vary with the

phase, a new modulation scheme called
quadrature amplitude modulation (QAM) is
obtained.

q The constellation diagram of 16-ary QAM consists of

a square lattice of signal points.

7/23/2021
 The general form of an M-ary QAM signal can
be defined as

where
Emin is the energy of the signal with the lowest
amplitude and ai and bi are a pair of independent
integers chosen according to the location of the
particular signal point.

In M-ary QAM energy per symbol and also distance

between possible symbol states is not a constant.

7/23/2021
q It reasons that particular values of S i (t) will be
detected with higher probability than others.

q The signal S i (t) may be expanded in terms of a pair of

basis functions defined as

7/23/2021
 Power Efficiency and Bandwidth :
q Power efficiency of QAM is superior to M-
ary PSK.

q Bandwidth efficiency of QAM is identical

to M-ary PSK.

7/23/2021
 Fig: signal constellation of M-ary QPSK and M-ary QAM(M=16)

7/23/2021
Bandwidth
f b ⎛ Eb f b ⎞
Efficiency W
= log2 ⎜ 1+
⎝ ⎠ η ⎟

fb = capacity W(bits per second)

W = bandwidth of the modulating baseband
signal (Hz)
Eb = energy per bit
η = noise power density (watts/Hz) Thus
Eb fb = total signal power
ηW = total noise power
fb
= bandwidth use
W efficiency

= bits per second per Hz

7/23/2021
M-ary Frequency Shift
Keying(MFSK)
q In M-ary FSK modulation the transmitted signals are defined
by:

where fc = nc/2Ts, for some fixed integer n.

The M transmitted signals are of equal energy and equal duration, and
the signal frequencies are separated by 1/2Ts Hertz, making the signals
orthogonal to one another.

7/23/2021
 M ary FSK Transmitter -
Receiver

7/23/2021
 M ary FSK PSD

Carrier frequencies should be mutually orthogonal. Even harmonics

of symbol frequencies can be selected.

7/23/2021
 Signal Space representation of
MFSK

7/23/2021
 Signal Space representation of 8 FSK

7/23/2021
 Bandwidth of M-ary FSK

Considerably large bandwidth as compared to M ary PSK

7/23/2021
 • M-ary FSK requires considerably increased bandwidth in
comparison
with M-ary PSK
• The probability of error for M-ary FSK decreases as M increases,

while for M-ary PSK The probability of error increases with M

7/23/2021
q The average probability of error based on the
union bound is given by

q Using only the leading terms of the binomial

expansion:

7/23/2021
Minimum Shift Keying
(MSK)
MSK is a continuous phase-frequency shift keying;

Why MSK?
-- Exploitation of Phase Information besides
frequency.

7/23/2021
Reason for Minimum Shift Keying, MSK
It is found that binary data consisting of sharp transitions between "one" and
"zero" states and vice versa potentially creates signals that have sidebands
extending out a long way from the carrier, and this creates problems for many
radio communications systems, as any sidebands outside the allowed
bandwidth cause interference to adjacent channels and any radio
communications links that may be using them.
MSK, minimum shift keying has the feature that there are no phase
discontinuities and this significantly reduces the bandwidth needed over other
forms of phase and frequency shift keying.

7/23/2021
• The minimum-shift keying (MSK) scheme is used in GSM, a pioneer and a
widely-used digital cellular mobile system. MSK can be viewed as either a
special case of binary continuous-phase frequency-shift keying (CPFSK) or a
special case of OQPSK.
• In expressing MSK as a special case of CPFSK, the change in carrier
frequency occurring every bit duration is equal to one half the bit rate,
which is the minimum separation possible for the two sinusoidal carriers to
be coherently orthogonal.

