ME1100 DIGITAL RF COMMUNICATIONS

This courseware product contains scholarly and technical information and

is protected by copyright laws and international treaties. No part of this
publication may be reproduced by any means, be it transmitted,
transcribed, photocopied, stored in a retrieval system, or translated into
any language in any form, without the prior written permission of Acehub
Vista Sdn. Bhd.  

The use of the courseware product and all other products developed
and/or distributed by Acehub Vista Sdn. Bhd. are subject to the applicable
License Agreement.

For further information, see the Courseware Product License Agreement.

http://dreamcatcher.asia/cw

1
4. Baseband Pulse Transmission and
Digital Modulation Techniques

2
 Definition of Baseband
• In communications, baseband is the term that describes
signals and systems whose range of frequencies is
measured from close to 0 Hz to a cut-off frequency, a
maximum bandwidth or highest signal frequency.
• It is sometimes used as a noun for a band of frequencies
starting close to zero.
• Baseband can often be considered a synonym to low-pass
or non-modulated.

fc

BW 3
 Possible Types of Signal Transmission
Analog signal and baseband transmission
Baseband channel
Analog Analog * microphone
source destination

Analog transmission using modulation and demodulation

channel
Analog Modulator Demodulator Analog
source (Tx) (Rv) destination

* Radio, TV

Digital signal transmission on digital channel

Digital channel
Digital Digital
Coder Decoder
source destination
* Data file in computer
4
Possible Types of Signal Transmission (Cont)
Digital signal transmission by modem
Analog channel
Digital Digital
Modem Modem
source destination
* Data through internet
Analog signal digitized transmission
Digital channel
Analog A/D Decoding Analog
source & coding & D/A destination
* Digitized voice, CD

Analog signal digitized transmitted by modem

Analog channel
Analog A/D Decoding Analog
Modem Modem
source & coding & D/A destination
* Modern comm system
5
 Baseband Pulse Transmission
• In baseband transmission, a bit stream (e.g., a PCM
signal) is encoded into pulses without modulating a
high-frequency carrier. The digital pulses are then
transmitted over a baseband or a low pass channel.

• The main concern in baseband pulse transmission is

shaping the transmitted pulses to bring the Inter-
Symbol Interference (ISI) under control.

• In baseband transmission, the required bandwidth is

proportional to the bit rate; to send bits faster, more
bandwidth is required.

6
 Baseband Pulse Transmission (cont.)
Digital Signal :101011..
101011..

5V 5V
Line
Code
0V Time Recovery
Wide bandwidth Circuit

Pulse Transmission
Shaping Medium
(Wired)
Pulse
Time
7
 Bandwidth of Baseband Channel
Amplitude

Frequency

0 f
Baseband or Low-Pass Channel with Wide-bandwidth

Amplitude

Frequency

0 f
Baseband or Low-Pass Channel with Narrow-bandwidth
8
 Bandwidth of Baseband Channel (cont.)
• Baseband transmission of a digital signal that preserves the
shape of the digital signal is possible only if the low-pass
channel has an infinite or very wide bandwidth.

Input Signal Bandwidth Output Signal

(Ideal Pulse) Bandwidth
f f f
0 f1 f2 f1 f2
∞
t Wide-bandwidth t
Channel
Input Signal Output Signal
(Ideal Pulse)

9
 Digital vs. Analog
Analog: Continuous signal Digital : Discrete signal
1.5 1.5

1 1

0.5 0.5

Voltage
Voltage

0 0

-0.5 -0.5

-1 -1

-1.5 -1.5
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
Time Time

Every amplitude Decision on symbol from

needed for reception a pre-defined alphabet
10
 Line Coding
• Binary 1’s and 0’s, such as the PCM bit stream, may
be represented in various rectangular pulses called
line codes.
• There are two major categories of line codes: RZ
(return-to-zero) and NRZ (nonreturn-to-zero).
• In RZ coding, a waveform returns to zero-volt level for
a portion (usually at the middle) of the bit interval.

11
 Binary Line Codes
• Unipolar Signaling: A binary 1 is represented by a positive
level (+A volts) and a binary 0 by zero level.

• Polar Signaling: Binary 1’s and 0’s are represented by equal

positive and negative levels (+A and –A volts), respectively.

• Bipolar Signaling: Binary 1’s are represented by alternate

positive and negative levels. The binary 0’s are represented by
zero level. Also called Alternate Mark Inversion (AMI).

• Manchester Signaling: Each binary 1 is represented by a

positive half-bit period pulse followed by a negative half-bit
period pulse, while a binary 0 is represented by a negative half-
bit period pulse followed by a positive half-bit period pulse. Also
called Split Phase Encoding.

12
 Binary Line Codes (cont’d)
Binary bits : 0 0 0 1 1 0 1 0 1
Volts (V)
5
Unipolar NRZ
0 Time (s)

5 Volts (V)

Polar NRZ
0 Time (s)

-5
Bit period - Tb
Volts (V)
5
Unipolar RZ
0 Time (s)
Volts (V)
5
Bipolar RZ
0 Time (s)

-5
Volts (V)
5
Manchester 0 Time (s)
NRZ
-5

Figure 4.1 – Binary Signaling Format 13

 Line Code Properties
• Self-synchronisation: When there is a long string of 1’s or 0’s
(DC component), a receiver is unable to identify a bit interval.
The receiver will lose bit synchronization with the transmitter. A
line code (e.g., Manchester NRZ) builds in a timing information
by having a transition within a bit period. The receiver will extract
this timing information to identify a bit interval for maintaining bit
synchronization with the transmitter.
• Low probability of the bit error: A line code is able to reduce
the effects of noise and ISI during transmission over a baseband
channel. This leads to reduced bit errors and hence low bit error
probability.
• Transparency: The data protocol and line code are designed so
that every possible sequence of data can be transmitted and
received.

14
 Unipolar NRZ
• Advantages:
• Simplicity in implementation.
• Does not require a lot of bandwidth for transmission.

• Disadvantages:
•Presence of DC level (indicated by spectral line at 0 Hz).
•Contains low frequency components.
•Does not have any error correction capability.
•Does not posses any clocking component for ease of
synchronisation.
•Is not Transparent. Long string of zeros causes loss of
synchronisation.

15
 Unipolar RZ
• Advantages:
• Simplicity in implementation.
• Presence of a spectral line at symbol rate which can be
used as symbol timing clock signal.

• Disadvantages:
•Presence of DC level (indicated by spectral line at 0 Hz).
•Continuous part is non-zero at 0 Hz.
•Does not have any error correction capability.
•Occupies twice as much bandwidth as Unipolar NRZ.
•Is not Transparent

16
 Polar NRZ
• Advantages:
• Simplicity in implementation.
• No DC component.

• Disadvantages:
•Continuous part is non-zero at 0 Hz.
•Does not have any error correction capability.
•Does not posses any clocking component for ease of
synchronisation.
•Is not transparent.

