Fundamentals of

Digital Modulation
Dr. Ahmed Bassyouni
Research Professor
Electrical Engineering Dept.
Boise State University
 Outline
 Digital modulation overview
 Power spectrum density and some Fourier basics
 Nyquist Criterion and Raised cosine filter
 Digital Modulation Techniques
– Linear Modulation
BPSK, DPSK
 QPSK, offset QPSK, /4 QPSK

– Constant envelope modulation

BFSK, MSK, GMSK
– Combined linear and constant envelope modulation
 MPSK
 QAM
 MFSK and OFDM
 Spread Spectrum
 Modulation over Fading and Multipath Channel

[2]
 Digital Modulation
 The input is discrete signal
– Time sequences of pulses or symbols
 Offers many advantages
– Robustness to channel impairments
– Easier multiplexing of various sources of information: voice, data,
video.
– Can accommodate digital error-control codes
– Enables encryption of the transferred signals
 More secure link

[3]
 Digital Modulation Example

The modulating signal is represented as a time-sequence of symbols

or pulses.

Each symbol has m finite states: That means each symbol carries n bits
of information where n = log2m bits/symbol.

...
Modulator
0 1 2 3 T

One symbol
(has m states – voltage levels)
(represents n = log2m bits of information)

[4]
 Factors that Influence Choice of Digital
Modulation Techniques
 A desired modulation scheme
– Provides low bit-error rates at low SNRs
 Power efficiency
– Performs well in multipath and fading conditions
– Occupies minimum RF channel bandwidth
 Bandwidth efficiency
– Is easy and cost-effective to implement
 Depending on the demands of a particular system or
application, tradeoffs are made when selecting a digital
modulation scheme.

[5]
 Power Efficiency of Modulation
 Power efficiency is the ability of the modulation technique to
preserve fidelity of the message at low power levels.
 Usually in order to obtain good fidelity, the signal power needs
to be increased.
– Tradeoff between fidelity and signal power
– Power efficiency describes how efficient this tradeoff is made

 Eb 
Power Efficiency :  p   required at the receiver input for certain PER 
 N0 
Eb: signal energy per bit
N0: noise power spectral density
PER: probability of error

[6]
 Bandwidth Efficiency of Modulation
 Ability of a modulation scheme to accommodate data within a
limited bandwidth.
 Bandwidth efficiency reflect how efficiently the allocated
bandwidth is utilized

R
Bandwidth Efficiency :  B  bps/Hz
B
R: the data rate (bps)
B: bandwidth occupied by the modulated RF signal

[7]
 Shannon’s Bound
 There is a fundamental upper bound on achievable bandwidth
efficiency. Shannon’s theorem gives the relationship between
the channel bandwidth and the maximum data rate that can be
transmitted over this channel considering also the noise present
in the channel.
Shannon’s Theorem

C S
 B max   log 2 (1  )
B N
C: channel capacity (maximum data-rate) (bps)
B: RF bandwidth
S/N: signal-to-noise ratio (no unit)

[8]
 Tradeoff between BW Efficiency and Power Efficiency

 There is a tradeoff between bandwidth efficiency and power

efficiency
– Adding error control codes
 Improves the power efficiency
 Reduces the requires received power for a particular bit

error rate
 Decreases the bandwidth efficiency
 Increases the bandwidth occupancy

– M-ary keying modulation

 Increases the bandwidth efficiency
 Decreases the power efficiency
 More power is requires at the receiver

– M-FSK keying modulation

 Increase the power efficiency
 Decrease the bandwidth efficiency

[9]
 Example
 SNR for a wireless channel is 30dB and RF bandwidth is
200kHz. Compute the theoretical maximum data rate that can
be transmitted over this channel?
 Answer:
 30dB 
S  10 
 10  
N
S
C  B log 2 (1  )  2 x105 log 2 (1  1000)  1.99Mbps
N

 Example 6.6
 Example 6.7

[10]
 Noiseless Channels and Nyquist Theorem
 For a noiseless channel, Nyquist theorem gives the relationship
between the channel bandwidth and maximum data rate that
can be transmitted over this channel

Nyquist Theorem

C  2B log 2 m
C: channel capacity (bps)
B: RF bandwidth
m: number of finite states in a symbol of transmitted signal

Example: A noiseless channel with 3kHz bandwidth can only transmit

maximum of 6Kbps if the symbols are binary symbols.