7/23/2021
 Representation of a MSK signal

7/23/2021
Signal Space Representation of
MSK

7/23/2021
MSK
Transmitter
mI(t)

mQ(t)

7/23/2021
MSK Receiver

7/23/2021
 Waveforms for MSK

Input x(t)

7/23/2021
7/23/2021
MSK wave

7/23/2021
7/23/2021
7/23/2021
 2. A system transmits binary data at the rate 2.5 x 106 bits/sec. During the
course of transmission white Gaussian Noise of zero mean and PSD 10-11
W/Hz is added to signal. In the absence of noise amplitude of sinusoidal
wave for digit 1 or 0 is 1 mV. Determine average probability of error for
following systems

i) Coherent BPSK
Pe = ½ erfc(√E b /No )

ii)16 PSK

iii) 16 QAM
7/23/2021
7/23/2021
7/23/2021
 Intersymbol
q Interference
Intersymbol interference (ISI) occurs when a pulse spreads out in such a way that it
interferes with adjacent pulses at the sample instant.
q Example: assume polar NRZ line code. The channel outputs are shown as spreaded (width Tb
becomes 2Tb) pulses shown (Spreading due to bandlimited channel characteristics).
Channel Input Channel Output
Pulse width Tb Pulse width Tb
Data 1

− 0 Tb − 0 Tb
Tb Tb
Data 0
− 0 Tb − 0 Tb
Tb Tb

7/23/2021
 Inter-Symbol Interference (ISI)
q ISI in the detection process due to the filtering effects of
the system

q Overall equivalent system transfer function

H ( f ) = Ht ( f )Hc ( f )Hr ( f )

q creates echoes and hence time

dispersion

q causes ISI at sampling time

z k = s k + n k + ∑ αi s i
i≠ k

7/23/2021
 Inter-symbol
qinterference
Baseband system model
x1 x2
zk
{x } k Tx filter Channel r(t) Rx. filter {xkˆ
ht (t) hc (t) hr (t) Detecto
T Ht ( f ) Hc ( f ) Hr ( f )
t = kT
r }
x3 T n(t)
q Equivalent model

x1 x2
Equivalent system zk
{x }
k
z(t) {xkˆ
h(t) Detecto
T H(f) t = kT
r }
x3 T ˆ
n (t)
filtered noise
H ( f ) = Ht ( f )Hc ( f )Hr ( f )

7/23/2021
Intersymbol
Interference
q For the input data
stream: 1 1 0 1 0 1
A
0 Tb 2Tb 3Tb 4Tb 5Tb

q The channel output is the superposition of each bit’s

output:

0 Tb 2T 3T 4T 5T
Resultant
b b b b
Channel
1 0 1 1 0 1
Output
Waveform

0 Tb 2T 3T 4T 5T
b b b b

7/23/2021
 Pulse Shaping to reduce Interchannel and
Intersymbol interference
q Duobinary Encoding : Waveform transmitted is combination of two bits.

q Nyquist Criterion for Distortionless Baseband Binary Transmission

q QPR Encoding and decoding

By reduction in bandwidth we canhave reduction in ICI and fit

more channels in given spectrum

7/23/2021
 Duobinary Encoding

Causes reduction in
bandwidth

7/23/2021
Frequency of d(k) = 1/2Tb
Frequency of VD(k) = 1/4Tb

7/23/2021
Nyquist’s Criterion for Distortionless Baseband Binary
Transmission
▪ The receiver reconstructs data sequence {bk} by extracting and
decoding the corresponding sequence of weights {Ak} from the
output y(t).

▪ The extraction involves sampling the output y(t) at some time

t=iTb .

▪ The decoding requires that the weighted pulse contribution as

shown in equation be free from ISI due to overlapping trails of
other weighted contributions by k≠i.

52
7/23/202
Nyquist’s Criterion for Distortionless Baseband Binary
Transmission
This requires that we control the received
pulse p(t), 𝑖=𝑘 Equation (A)
( ) 1,
𝑝 𝑖𝑇𝑏 − 𝑘𝑇𝑏 = { 0, 𝑖 ≠ 𝑘 Time domain
∞
condition for
If so, ( )
𝑦 𝑡 = 𝜇𝐴 𝑖 + 𝜇
(
∑ 𝐴𝑘 𝑝 𝑖𝑇𝑏 − 𝑘𝑇𝑏
) will be
zero ISI
𝑘=−∞
𝑘≠𝑖