17
 Bipolar RZ
• Advantages:
• No DC component.
• Occupies less bandwidth than unipolar and polar RZ
schemes.
• Suitable for transmission over AC coupled lines.
• Possesses single error detection capability.
• Clock can be extracted by rectifying (a copy of) the
received signal.

• Disadvantages:
•Is not Transparent.

18
 Manchester
• Advantages:
• No DC component.
• Suitable for transmission over AC coupled lines.
• Easy to synchronise with.
• Is Transparent.

• Disadvantages:
• Because of the greater number of transitions it occupies a
significantly large bandwidth.
• Does not have error detection capability.

19
 Eye Patterns
• An eye pattern derives its name as it resembles a human eye. It
is the synchronized superposition of all possible realizations of
the signal of interest viewed within a signaling interval.
• To construct an eye pattern, we plot the received signal against
time on a fixed-interval axis, at the end of the fixed time interval,
wrap around to the beginning of the time axis. Thus the pattern
consists of many overlapping curves.
• An eye pattern provides the following information:
– Timing error allowed on the sampler is given by the horizontal
width of the eye. The preferred time for sampling is at the point
where the vertical opening of the eye is the largest.
– Sensitivity to the timing error is given by the slope of the open
eye (evaluated at, or near, the zero-crossing point).
– Noise margin of the system is given by the eye opening height.

20
 Eye Patterns (cont’d)
Best
sampling
Slope = sensitivity time Distortion at
to timing error sampling time

Noise margin

Time interval over which the Distortion of

received signal can be sampled zero-crossings
Figure 4.2 – Eye Pattern Interpretation
21
 Eye Patterns (cont’d)
• When a received line code waveform contains channel
impairments (such as distortion, ISI, and noise), traces
from the upper portion of the eye pattern cross traces
from the lower portion. This causes the eye opening to
become small.
• When the amount of channel impairments increases,
the opening becomes even smaller. This greatly
reduces the noise margin and results in more bit errors.
• Hence, eye patterns are used to assess the quality of a
received line code waveform. In addition, eye patterns
are a tool for evaluating the performance of a
baseband pulse transmission system.

22
 Eye Patterns (cont’d)
Polar NRZ Waveforms Eye Patterns (diagrams)
Binary bits : 1 0 1 0 0 1 Distortion on zero crossings

Amplitude
(V)
5V

Time (s) Time (s)

-5V
Noise
Minimal Distortion with ISI Margin
Tb
Amplitude
(V) 5V

Time (s) Time (s)

-5V
Distortion with ISI and noise corruption Noise
Margin
Distortion on
zero crossings
Figure 4.3 – Distorted Polar NRZ Waveforms and Corresponding Eye Patterns

23
 Bandwidth of Digital Signal
V0(t)
Observation:
Observation:pulse
pulsewith
withsmaller
smaller
rise/fall
rise/falltime
timeextends
extendsfurther
furtherout
outon
on
Ideal T = 10nsec the
thefrequency
frequencyaxisaxis(Sinc
(Sincresponse).
response).
 = 0 nsec

t
|V(f)|
8
V1(t) 1 10

8 10
9 V1(f)
T = 10nsec
 = 0.5nsec
V( i 2   f  1  To1 )
6 10
9 V0(f) (Infinite BW)
1

t V( i 2   f  2  To2 )
V2(t) 1 4 10
9 V2(f)

9
2 10
T = 10nsec
 = 3.0nsec
0
8 8 8 8 9
0 2 10 4 10 6 10 8 10 1 10
t f
i f

24
 System Filtering, P(f)

Pulse Shaping Pulse Shaping Data

Data Low Pass Filter TRANSMISSION Low Pass Filter out
in at transmitter CHANNEL at receiver
Xin(t) Xout(t)
HT(f) HR(f)
Channel (filter)
characteristics
HC(f)
Front-End Front-End
Transmitter Receiver The recovered pulse is rounded
Input flat-top pulses (to be sent to sampling and
System filtering P(f) decoding circuits)

Figure 4.4 – General Baseband Pulse Transmission System

• A baseband pulse transmission system consists of three low pass

filtering at the transmitter, channel, and receiver, respectively. The
combined effect is called system low pass filtering.
• When line code flat-top pulses are transmitted over the system, the
system filtering will cause the pulses to have rounded tops at
system output.
25
Effect of Bandlimiting the Pulse Spectrum

Low Pass
Filter

t t

 Causes “ringing” or ripples in time domain

 Then, pulse spread in time domain causes inter-symbol interference

26
 Inter-Symbol Interference
• If these flat-top pulses are filtered improperly by the
system, they will spread in time to form so-called tails
or ripples. These tails spread into adjacent bit intervals
to interfere with the interpretation of the adjacent
symbols. This interference is known as ISI (inter-
symbol interference).
• If an ISI causes a wrong interpretation of a symbol
(e.g., a pulse symbol is interpreted as binary 0 instead
of binary 1, and a zero-level symbol is interpreted as
binary 1 instead of binary 0), then a bit error arises. ISI
is a major source of bit errors in the receivers.

27
 Inter-Symbol Interference (cont’d)
Transmitted waveforms, Xin (t) Individual pulse responses Received waveforms, Xout (t)
(Unipolar NRZ)
Inter-symbol
Amplitude (V) Recovered bits : 0 1 1 1
interference
Threshold
0 1 0 1 voltage, Vt
5V (2.5 V)
2.5V

Os Time (s) t
Amplitude (V) Ts Ts Ts
Threshold
voltage, Vt
0 1 1 1
5V (2.5 V)

2.5V

Os Time (s)
Ts
Ts Ts

Recovered bits : 0 1 1 1
Sampling points Sampling points Sampling points
(transmitter clock) (receiver clock) (receiver clock)
Note : Vr > Vt => “1”
Figure 4.5 – Examples of ISI on the Received Pulse in a Binary Communication System and Vr < Vt => “0”
28
 Pulse Shaping: Ideal Nyquist Filtering
• Two approaches are used to control ISI: pulse shaping and
equalization.
• Pulse shaping is a process of using system filtering to change
the shape of transmitted pulses. Two pulse shaping or Nyquist
filtering techniques: ideal Nyquist and raised cosine.
• Harry Nyquist stated that it is possible to transmit independent
symbols through a baseband system at a symbol rate RS  2W
without ISI, where W is called minimum Nyquist bandwidth.
• This occurs when the system spectrum P(f) is made rectangular
with bandwidth W. The system time response p(t) is a sinc pulse.
• The received sinc pulses have been shaped in such a way that
their tails will cross the zero level at every sampling instants to
eliminate the ISI. This is an ideal Nyquist pulse shaping.