[11]
 Power Spectral Density of Digital Signals and Bandwith

 What does signal bandwidth mean?

 Answer is based on Power Spectral Density (PSD) of Signals
 For a random signal w(t), PSD is defined as:

W (f)2 
Pw( f )  lim  T 

T  T
 
WT ( f ) is th fourier transform of wT (t )
 T T
w(t )   t 
wT (t )   2 2
0 elsewhere

[12]
 Fourier Analysis
 Joseph Fourier has shown that any periodic function F(f) with period T, can
be constructed by summing a (possibly infinite) number of sin’s and cos’s.

a0 
F (t )   an cos(nT t )  bn sin(nT t )
2 n 1
2
T   2f 0
T

 Such a decomposition is called Fourier series and the coefficients are called
the Fourier coefficients.
 A line graph of the amplitudes of the Fourier series components can be
drawn as a function of frequency. Such a graph is called a spectrum or
frequency spectrum. f0 is called the fundamental frequency.
 The nth term is called nth harmonic. The coefficients of the nth harmonic are
an and bn.

[13]
 Fourier Analysis
 The coefficients can be obtained from the periodic function F(t)
as follows:
T
2
a0   F (t )dt
T 0
T
2
an   F (t ) cos n T tdt , n  1,2,...
T 0
T
2
bn   F (t ) sin n T tdt , n  1,2,...
T 0

[14]
 Example: A Periodic Function
 Find the Fourier series of the periodic function f(x), where one
period of f(x) is defined as: f(x) = x, -p < x < p

T=2


 0  2



T  2

[15]
 Example: Its Fourier Approximation
Domain: [-, ]
4 4
x x

3 1 harmonic
2*sin(x)
3 2 harmonics 2*(sin(x)-sin(2*x)/2)

2 2

1 1

0 0

-1 -1

-2 -2

-3 -3

-4 -4
-3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3

4 4
x x
2*(sin(x)-(sin(2*x)/2)+(sin(3*x)/3)) 2*(sin(x)-(sin(2*x)/2)+(sin(3*x)/3)-(sin(4*x)/4))
3 3 harmonics 3 4 harmonics
2 2

1 1

0 0

-1 -1

-2 -2

-3 -3

-4 -4
-3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3

[16]
 Example: Frequency Spectrum
Magnitude : an2  bn2
2.5

For First 10 harmonics

2

1.5

0.5

0
0 1 2 3 4 5 6 7 8 9 10 11 12
Harmonics

Each harmonic corresponds to a frequency that is multiple of the fundamental

frequency

[17]
 Complex Form of Fourier Series

By substituting Euler’s Expression into Fourier expansion:

e j  cos   j sin 
It can be shown that the following is true:

F (t )   n dt
c e jnT t

n  
T
2
1

 jnT t
cn  F (t ) e dt , n  ...  2,1,0,1,2,....
T T

2
cn are the complex Fourier coefficients

[18]
 Digital Modulation - Continues
 Line Coding
– Base-band signals are represented as line codes

1 0 1 0 1 0 1
Tb
V
0 Unipolar
Tb NRZ
V
Bipolar
RZ
-V
V
Manchester
-V NRZ

Tb

[19]
PSD of various line codes

[20]
Summary of line coding schemes

Plus HDB3 and B8ZS

[21]
 Baseband binary data transmission system.
 ISI arises when the channel is dispersive
 Frequency limited -> time unlimited -> ISI
 Time limited -> bandwidth unlimited -> bandpass channel ->
time unlimited -> ISI

p(t)