( )
𝑦 𝑡 = 𝜇𝐴𝑖
This implies zero ISI. Condition of Equation (A)
assures
perfect reception in the absence of noise.
Nyquist’s Criterion for Distortionless Baseband
Binary
Transmission
1, Equation (A)
( ) = {𝑖 = 𝑘
𝑝 𝑖 − 𝑘 𝑇𝑏 Time domain
0, 𝑖 ≠ 𝑘
condition for
Frequency domain condition for the perfect reception
zero ISI
Let{p(nTb)} represents the impulses at which p(t) is sampled for
decision at a rate of Tb. The Fourier transform of these impulses
is given by, ∞
( )
𝑃 𝛿 𝑓 = 𝑅𝑏
R =1/Tb is bit
∑ 𝑃(𝑓 − 𝑛𝑅 𝑏 ) Equation ( ) b
𝑛 rate is spectrum
P(f)
∞ =−∞ B
of p(t)
( ) ( ) ( )
𝑃𝛿 𝑡 = ∑ 𝑝 𝑛𝑇𝑏 𝛿 𝑡 − 𝑛𝑇 𝑏 ∞
𝑛 =−∞ ( )
𝑃𝛿 𝑓 = ∫ 𝑃𝛿 (𝑡)𝑒−𝑗 2𝜋𝑓𝑡 𝑑𝑡
78//2234/ Department of ECE 71
Nyquist’s Criterion for Distortionless Baseband
Binary
Transmissio
∞ ∞
n
( ) ( ) ( )
𝑃𝛿 𝑓 = ∫ ∑ 𝑝 𝑛𝑇𝑏 𝛿 𝑡 − 𝑛𝑇 𝑏 𝑒−𝑗 2𝜋𝑓𝑡 𝑑𝑡
−∞ 𝑛=−∞
use, n=i-k (refer
equation A) ∞ ∞
( ) ( ) ( )
𝑃𝛿 𝑓 = ∫ ∑ 𝑝 [𝑖 − 𝑘]𝑇 𝑏 𝛿 𝑡 − [𝑖 − 𝑘]𝑇𝑏 𝑒−𝑗 2𝜋𝑓𝑡 𝑑𝑡
−∞ 𝑛=−∞

(
Substitute, 𝑝 𝑖 − 𝑘 𝑇 𝑏
) = {1,0, 𝑖𝑖≠
=𝑘
𝑘 in the above equation
∞

I‫𝑒 𝑡 𝛿 𝑜 𝑝 ∫ ﻟ‬−𝑗 2𝜋𝑓𝑡 𝑑𝑡,

( ) ( ) 𝑖=𝑘
−∞
( )
𝑃𝛿 𝑓 ❪ ∞
= I ( ) 𝑖≠𝑘
∫ 0. 𝛿 𝑡 𝑒−𝑗 2𝜋𝑓𝑡 𝑑𝑡 ,
𝗅 −∞
7/23/2021 55
 Nyquist’s Criterion for Distortionless Baseband
Binary
Transmissio
∞ ∞
n
( ) ( ) −𝑗 2𝜋𝑓𝑡 ( ) ( )
𝑃𝛿 𝑓 = ∫ 𝑝(𝑜)𝛿 𝑡 𝑒 𝑑𝑡 = 𝑝 𝑜 ∫ 𝛿 𝑡 𝑒−𝑗 2𝜋𝑓𝑡 𝑑𝑡 = 1
−∞ −∞ Equation
(C)
Using Equation (C) in (B), we get
∞
( )
𝑅𝑏 ∑ 𝑃 𝑓 − 𝑛𝑅 𝑏 = 1
𝑛=−∞
∞ Equation
1 (D)
( )
∑ 𝑃 𝑓 − 𝑛𝑅 𝑏 = 𝑅𝑏 =
𝑇𝑏
Equation (D) is the Frequency domain
𝑛 =−∞ condition for zero
ISI (perfect reception) in the absence of noise.

7/23/2021 56
 (1) IDEAL SOLUTION

𝑅𝑏
𝐵0 = 2

1 𝑓
( )
𝑃 𝑓 = 2𝐵 𝑟𝑒𝑐𝑡 2𝐵0 )
0(
Signal waveform that produces zero
( ) ISI is defined by sinc function.
𝑝( 𝑡 𝑠𝑖𝑛 2𝜋𝐵 0𝑡 ( )
2𝜋𝐵 0 𝑡 =sinc 2𝐵 0𝑡
=
)
If the waveform y(t) is sampled
at the instants time t=0,
of pulses will
not
±Tb,interfere
±2Tb, ..with
then each
other