29
Pulse Shaping: Ideal Nyquist Filtering (cont’d)
P(t)
(a) System ideal spectrum (zero ISI) 1
P(f ) = X(f)  He(f) = 1/(2W)  rect[f / (2W)]
0.8
(b) Shaped sinc pulse at system output
(also called system time impulse response)

Impulse Response
0.6 Tb
p(t) = sinc(2Wt)
Tb  1 /( 2W ) 0.4

W  1 / 2Tb  Rb / 2 0.2
= minimum Nyquist bandwidth
0

|P(f)| -0.2

-0.4
Tb  1 /( 2W ) -6 -4 -2 0 2 4 6
Normalised Time (t/Tb)

Sampling Instant
f Tb
0
–W = –1/2Tb W = 1/2Tb
Figure 4.6 – Criteria for Zero ISI
30
 Pulse Shaping: Ideal Nyquist Filtering (cont.)
5
1
0 Pulse Response
-5 Nyquist Brickwall Filter
0.8
-10

Impulse Response
0.6
Response (dB)

-15
-20 0.4
-25
-30
0.2

-35 0
-40
-45 -0.2

-50
0 0.5 1 1.5 2 2.5 3 3.5 4 -0.4
-6 -4 -2 0 2 4 6
Normalised Frequency (f/Rb)
Normalised Time (t/Tb)

Figure 4.6(b) – Criteria for zero ISI

 Nyquist filter - achieves zero crossings at integer multiples of symbol period

 Zero crossings at symbol interval - no ISI at sample point
31
Pulse Shaping: Ideal Nyquist Filtering (cont’d)
Binary bits : 0 1 1 0 1 0 1 0

0.8

0.6

0.4
Voltage (V)

0.2

-0.2

-0.4

-0.6

-0.8

-1

-5 -4 -3 -2 -1 0 1 2 3 4 5
Normalised Time (t/Ts)
 Filtered pulses with their tails crossing zero level at all sampling point
Figure 4.7 – Zero ISI Transmission
32
Pulse Shaping: Raised Cosine Filtering
• Two problems in ideal Nyquist filtering:
– The rectangular spectrum is not physically realizable.
– A small sampling instant error (i.e., a sampling instant is slightly shifted)
will result in ISI.
• To solve these two problems, raised cosine filtering is
recommended. A sinusoidal (raised cosine) roll-off is introduced to
the system spectrum. The degree of roll-off is determined by a roll-
off factor r, which varies from 0 to 1. When r = 0, we have the
rectangular spectrum with bandwidth W. At r = 1, the system
spectrum has the most gradual roll-offs with bandwidth 2W. System
spectra with sinusoidal roll-offs are physically realizable.
• As r increases from 0 to 1, the transmitted sinc pulses are shaped to
have their tail amplitudes reduced. At r = 1, the tails have additional
zero-crossings. Smaller tail amplitudes and additional zero-
crossings will help reduce ISI, which is caused by sampling instant
errors.

33
 Pulse Shaping: Raised Cosine Filtering (cont’d)
• The raised cosine Nyquist filter has the transfer function :
1, f  f1

 1    ( f  f1   B is the absolute bandwidth
Pcos ine ( f )   1  cos   , f1  f  B
2 2 f f0 is the 6 dB bandwidth of the filter
     
 r is the roll-off factor
0, f B

f = B – f0 f f
|Pcosine (f)|
f1 = f0 – f 1

r = f / f0
0.5

f
B  f0  f1 f1 f0 B
Figure 4.8 – Raised Cosine Filter Characteristics Source : Ref [1]
34
 Pulse Shaping: Raised Cosine Filtering (cont.)
1 ralpha=0.0
=0
1 ralpha=0.5
= 0.5
ralpha=0.0
=0
ralpha=1.0
= 1.0
ralpha=0.5
= 0.5
ralpha=1.0 0.8

Norm Impulse Response

= 1.0
Norm Freq Response

0.8
0.6

0.6 0.4

0.2
0.4

0.2
-0.2

0 -0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 -4 -3 -2 -1 0 1 2 3 4
Norm. Freq (f/Rb) Norm. Time (t/Tb)
Figure 4.8(b) – Raised cosine filter characteristics

 Smaller r, less bandwidth but higher sidelobe levels

 r=0.2 to 0.5 typically used currently
 Bandwidth is given as (1  r )R S where R is symbol rate
BW  s
2 35
 Raised Cosine Filters and Root Raised
Cosine Filters

Receiver
Transmitter
Ÿ Filter noise
Ÿ Spectral containment
Ÿ Max SNR @ decision pt

Root Raised Root Raised

Data DEMOD
Cosine Cosine

Root Raised Cosine filters at TX & RX form overall Raised Cosine response

36
 FIR Implementation
Impulse
T T T T T T T
Data Train
k0 k1 k2 k3 k4 k5 k6 k7

1
FIR Response

0.5

0
Pulse Shaped
Data Stream
-4 -3 -2 -1 0 1 2 3 4
Time (t/Tb)

1
FIR Response

0.5

0 5 10 15 20 25 30
Tap Number
37
 Review Question
What is the minimum required bandwidth of a baseband
channel if we need to send 1 Mbps by using baseband
transmission?
Ans
(1  r )R S
From BW  , the minimum bandwidth is when r=0
2

Therefore, minimum bandwidth to send 1Mbps, is given as:

B = Rs/2 = 1 Mbps/2 = 500 kHz.

?
38
 Probability of Error
2 2

1.5 1.5

1 1

0.5 0.5

0 0

-0.5 -0.5

-1 -1

-1.5 -1.5

-2 -2
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Sampling Instants Sampling Instants

• For a simple threshold detection system:

• Erroneous decision made when noise causes signal to cross threshold
• Probability of error related to signal-to-noise ratio
39
 Probability of Error (cont.)
0
10

Assumes AGWN
Perfect sampling -2

Probability of Error, Pe
10

Bipolar signal (+1,-1)

-4

Typical operating points: 10

• Wireless voice10-3
• Wireless data 10-6 10
-6

• Lightwave <10-9 to 10-12

-8
10
0 2 4 6 8 10 12 14 16 18 20
SNR (dB)

Figure 4.9 – Probability of Error

40
 Matched Filter
• Filter at RX to reduce noise Data Signal

•Choose filter response to be mirror

image of incoming signal

•Filter is ‘matched’ e.g. RRCF pair Filter

h(t)

• Signal output from filter: E Symbol

Timing

• Noise output from filter: NO Noise

•Output SNR depends on signal

energy (E) and noise power spectral
density (NO)
41
 Probability of Error vs. Eb/No
0
10
 Performance of bipolar
NRZ in AWGN
-2
10

Probability of Error, Pe
 Eb/No - measure of power
efficiency 10
-4

Typical implementation: -6
10
RRCF pair
-8
10
0 2 4 6 8 10 12 14 16
Eb/No (dB)
Figure 4.10 – Probability of Error vs Eb/No

42
 Increasing Number of Symbol States
 Represent groups of bits as 1

Binary Signal
symbols 0.8
0.6

 Increase bandwidth efficiency by 0.4

reducing symbol rate

0.2
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
 e.g. Two bits encoded into one Time (bit periods)
symbol
2

4-level Signal
1

 Symbol period: TS=2TbRS=0.5Rb 0

 Trade-off against noise immunity -1

-2
0 1 2 3 4 5 6 7
Time (symbol periods)
Binary Input Output
(l = 2 bits) Level (V)
Binary signal L – level
11 + 1.5 multilevel signal
10 + 0.5
w1(t) or w2(t) k – Bit digital-
00 - 0.5 to-analog
01 - 1.5 D symbols/sec = R bits/sec converter D symbols/sec = R/l

L = 2k and R bits/sec
43
 Example
• Modem uses 32 phases to transmit 5 bits of information
• Phase change every 0.5 ms (symbol period TS)

• Symbol rate: RS=1/TS=2000 symbols/sec

• Channel capacity: 5xRS=10,000 bits/sec

• Do not confuse symbol (baud) rate with information (bit)

rate!