[22]
 ISI Example
sequence sent 1 0 1
sequence received 1 1(!) 1

Signal received

Threshold

0 t

-3T -2T -T 0 T 2T 3T 4T 5T

Sequence of three pulses (1, 0, 1)

sent at a rate 1/T
[23]
 ISI

 Nyquist three criteria

– Pulse amplitudes can be detected
correctly despite pulse spreading or
overlapping, if there is no ISI at the
decision-making instants
 1: At sampling points, no ISI
 2: At threshold, no ISI
 3: Areas within symbol period is
zero, then no ISI
– At least 14 points in the finals
 4 point for questions
 10 point like the homework

[24]
 1st Nyquist Criterion: Time domain
p(t): impulse response of a transmission system (infinite length)
p(t)
1
 shaping function

0 no ISI !
t
1
T
2 fN t0 2t0

Equally spaced zeros,

1
-1 interval T
2 fn

[25]
 1st Nyquist Criterion: Time domain

Suppose 1/T is the sample rate

The necessary and sufficient condition for p(t) to satisfy

1, n  0
pnT   
0, n  0
Is that its Fourier transform P(f) satisfy


 P f  m T   T
m  

[26]
1st Nyquist Criterion: Frequency domain


 P f  m T   T
m  

f
0 fa  2 f N 4 fN
(limited bandwidth)
[27]
 Proof

Fourier Transform pt    P f exp  j 2ft df


At t=T pnT    P f exp  j 2fnT df

  2 m 1 2T
pnT     P f  exp  j 2fnT df
2 m 1 2T
m  


 P f  m T  exp  j 2fnT df

1 2T

1 2T
m  


 P f  m T exp  j 2fnT df

1 2T

1 2T
m  



1 2T
B f  exp  j 2fnT df B f    P f  m T 
1 2T m  

[28]
 Proof
 
B f    P f  m T  B f    b exp  j 2nfT 
n
m   n  

B f exp  j 2nfT 
1 2T
bn  T 
1 2T

bn  Tp nT  T n  0
bn  
0 n  0

B f   T  P f  m T   T
m  

[29]
 Sample rate vs. bandwidth
 W is the bandwidth of P(f)
 When 1/T > 2W, no function to satisfy Nyquist condition.

P(f)

[30]
 Sample rate vs. bandwidth

sin t T  t  T ,  f  W 
pt    sinc  P f    ,
t T  0, otherwise 

[31]
 Sample rate vs. bandwidth
 When 1/T < 2W, numbers of choices to satisfy Nyquist
condition

 A typical one is the raised cosine function

[32]
 Cosine rolloff/Raised cosine filter
 Slightly notation different from the book. But it is the same
sin( Tt ) cos( r Tt )
prc0 (t )  
T t
1  (2 r Tt ) 2

r : rolloff factor 0  r 1

1 f  (1  r ) 21T

Prc0 ( j 2f )  1
2
1  cos( 
2r ( Tf  r  1))  if 1
2T (1  r )  f  21T (1  r )

0 f  1
2T (1  r )

[33]
 Raised cosine shaping
 Tradeoff: higher r, higher bandwidth, but smoother in time.
 W
P(ω)
r=0
r = 0.25
r = 0.50
r = 0.75
r = 1.00

W 2w ω
p(t)

π π
 
W W
0

0 t

[34]
Modulated time domain

[35]
 Cosine rolloff filter: Bandwidth efficiency
 Vestigial spectrum

data rate 1/ T 2 bit/s

 rc   
bandwidth (1  r ) / 2T 1  r Hz

bit/s 2 bit/s
1   2
Hz (1  r ) Hz
 
2nd Nyquist (r=1) r=0

[36]
 2nd Nyquist Criterion
 Values at the pulse edge are distortion-less
 p(t) =0.5, when t= -T/2 or T/2; p(t)=0, when t=(2k-1)T/2, k≠0,1
-1/T ≤ f ≤ 1/T