7/23/2021
 (1) IDEAL SOLUTION

Although this choice of pulse p(t) achieves economy in

bandwidth in that it solves the problem of zero
intersymbol interference with the minimum bandwidth
possible.
There are two practical difficulties
1. It requires that the amplitude characteristics of
P(f) flat from
–B0 to +B0. this is physically unrealizable because of
the abrupt transitions at ±B 0.
2. The function p(t) decreases at a slow rate of decay that is
caused by the discontinuity of P(f) at ±B 0
58
7/23/202
 (2) PRACTICAL SOLUTION

To overcome the practical difficulties posed by the ideal

solution by extending the bandwidth from B0=Rb/2 to an
adjustable value between B0 and 2B0.

While doing so we permit three components in the series

on the left side of equation (D),
1
( ) ( ) ( ) , −𝐵 ≤ 𝑓 ≤ 𝐵
𝑃 𝑓 + 𝑃 𝑓−2𝐵0 + 𝑃 𝑓+2𝐵0 = 2𝐵 0 0 0

A particular form of P(f) that embodies many desirable

features is constructed by a raised cosine spectrum.

59
7/23/202
 60
7/23/202
 Frequency Response

Time Response

7/23/2021
 QPR Encoder

If QPSK is used with d(t), bandwidth would be f b.

With QPR bandwidth is fb /4

7/23/2021
 QPR Decoder

7/23/2021
Some Issues in Transmission and
Reception
1. Attenuation, Filtering and Corruption Solution:
Regenerative Repeater

2. Noise, Jitter, ISI: Eye Pattern for

measurement
3. Channel Distortion: Solution: Equalizers

7/23/2021
 Equalizers: Combating
qISIEqualization is the process of nullifying the adverse effect of
a communication channel
q Any real channel needs equalization to approach flat
frequency response
q Equalizers need experimental adjustment in the field
because we don’t know the channel characteristics exactly

q System Performance can be improved using equalizer

7/23/2021
 Equalizing filters
q…Baseband system model
a1
∑k akδ (t − kT) Tx filter Channel r(t) Equalizer Rx. filter z(t) zk {a ˆ k
ht (t) hc (t) he (t) hr (t) Detecto
t = kT
r
}
T a2 a Ht ( f ) Hc ( f ) He ( f ) Hr ( f )
n(t)

q 3Equivalent model
H ( f ) = Ht ( f )Hc ( f )Hr ( f )
a1
Equivalent system zk
∑k a δ (t − kT)
z(t) x(t) Equalizer z(t) {a ˆ k
k
h(t) he (t) Detecto
t = kT
r
}
H(f) He ( f )
T a2 a ˆ
n (
3 t) noise
filtered
ˆ
n (t) = n(t) * hr
(t)
7/23/2021
 Practicall
q Channel frequency transfer function is not known.
y
q Usually transmit and receive filters are matched H

(f) = H t(f) H r(f)

q Overall System function is written as H (f) =

Ht(f)Hc(f) H r(f) H e(f)

q Hence Equalizer transfer function should

compensate for channel distortion

7/23/2021
 Equalization –
qcont’d
Equalization using
q MLSE (Maximum likelihood sequence estimation)
q Filtering
q Transversal filtering
q Zero-forcing equalizer

q Minimum mean square error (MSE) equalizer

q Decision feedback

q Using the past decisions to remove the ISI contributed

by them
q Adaptive equalizer

7/23/2021
 Equalization by transversal
• filtering
Transversal filter:
– A weighted tap delayed line that reduces the effect of ISI
by
proper adjustment of the filter taps.

7/23/2021
 Transversal equalizing filter
q Zero-forcing equalizer:
q
… The filter taps are adjusted such that the equalizer
output is forced to be zero at N sample points on each
side:
Adjust k=0
⎧1
z(k ) =⎨
{c N
n n=− N ⎩0 k = ± 1,...,± N
}
q Mean Square Error (MSE) equalizer:
q The filter taps are adjusted such that the MSE of ISI and
noise power at the equalizer output is minimized.
Adjust