44
 Bit Rate Vs Baud/Symbol Rate
• Bit rate is the number of bits per second. Baud or Symbol
rate is the number of signal elements per second.

• In the analog transmission of digital data, the baud or

symbol rate is less than or equal to the bit rate.

• The term “baud” originates from the French engineer Emile

Baudot, who invented the 5-bit teletype code.

• Baud rate refers to the number of signal or symbol changes

that occur per second. A symbol is one of several voltage,
frequency, or phase changes.

45
 Limitation on Symbol States
• Bandwidth efficiency achieved by increasing symbol states
• Limit set by POWER EFFICIENCY
• n bits encoded into M=2n symbol states

Example:
• Information rate: 2 Mbits/s
• Channel bandwidth: 75 kHz
• Symbol rate: 150 ksymbol/s
• Bits/sym required: (2x106/150x103)=14 bits/symbol
• No. of symbol states: M=214=16,384 states
• Assume symbols correctly identified if separated by 5 mV
• Pk-pk voltage of waveform: 82 V
46
 Power-Bandwidth Efficiency Trade-off

Increase symbol states to

increase bandwidth efficiency
101.0
Harder to distinguish symbols in 8
7

Bandwidth Efficiency Rb/B

6
5
presence of noise 4
3

Hence reduced power efficiency Case: Rb<C

100.0
8
7
6 Practical radios operate here
Power/BW efficiency trade-off 5
4 i.e. theoretical capacity always better
than what can be achieved practically
C  B log 21  SNR 
3

C  Eb C  10-1.0

 log 21  
-6.0 0.0 6.0 12.0 18.0 24.0 30.0 36.0
shannon curve Apr. 11, 2001 Power Efficiency Eb/NO (dB)
B  NO B 

47
 Review Question
A computer has an output bit rate of 56 kbps, and three
successive binary digits are coded into one of eight possible
amplitude levels. The resulting eight-level pulses are then
shaped using raised cosine filtering during transmission over a
baseband channel. Find the transmission bandwidths for the
following roll-off factors: r = 0.25, 0.5, 0.75 and 1.

?
48
 Review Question
Ans
Bit rate RS=Rb = 56 kbps, number of bits per symbol k = 3,
8-level signalling
RS
Transmission bandwidth BT = (1+ r)
2k
Roll-off Factor Transmission
Bandwidth
r = 0.25 BT = 11.67 kHz
r = 0.5 BT = 14 kHz
r = 0.75 BT = 16.4 kHz
r=1 BT = 18.7 kHz 49
 Review Question
An analogue source is sampled, quantised, and encoded into
a binary PCM signal. Each quantised sample is encoded into
a code-word, which is then line coded into three information
pulses plus a synchronising pulse. The line code pulses can
take on four possible levels, and are transmitted over a
channel of bandwidth 6 kHz using RRC filtering r=0.5.

1.Determine the symbol rate of the line code pulses.

2.Find the information bit rate of the PCM signal.

?
50
 Review Question
Ans
one sample 3 info pulses 1 sync pulse

6-bit codeword 2 bits 2 bits 2 bits 2 bits

No. of signalling levels M = 4 = 2k, No. of bits per symbol k = 2

Channel bandwidth = transmission bandwidth BT = 6 kHz
(1) RS (1+ r)
BT =B =
2
RS (1+ 0.5)
6, 000 =
2
Symbol rate RS =8, 000 symbols/s
51
 Review Question
Ans

(2)
Only ¾ symbols carry information out of 8,000 symbols in one
second
Info symbol rate = ¾ x 8,000 = 6,000 info symbols/s
Info bit rate = info symbol rate x k = 6,000 x 2 = 12,000 bps

52
 Advantages of Digital Modulation
• Spectral efficiency – use of a narrow bandwidth to
send a large amount of data
– Effective use of limited frequency resources

• Good privacy and security features

– Digital encryption techniques may be employed

• Lower power consumption

• Repeatable, more easily produced

• Reduced device size

53
 Modulation for Wireless
• Media
– Carrier

V(t) = A cos(2πfc t + Φ)

• The three essential parameters:

– Amplitude value A(t) ― Amplitude Modulation
– Frequency value f(t) ― Frequency Modulation
– Phase value (t) ― Phase Modulation

54
Analog Modulation Application Examples
• Amplitude Modulation
– AM radio

• Frequency Modulation
– FM radio, TV audio signal

• Phase Modulation
– TV color image signal (including amplitude modulation)

55
 Transmission of a Digital Message
• Basically, it is the same as the analog modulation methods.

– ASK: Amplitude shift keying

V(t) = A(t) cos(2πfc t + Φ)

– FSK: Frequency shift keying

V(t) = A(t) cos(2πf(t) t + Φ)

– PSK: Phase shift keying

V(t) = A(t) cos(2πf(t) t + Φ(t))

• Digital modulation: amplitude, frequency, and/or phase are used

to represent a digital state.

56
Transmission over Bandpass Channels
• In passband transmission, the baseband message
signal is modulated onto a high-frequency sinusoidal
carrier. The modulated signal is then transmitted over
a bandpass channel.
• The major concern is the optimum design of the
receiver so as to minimize the probability of symbol
error in the presence of channel noise.
• AM, FM, and PM are analog bandpass modulations.
ASK, FSK, and PSK are digital bandpass
modulations.
• Examples of bandpass channels are telephone lines,
microwave radio links, and satelite paths.

57
 Definition of Bandpass
• A bandpass, is the term is often used in contrast to
baseband, which refers to a signal which has been
translated to a higher frequency or modulation.
• It is often for transmission by Frequency Division
Multiplexing (FDM).
• It can often be considered a synonym to passband,
carrier-modulated or radio frequency.