Pr ( f )  Re[  (1) n P ( f  n / T )]  T cos( fT / 2)
n  

PI ( f )  Im[  (1) n P( f  n / T )]  0
n  

[37]
Example

[38]
 3rd Nyquist Criterion
 Within each symbol period, the integration of signal (area) is
proportional to the integration of the transmit signal (area)
 ( wt ) / 2 
 ,w
 sin( wT / 2) T
P( w)  
 0, 
 w 
 T
 /T
1 ( wt / 2)
p(t )  
jwt
e dw
2  / T
sin( wT / 2)

2 n1T
1, n0
A  2 n1 p(t )dt  
2

2
T
0, n0
[39]
 Cosine rolloff filter: Eye pattern

2nd Nyquist
1st Nyquist: 1st Nyquist:

2nd Nyquist: 2nd Nyquist:

1st Nyquist

1st Nyquist: 1st Nyquist:

2nd Nyquist: 2nd Nyquist:

[40]
 Gaussian Pulse Shaping Filter
 Tradeoff between bandwidth and ISI
 Example 6.8

[41]
 Gaussian minimum shift keying
 GMSK is similar to MSK except it incorporates a premodulation Gaussian
LPF

 Achieves smooth phase transitions between signal states which can

significantly reduce bandwidth requirements
 There are no well-defined phase transitions to detect for bit synchronization
at the receiving end.
 With smoother phase transitions, there is an increased chance in intersymbol
interference which increases the complexity of the receiver.
 Used extensively in 2nd generation digital cellular and cordless telephone
apps. such as GSM

[42]
 spread-spectrum transmission
 Three advantages over fixed spectrum
– Spread-spectrum signals are highly resistant to noise and
interference. The process of re-collecting a spread signal spreads
out noise and interference, causing them to recede into the
background.
– Spread-spectrum signals are difficult to intercept. A Frequency-
Hop spread-spectrum signal sounds like a momentary noise burst
or simply an increase in the background noise for short
Frequency-Hop codes on any narrowband receiver except a
Frequency-Hop spread-spectrum receiver using the exact same
channel sequence as was used by the transmitter.
– Spread-spectrum transmissions can share a frequency band with
many types of conventional transmissions with minimal
interference. The spread-spectrum signals add minimal noise to
the narrow-frequency communications, and vice versa. As a
result, bandwidth can be utilized more efficiently.
[43]
 PN Sequence Generator
 Pseudorandom sequence
– Randomness and noise properties
– Walsh, M-sequence, Gold, Kasami, Z4
– Provide signal privacy

[44]
 Direct Sequence (DS)-CDMA
 It phase-modulates a sine wave pseudo-randomly with a
continuous string of pseudo-noise code symbols called "chips",
each of which has a much shorter duration than an information
bit. That is, each information bit is modulated by a sequence of
much faster chips. Therefore, the chip rate is much higher than
the information signal bit rate.
 It uses a signal structure in which the sequence of chips
produced by the transmitter is known a priori by the receiver.
The receiver can then use the same PN sequence to counteract
the effect of the PN sequence on the received signal in order to
reconstruct the information signal.

[45]
 Direct Sequence Spread Spectrum
 Unique code to
differentiate all users
 Sequence used for
spreading have low
cross-correlations
 Allow many users to
occupy all the
frequency/bandwidth
allocations at that
same time
 Processing gain is the
system capacity
– How many users
the system can
support

[46]
 Spreading & Despreading
 Spreading
– Source signal is multiplied by a PN signal: 6.134, 6.135
Ts ChipRate
 Processing Gain: Gp  
 Despreading Tc DataRate
– Spread signal is multiplied by the spreading code
 Polar {±1} signal representation