{c n} N
[
min E (z(kT) − ka 2

]
n=− N

)
7/23/2021
 Orthogonal Frequency Division
Multiplexing OFDM

7/23/2021
 Advantages of
q
OFDM
Allows carriers to overlap (no guard band), resulting in
lesser wasted bandwidth without any Inter Carrier Interference (ICI)

q High data rate distributed over multiple carriers resulting in lower

symbol rate (more immune to ISI)

q Permits higher data rate as compared to FDM

q Increased security and bandwidth efficiency possible using
CDMA – OFDM (MC-CDMA)

q Simple guard intervals make the system more robust to multipath

effects.
7/23/2021
 What is OFDM ??
Orthogonality in time domain…

7/23/2021
 What is OFDM ??
Orthogonality in frequency domain…

• Each carrier is modulated using BPSK / QPSK / M-ary QAM

• Frequency response for each carrier is a Sinc(X)
• Overlap of frequency response is possible as against FDM where inter-
carrier spacing is a must
• Frequency responses of the carriers overlap at zero crossings avoiding
Inter Carrier Interference
7/23/2021
OFDM Methodology

7/23/2021
7/23/2021
7/23/2021
7/23/2021
The IFFT takes in N source symbols at a time where N is the number of
subcarriers in the system. Each of these N input symbols has a symbol
period of T seconds.
Output of the IFFT is N orthogonal sinusoids. These orthogonal sinusoids
each have a different frequency and the lowest frequency is DC.
The input symbols are complex values representing the mapped
constellation point and therefore specify both the amplitude and phase of
the sinusoid for that subcarrier.
The IFFT output is the summation of all N sinusoids. Thus, the IFFT block
provides a simple way to modulate data onto N orthogonal subcarriers. The
block of N output samples from the IFFT make up a single OFDM symbol.

7/23/2021
 IEEE 802.11a:

Frequency Bands: 5.150-5.350 GHz and 5.725-5.825 GHz (12

channels) Modulation OFDM

Range: 100m

IEEE 802.11g

Frequency Bands: 2.412-2.472GHz

Modulation: OFDM
Range: 300m

7/23/2021
 Reference : https://rfmw.em.keysight.com/wireless/helpfiles/89600b/
webhelp/subsyste ms/wlan-ofdm/content/ofdm_basicprinciplesoverview.
htm
7/23/2021
7/23/2021
 Unit 3 related
experiments

7/23/2021
%Comparison of Constellation diagrams of 16-PSK and 16-QAM
%in presence of noise, SNR=20 dB
%16-PSK

M=input('Number_Symbols='); SNR=input('SNR of QPSK system

in dB='); x1=randint(1000,1,M);
y1=pskmod(x1,M); y1n=awgn(y1,SNR,'measured');
scatterplot(y1n); y1r=pskdemod(y1n,M); [num_error,
er_rate]=symerr(x1,y1r)
%Constellation diagram of QAM in presence of noise with SNR=5dB
y2=qammod(x1,M); y2n=awgn(y2,SNR,'measured');
scatterplot(y2n); y2r=qamdemod(y2n,M); [num_error,
er_rate]=symerr(x1,y2r)

7/23/2021
SNR = 20 dB

Scatter plot
4

Output: 2
No, of symbols = 16
1
SNR = 20 dB

Quadrature
0

-1
Scatter plot
-2

1 -3

-4
-4 -2 0 2 4
0.5
In-Phase
Quadratur

0
e

-0.5

-1

-1 -0.5 0 0.5 1
In-Phase

7/23/2021
 Scatter plot

5
No, of symbols = 4

16 SNR = 10 dB 3

Quadrature
0

-1

-2

-3

-4

-5
-5 0 5
In-Phase
Scatter plot

1.5

0.5
Quadratur

0
e

-0.5

-1

-1.5

-1.5 -1 -0.5 1 1.5

0 0.5
In-Phase

7/23/2021
 Reference
s Unit III: Digital Modulation-II
Generation, Reception, Signal Space T1 290-294, 300-312,
Representation and Probability of Error 602-610
Calculation for Quadrature Amplitude Shift
Keying (QASK), M-ary FSK (MFSK), Minimum Shift
Keying (MSK)

Pulse Shaping to reduce Interchannel and

Intersymbol Interference T1 316-319, 322-330

some Issues in transmission and reception T1 331-334

Orthogonal Frequency Division Multiplexing

(OFDM) T1 335-340

Comparison of digital modulation systems T1

614-616

7/23/2021
 Thank
You

7/23/2021