BW

fRF
58
 Bandpass Modulation
|F(f )| bandpass
fc(t) bandpass spectrum
m(t) baseband spectrum modulation

f
–fc fc, carrier frequency
modulating baseband signal m(t) bandpass signal fc(t)

~
f(t) r(t) f(t)
~
m(t)
Baseband Transmission Baseband
Signal Modulator medium Demodulator Signal
processing (channel) processing

Carrier Carrier

bandpass modulation bandpass demodulation

Figure 4.11 – Bandpass Communication System

59
Digital Modulation - Polar Display: Magnitude &
Phase Represented Together

ag
M
Phase
0 deg
(carrier phase reference)


Magnitude is an absolute value

Phase is relative to a reference signal
60
Digital Modulation - Signal Changes

Ma stant
t ude
ag
M

gn i
Con
Constant Phase
0 deg Phase
0 deg
(Reference)
AM-Magnitude Change PM-Phase Change

0 deg
0 deg

AM & PM -Both Change FM-Frequency Change

61
 Polar vs IQ Format
"Q"

Q-Value
{ 0 deg

{
"I"
Project signal
to "I" and "Q" axes
I-Value

Polar to Rectangular Conversion

62
 IQ Representation in Modulated Carrier

V (t )  A(t ) cos[2f c t   (t )]  A(t ) cos (t ) cos 2f c t  A(t ) sin  (t ) sin 2f c t

I (t) Complex BB signal Q (t)

A(t )  I (t ) 2  Q(t ) 2  (t )  tan 1[Q(t ) / I (t )]

I component of V(t) I component of BB

Q
V(t)

Q (t) A (t)
 (t )

Q component of BB I (t)
I
Q component of V(t)

63
 Constellation / IQ Diagram

carrier
(reference)
Q - Quadrature (90O)

01
I - In-phase (0O)
10 00 modulated
11 signal

Figure 4.12 – IQ Constellations

64
 Amplitude Shift Keying (ASK)
• In ASK, baseband pulses f(t) amplitude-modulate a carrier of Acos(ct).
• A binary 1 symbol (e.g., a +ve pulse) turns on the carrier and a binary 0
symbol (e.g., a zero level) turns off the carrier. Thus, ASK is also called
On Off Keying (OOK).
• Could be accomplished with an AM system.
• ASK signal: f c (t )  b(t ) A cos(ωct )
Binary Bit b(t) 0 1 0 1 0 1 1

539 x 295 pixels, 28kB

Baseband pulses f(t)
Tb Time (s)

0 1 0 1 0 1 1

Modulated ASK
signal fc(t)

Time (s)
Figure 4.13 – ASK Signal Tb 65
 Amplitude Shift Keying (ASK) (cont’d)
• Multiplication of pulses f(t) and carrier cos(ct + ) will shift the f(t)
spectrum up to carrier frequency c:
1
f (t ) cos(ωct  θ)  [e jθ F(ω  ωc )  e  jθ F(ω  ωc )]
2

|F(f)|
Pulse baseband A
spectrum
f
–B +B
O
|FC(f)|
ASK DSB-SC spectrum
Lower Upper
sideband sideband

½A ½A

f
– fC – B – fC – fC + B O fC – B fC fC + B

Figure 4.13b – ASK Spectrum Tx bw of ASK : BT  2 B BT

66
where B is baseband bw
 ASK in IQ Domain
• ASK(OOK)
– I: In-phase component
– Q: Quadrature component
I
Q
0
I
(0) (1) time
Q

Amplitude variation on the I axis 0

time
Figure 4.14 – IQ Domain for ASK
67
 Review Question
Find the transmission bandwidth for an ASK signal with
pulse shaping, r = 0.5, as shown below:

?
Amplitude/volt

0 1 0 1 0 0 1

Time/s

1 second
Ans
Bit Rate = Symbol Rate = 7

Tx BW of ASK  BT  2 B  (1  r ) RS  (1  0.5)7  10.5 Hz

68
 Frequency Shift Keying (FSK)
• In FSK, baseband pulses f(t) frequency-modulate a carrier of
Acos(ct).
• A binary 1 symbol (e.g., a +ve pulse) produces a frequency f1 =
fc + f, and a binary 0 symbol (e.g., a zero level) produces a
frequency f2 = fc - f . f is the frequency deviation about fc.
• Could be done with a simple FM system.
• FSK signal: f c (t )  A cos[(ωc  ω)t ]

5V
f2 1 0 1 0 0

f1
Time (s)

–5V
Tb Tb
Figure 4.15 – FSK Signal 69
 Frequency Shift Keying (FSK) (cont’d)
1 0 1 0 0

Time (s)
f2 = cos (ωC + Δω)t f1 = cos (ωC - Δω)t

|FC(f)|
f1 FSK spectrum
f2
fC

f
B Δf Δf B
fC – Δf fC + Δf
fC – Δf – B fC + Δf + B
2Δf + 2B

Figure 4.16 – Time Domain and Spectrum for the FSK Signal

Modulation index of FSK: f

β where, B is the baseband bandwidth
B
Transmission bandwidth of the FSK signal: BT  2f  2 B  2 B(1  β)
70
 FSK in IQ
• A frequency change causes a
constant-rate phase change vs. the
reference carrier.
• The amplitude remains constant on
the IQ circle.
Phase
1 Symbol only turns π(ex)

Q π

Time
1Sp 2Sp 3Sp 4Sp 5Sp
I

• If the phase change is 90 －π

degrees in one symbol period, the
modulation type is called Minimum Pos. offset Neg. offset Pos. offset
Shift Keying (remember this one).
Figure 4.17 – FSK in IQ 71
 Review Question
An FSK modulated signal with pulse shaping r = 1, has a
bandwidth of 100 kHz which spans from 300 kHz to 400 kHz.
What is the carrier frequency and the bit rate of this
transmission, if the frequency deviation chosen is 25 kHz.

Ans
The carrier frequency should be center of the span
frequencies (400+300)/2 = 350 KHz. If 2f = 50 kHz,
this means B  2f  2 B  100 kHz
T

2f  RS (1  r )  100 kHz

50  2 RS  100  RS  25 kbps
?
72
 Phase Shift Keying (PSK)
• In PSK, baseband pulses f(t) phase-modulate a carrier of
Acos(ct).
• A binary 0 symbol (e.g., a +ve pulse) produces +Acos(ct), and
a binary 1 symbol (e.g., a zero level) produces –Acos(ct).
• PSK signal: f c (t )   A cos(ωct )

• The PSK signal has the same double-sideband characteristic as

ASK, hence the transmission bandwidth of the PSK signal: B  2 B
T
PSK signal fc(t)

5V 1 0 1 0 0

Time (s)

–5V
Tb
Figure 4.18 – PSK Signal
73
 Representation of PSK in IQ
• PSK
– Specifically, Binary Phase Shift Keying (BPSK)
I

Q
0
(1) I
(0)

Change phase to 180°

0
relative to reference

Figure 4.19 – IQ Domain for the PSK Signal

74
 Review Question
Find the transmission bandwidth for an PSK signal with
pulse shaping, r = 0.5, as shown below:

Amplitude

0 1 0 1 1 0
?
Time

1 second
Ans
Bit Rate = Symbol Rate = 6

Tx BW of PSK  BT  2 B  (1  r ) RS  (1  0.5)6  9 Hz
75
Transmission over Non-Linear Channels
Binary Code 0 1 0 0 1 1
(NRZ) t

B-ASK t

B-FSK t

B-PSK t

Figure 4.20 – Digital Carrier Modulation

• PSK and FSK signals have a constant envelope. This feature makes them impervious to
amplitude non-linearity, which is commonly encountered in microwave and satelite channels.
• Therefore in practice, PSK and FSK signals are preferred to ASK signals for transmission
over non-linear channels.