[47]
Direct Sequence Spreading

[48]
Spreading & Despreading

[49]
 CDMA – Multiple Users
 One user’s information
is the other’s
interferences
 If the interference
structure can be
explored, multiuser
detection
– Match filter
– Decorrelator
– MMSE decodor
– Successive cancellation
– Decision feedback

[50]
 Frequency Hopping Spread Spectrum
 Frequency-hopping
spread spectrum
(FHSS) is a spread-
spectrum method of
transmitting radio
signals by rapidly
switching a carrier
among many
frequency channels,
using a
pseudorandom
sequence known to
both transmitter and
receiver.
 Military, bluetooth

[51]
 Baseline: Stationary Channel

y: the received signal

x: the transmitted signal with
BPSK modulation amplitude a
w: white noise with power N0

[52]
 Baseline: Stationary Channel

y: the received signal

BPSK modulation x: the transmitted signal with
amplitude a
w: white noise with power N0

a
Pe  Q( )  Q( 2SNR )  e  SNR
N0 / 2
 2
1 a

u / 2
where, Q(x)  du, SNR 
2
e
2 x N0
Error probability decays exponentially with signal-noise-ratio
(SNR).

[53]
 Flat Fading Channel

Assume h is Gaussian random:

BPSK:

Conditional on h,

6.154, 6.155

Averaged over h, which follows chi-square distribution

at high SNR.

[54]
Irreducible Bit Error Rate due to multipath

[55]
Simulation of Fading and Multipath

[56]
Irreducible BER due to fading

[57]
Irreducible BER due to fading

[58]
BER due to fading & multipath

[59]
Questions?

[60]
 Biography
Dr. Ahmed Bassyouni has received the B.Sc. degree in Electrical Engineering, M.Sc. and
Ph.D. degrees in Adaptive controls from the Faculty of Engineering, Alexandria
University in 1982, 1987, and 1992. Post Doctoral Research on RF Transceivers with the
Caltech Institute, and Boise State University in 1995, and 2001.
Dr. Bassyouni is an expert of RF and Microwave applications oriented to air defense
systems technology, and a designer of complex sensors such as phased array surveillance
radars, fire control radars, MTI, SAR, Airborne, Mortars Fires tracking radars, and
integrated UAV battle field systems. Currently he is a consultant with the Sensors and
Systems in New York. He was a Research Professor with the Department of Electrical
Engineering, Boise State University, Idaho (1997-2002); Visitor Professor with the
Department of Electrical Engineering at the University of Arkansas in (1993-1995);
Teaching Assistant (1987-1992); Instructor of Engineering Science with the Air Defense
College in Alexandria, Egypt (1981-1987).
Dr. Bassyouni is a senior member, technical committee member, and co-chair conferences
participant, of the IEEE, International Society for Optical Engineering (SPIE), (American
Institute of Aeronautics and Astronautics) AIAA, (Society of Manufacturing Engineers)
SME, and (Robotic Institute) RI, the New York Academy of Sciences, and the American
Society of Engineering Education (ASEE). Also, He is an active member with the
American Society of Quality (ASQ), the American Association for the Advancement of
Science (AAAS), and the International Association of Online Engineering (IAOE). He is a
member of Tau Beta Pi (The Engineering Honor Society), Sigma Xi (The Scientific
Research Society), Phi Beta Delta (The International Honor Society). He is the author and
coauthor of seven books, and over 120 papers, trades analysis documents and
presentations in the areas of adaptive controls, robotics, RF/microwave systems, and radar
systems. He leads the radars and missile control systems projects from the requirements to
the prototype field testing, participating with innovative ideas, and advanced design
techniques. He is a Certified Reliability Engineer (C.R.E.), and a Certified System
Engineer (CSE). Dr. Bassyouni provides consulting service to the USA companies, and
organizations of defense industry. He has been awarded over 25 research grants to pursue
his work in integrated sensors.

[61]