76
Binary Modulation Schemes (ASK/FSK/PSK)
 Binary modulation schemes offer a theoretical best case bandwidth
efficiency of 1 b/s/Hz, (10 kb/s need at least 10 kHz BW).
 For modern wireless communication systems required high data
rates in congested radio spectrum, these schemes do not have
sufficient bandwidth efficiency.
 PSK and FSK have a constant envelope which makes them
impervious to amplitude non-linearity (commonly encountered in
microwave and satellite channels).

77
 M-ary Modulation
• In baseband systems, a symbol is a pulse or a zero level. In
bandpass systems, a symbol is a modulated carrier or a zero level.
• M-ary (e.g., 4-ary) modulation is also called multilevel signaling,
where successive bits are combined to form a longer symbol.
• No. of signaling levels M = 2k, where k is the no. of bits per
symbol.
• Symbol or baud rate RS = Rb / k, where Rb is the bit rate.
• Symbol period TS = 1 / RS and bit period Tb = 1 / Rb.
• Multilevel signaling results in a smaller transmission bandwidth:
– BT = B / k for baseband transmission
– BT = 2B / k for bandpass transmission
– 6 dB bandwidth of the bandpass signal = (2B / k)(1/(1+r))
– B = (Rb / 2)(1+r), where r is the roll-off factor for pulse shaping

78
 M-ary Modulation (cont’d)
L = 2l, level
Binary Input Output
(l = 2 bits) Level (V) Binary data/bit multilevel signal
11 + 1.5 11101100010011 L = 4, multilevel signal 1.5V, 0.5V, 1.5V, -0.5V, …..
10 + 0.5 converter (convert bit
00 - 0.5 1 bit per symbol : to analog amplitude) 2 bit per symbol :
01 - 1.5
D symbols/sec = R bits/sec D symbols/sec = R/l
bit stream: 1 1 1 0 1 1 0 0 0 1 0 0 1 1 and R bits/sec
1
Binary Signal

0.8
0.6
0.4  Represent groups of bits as symbols.
0.2
0
 Increase bandwidth efficiency by
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 reducing symbol rate
Time (bit periods)
2  e.g. Two bits encoded into one symbol.
4-level Signal

1  Symbol period: TS = 2Tb  RS = 0.5Rb

Tb
0  Trade-off against noise immunity.
-1
Ts
T0
-2
0 1 2 3 4 5 6 7
Time (symbol periods)
Rectangular Pulse Shape, Tb = 1 ms 4 levels = 2 bits per symbol
79
Figure 4.21 – M-ary Modulation
 M-ary Modulation (cont’d)
• Multi-symbol systems include multi-phase, multi-amplitude, and
combined multi-phase/multi-amplitude signaling schemes.
• These are commonly used in telephone, microwave, and satelite
data communications to achieve higher spectrum efficiency.
• Two successive binary pulses are combined to produce four
symbols of 00, 01, 10, 11. The symbols modulate a carrier into
four phases i. The ith modulated signal of i is:
T T
si (t )  cos(ωct  θ i ), i  1, 2, 3 or 4,  t 
2 2
• Two possible choices for the four phase angles are:

π π 3π π i
θ i  0,  , π or  ,  (phases are apart)
2 4 4 2

80
 M-ary Modulation (cont’d)
• Signals of this type are called quaternary/quadrature PSK
(QPSK).
• A QPSK signal can be represented in a 2-dimensional
constellation diagram. The horizontal axis is called In-Phase,
and the vertical axis is named Quadrature. The points represent
the four phases i which correspond to four symbols.
Quadrature Quadrature
“01” “01” “00”

“10” “00”
In-Phase In-Phase

“11” “11” “10”

Figure 4.22 – Two Different QPSK Signal Constellations

81
 M-ary Modulation (cont’d)
I channel
I Ch. sig
Sym rate R S = R b /2

P.Shape
Filter Rb/2 symbols/s
01 00
(Binary input) Carrie r cos ωCt
Data Serial-
Bit rate R b
Parallel
90 o

sin ωCt 11 10
0 0 0 P.Shape
1
Time Filter

1/Rb T = 2/Rb Q channel

Sym rate R S
= R b /2 T = 2/Rb T = 2/Rb
Q Ch. sig
Figure 4.23 – M-ary Modulation (QPSK) Generation

 Twice bandwidth efficiency of PSK 00 01

cos (ωCt+45o) cos (ωCt+135o)
 2 bits grouped per symbol defining 4 phase states
 Often Gray-coded
82
 M-ary Modulation (cont’d)
• Other PSK signals are 2-phase/binary PSK (BPSK), 8-PSK, 16-PSK,
etc.
• No. of signaling levels M = 2k, where k is the no. of bits per symbol.
 2 (t )
“0” “1”  2 (t )
s1 s2 s3 “011”
BPSK (M=2) “010” “001”
 Eb Eb  1 (t ) s4 s2
Es
 2 (t )
“01” “00” “110” s“000”
1
s2 s1 s5  1 (t )

Es
“111” “100”
s6 s8
 1 (t ) QPSK (M=4)
“101” s7
s3 “11” “10”
s4 8-PSK (M=8)
Figure 4.24 – Constellation Diagram
83
Quadrature Amplitude Modulation (QAM)
• A QAM signal contains multiple phases and amplitudes:
si (t )  ri cos(ωc t  θ i )
• 16-state/level/symbol QAM is interpreted as having 4-level amplitude
modulation applied on each of two quadrature carriers.

“0000” “0001” “0011” “0010”

s1 s2 s3 s4

“1000” “1001” “1011” “1010” A 16-QAM

constellation

“1100” “1101” “1111” “1110”

“0100” “0101” “0111” “0110”

Figure 4.25 – Constellation Diagram for QAM

84
 Quadrature Amplitude Modulation (QAM)

• QAM is combining ASK and PSK in such a way that

have maximum contrast between each bit.
• Implementation is by modulation of a carrier signal in
both amplitude and phase
• Each unique combination of amplitude and phase is a
“symbol”
• Normally the number of amplitude shifts is fewer than
the number of phase shifts.
• Because amplitude changes are susceptible to noise
and require greater shift differences than do phase
changes

85
 Digital Modulation Block Diagram
Analog
Analog-Digital Converter
signal (sampling, quantizing)

Encoding Compression, 011010100101 Symbol Mapping 01 10 10 10 01 01

Raw Binary Error Correction, [Convert to
Data 110101 Encryption Symbols]

01 00
I - Signal I - Signal
Low Pass Filter

Q - Signal IQ Modulation
Q - Signal
10 11 Low Pass Filter

Modulation Mapping Baseband signal

RF signal
Up-conversion Filtering, RF Power
band shaping Amplifier

Figure 4.26 – Generation of Digital Modulation 86

 Analog-Digital Converter (Source Coding)
Waveform Coder

Analog
Analog-Digital Converter
signal (sampling, quantizing)

Encoding
...1101011101...
 Analog speech is sampled and then transformed in a digital stream.
 The higher the SNR, the better the speech quality perceived by a listener.
 Example : PCM (64 kbps) and Adaptive PCM (32 kbps).
 Source Coding (Waveform Coder/ Vocoders)
– Change analog input signal to an appropriate digital stream.
– Not dependent on transmission channel characteristics.
– Reduce natural redundancy present in human speech aw data comes
from the user.

87
 Raw Data Conversion
Analog
Analog-Digital Converter
signal (sampling, quantizing)

Compression,
Encoding 011010100101 Symbol Mapping
Error Correction, [Convert to
Raw Binary Interleaving,
Data Symbols]
110101 Encryption

• Raw data comes from the user

– Digitized voice, keystrokes, jpegs…

• Compression is employed for efficiency

• Error correction is applied for transmission quality

• Interleaving creates signal-dropout resistance

• Encryption is applied for security

88
 Data Bits to Symbols

Compression, 011010100101 Symbol Mapping 01 10 10 10 01 01

Error Correction, [Convert to
Encryption Symbols]

• Symbols are represented by the possible states of digital

modulation.
• Higher order modulation allows more bits per symbol.
• What does that mean?
– Mapping symbols to I and Q.

bit rate
Symbol rate =
the number of bits transmitted with each symbol

89
 IQ Mapping
• What is Mapping: • Example
– Translate a Symbol to a
Q
point in the IQ space
(01) (11)
01 00

10 11 I
Modulation Mapping

(00) (10)

Figure 4.27 – IQ Mapping

90
 Error Vector Magnitude (EVM)

Figure 4.28 – EVM

91
 Differential Modulation
• QPSK(Quadrature PSK) • DQPSK(Differential QPSK)
– Assign the value to – The value is based on the
points in the IQ space transitions between two points

Q Q
(01)
(01) (11) (00)

I I

(11) (10)
(00) (10)

00 = 0° 01 = −90°
Figure 4.30 – Differential Modulation
10 = +90° 11 = +180°
93
 Differential Modulation (cont’d)
QPSK p/4DQPSK

Both formats are two bits/symbol.

The differential format depends on transitions only, and only four of
the eight possible transitions are allowed.
Figure 4.31 – Constellation for QPSK and DQPSK

94
 Quadrature PSK (QPSK)
• Modulated signals shift between four phases.

• Each phase is 90 degree apart.

• Thus, each symbol has two bits.

• Uses two basic functions: sine and cosine.

• Hardware realization possible carrier signal

representing each symbol is created by the free
running of sine and cosine waves.

95
 More on QPSK
• Having two bits/symbol means the bit stream must be
broken into two separate streams.

• Signal expression:
s(t) = I(t)Cos(2π ft) + Q(t)Sin(2π ft)

96
 More on QPSK (cont’d)
• E.g., Consider the following bit streams [00111000] polar NRZ:

Amplitude (V)

+5

d2 d3 d4

Time (s)
d5 d6 d7
d0 d1

–5

0 T 2T 3T 4T 5T 6T 7T 8T

97
 More on QPSK (cont’d)
• Splitting the data into two separate bit streams:
I(t)

+5
d2 d4
Time (s)

d0 d6

–5
0 2T 4T 6T 8T

Q(t)

+5
d3
Time (s)

d5 d7
d1
–5
0 2T 4T 6T 8T
98
 More on QPSK (cont’d)
• QPSK signal:

d0 = -5V d2 = +5V d4 = +5V d6 = -5V

d1 = -5V d3 = +5V d5 = -5V d7 = -5V

S(t) Time (s)

0 2T 4T 6T 8T
00 11 10 00

If both bit streams change, a 180-deg phase transition is observed.

If either the I or Q channel changes, a 90-deg phase transition is
observed.

99
 QPSK Generation
I channel
I Ch. sig
Sym rate R S = R b /2
P.Shape
Filter
01 00
Carrier
Data Serial-
Bit rate R b
Parallel
90 o

11 10
P.Shape
Filter

Q channel
Sym rate R S
= R b /2
Q Ch. sig
Figure 4.32 – Generation of QPSK signal

 Twice bandwidth efficiency of PSK

 2 bits grouped per symbol defining 4 phase states
 Often Gray-coded
100
 QPSK Implementation

P.Shape Recons
DAC
FIR Filter Filter

Symbol RF LO
Data Encode 90 o
PA

P.Shape Recons
DAC
FIR Filter Filter

Baseband

I I
Symbol P.Shape Recons
Data Encode FIR Filter
DAC
Filter
PA
Q Q

NCO
RF LO
Baseband

Figure 4.33 – Implementation of QPSK

101
 QPSK Eye Diagrams

Raised cosine r=1.0 Raised cosine r=0.22

• Eye closure as a result of filtering
• Smaller alpha requires higher tolerance on clock circuits
• Results above use noiseless signal

102
 QPSK Demodulation

Matched Decision
Filter Circuit

Carrier
Recovery
RX Signal Sym Time Data Recovered
(RF/IF) Recovery Combine Data Stream
90 o

Matched Decision
Filter Circuit

 Consider as two BPSK demodulators operating in phase quadrature

 Require carrier recovery (coherent demodulation)

Figure 4.34 – QPSK Demodulation

103
 QPSK Carrier Recovery
Recovered
QPSK requires coherent demod IF Carrier

•No discrete component at carrier

•Require carrier recovery LPF

Modulated Phase
IF Signal Detector

90 O VCO

Figure 4.35 – QPSK Carrier Recovery LPF

104
QPSK Measurement Parameters

105
 Differential Encoding/Decoding
Output
XOR
Signal 11010010101001...
0 10011100110001
Output (del) 0 1001110011000...
XOR 10011100110001...
T b

Encoder
Coded Decoded
10011100110001
Signal Signal
XOR 11010010101001

T b
Signal 10011100110001
Signal (del) 0 10011100110001
Decoder XOR 11010010101001

Encode data based on change of state - does not matter if data inverted
• Input ‘1’ - change state of output bit
• Input ‘0’ - leave output state unchanged
106
 Differential Phase Shift Keying (DPSK)
Output
0 10011100110001
PSK signal phase
XOR    
Signal 11010010101001...
Output (del) 0 1001110011000...
XOR 10011100110001...
T b

Encoder Decoded
Signal
10011100110001 11010010101001
Coded Signal
    Filter

T b Signal   

Signal (del)    
Decoder Mixer OP 

Differentially encode data

Demodulate data using previous symbol - no need for LO!
107
 IQ Offset Modulation
Eye Constellation

Q
time
QPSK
I
time

Q
Offset time
QPSK
I
time

Offset QPSK minimizes power variations between symbols, since the signal
trajectory (vector diagram) does not go through zero (the origin).
Figure 4.36 – Constellation and Eye Diagram for QPSK and Offset QPSK
108
 Offset QPSK (OQPSK)
• Q bit streams is delayed by ½ symbol time so that I
and Q do not transit at the same time.

• Useful in high power application (e.g., satelite

communications).

• Phase transitions between symbols are limited to 90°.

109
 Offset QPSK (OQPSK) (cont’d)
Delaying the Q channel:
I(t)

+5

d2 d4
Time (s)

d0 d6
–5
–T 0 T 3T 5T 7T

Q(t)

+5
d3

Time (s)

d1 d5 d7

–5
0 2T 4T 6T 8T

110
 Offset QPSK (OQPSK) (cont’d)
OQPSK waveform

d0 = – 5V d2 = +5V d4 = +5V d6 = –5V

d1 = – 5V d3 = +5V d5 = –5V d7 = –5V

S(t)

0 T 2T 3T 4T 5T 6T 7T

Phase transition is limited to 90°.

111
 Offset QPSK (OQPSK) (cont’d)
OQPSK modulator:
I channel
I Ch. sig
Sym rate R S = R b /2

P.Shape
Filter
01 00
Carrier
Data Serial-
Bit rate R b
Parallel
90 o

11 10
P.Shape No 180 o phase
T b Filter transitions
Q channel
Sym rate R S
= R b /2
Q Ch. sig

Figure 4.37 – Offset QPSK Modulator

112
 Offset QPSK (OQPSK) (cont’d)
• I and Q channels do not transit at the same time.
(Limiting the phase transition to 90 degrees.)

• Unlike QPSK, the IQ transitions do not go through 0

in the constellation diagram.

• OQPSK is more constant-envelope than QPSK.

• Essential if the transmitter is operating close to the

saturating point of the amplifier.

113
 Offset QPSK (OQPSK) (cont’d)
Constellation diagram comparison between QPSK and OQPSK:

114
Offset QPSK (OQPSK) (cont’d)

ms

ms

115
 Constant-Envelope Modulation
QPSK MSK (GSM)

Amplitude (envelope) varies Amplitude (envelope) does

from zero to a nominal value not vary at all

116
 Higher Order Modulation
• 8PSK(8-PSK) • p/4 DQPSK
– Assign the value to – The value is based on the
points in the IQ space transitions between two points
– Three points per symbol – Eliminates zero crossings

Q Q
(110) (11)
(001) (011)

(01) (00)
(111)
I I
(100)
(10)
(000)
(010)
(101)

00 = +45° 01 = −135°
10 = +135°11 = −45°
117
 More on Higher Order Modulation
• 16QAM(16-Quadrature Amplitude Q
Modulation)
– Each IQ symbol location is
(000100) (001100) (011100) (010100) (110100) (111100)(101100) (100100)
represented by four data bits
• 64QAM (64-Quadrature Amplitude
Modulation) (000101) (001101) (011101) (010101) (110101) (111101)(101101) (100101)
– Each symbol is now worth five bits

Q (000111) (001111) (011111) (010111) (110111) (111111)(101111) (100111)

(000110) (001110) (011110) (010110) (110110) (111110)(101110) (100110) I

(0010) (0110) (1110) (1010)

(000010) (001010) (011010) (010010) (110010) (111010)(101010) (100010)

(0011) (0111) (1111) (1011)
I
(000011) (001011) (011011) (010011) (110011) (111011)(101011) (100011)
(0001) (0101) (1101) (1001)
(000001) (001001) (011001) (010001) (110001) (111001)(101001) (100001)

(0000) (0100) (1100) (1000)

(000000) (001000) (011000) (010000) (110000) (111000)(101000) (100000)

118
 Why Not Just Keep Going?
• Errors in IQ modulation create symbol errors in
transmission
• Vector Errors are created (what’s that?) Q
• Noise in the transmission channel creates symbol errors
• Inaccuracies in the receiver create errors
• Signal-to-noise requirements increase with higher order
modulations
(0010) (0110) (1110) (1010)

Q
(0011) (0111) (1111) (1011)
I
(01) (11)
(0001) (0101) (1101) (1001)

I
(0000) (0100) (1100) (1000)

(00) (10)

119
Power Efficiencies for M-PSK
M-PSK & M-QAM and M-QAM
Power Efficiencies
Probability of Symbol Error, PS 10-1

10-2

10-3
8-PSK 16-PSK

10-4
M=2,4 16-QAM 64-QAM
QAM
10-5 PSK

10-6
0 2 4 6 8 10 12 14 16 18 20
pow_eff MQAMPSK Apr. 10, 2001 Eb/NO (dB)

Figure 4.38 – Pe for different Modulation Scheme

120
 Review Question
A 4,800 bits/s data terminal is connected to a modem.
Calculate the transmission bandwidth required at the modem
output for each of the following schemes, for r = 0.5:

1.Binary ASK.

2.Binary FSK. Assume that the frequency deviates 3,600 Hz

about the carrier.

?
3. 16-QAM.

121
 Review Question
Ans
Bit rate Rb = 4,800 bits/s, r = 0.5, and B = (Rb / 2)(1 + r)
(1) BASK
BT = 2B = Rb(1 + ) = 4,800(1 + 0.5) = 7,200 Hz

(2) BFSK; Frequency deviation f = 3,600 Hz

BT = 2f + 2B
= 2f + Rb(1 + r)
= 2 x 3,600 + 7,200 = 14,400 Hz

(3)16-QAM; Number of bits per symbol k = log216 = 4

BT = 2B / k
= Rb(1 + r) / 4
= 7,200 / 4 = 1,800 Hz 122
 Review Question
A 9,600 bits/s transmission over a telephone line is desired.
For this purpose the line must be specially conditioned to
allow signal transmission over the frequency range of 300 to
3,000 Hz. Show that 16-state QAM, with r = 0.125 will provide
the desirable bit rate.
Ans
16 QAM; Rb = 9,600 bps, BT = | 300 – 3,000 Hz | = 2,700
Hz, r = 0.125, k = 4, and
B = (Rb / 2)(1 + r)

?
BT = 2B / k
BT = Rb(1 + r) / k
2,700 = Rb(1 + 0.125) / 4
Rb = 9,600 bps (Proven) 123
 Review Question
A PCM binary stream is produced using 8,000 samples/s rate
and 7-bit encoding. 32 synchronisation bits are inserted into
the binary stream for every 224 data bits. The resulting binary
stream is then transmitted in the form of two-level pulses over
a baseband channel, with r = 0.2. This is followed by digital
band-pass modulation prior to the transmission.
1.What is the PCM bit rate?
2.What is the bandwidth of the baseband transmission?
3.Determine the bandwidth of the pass-band transmission
under the following digital modulation schemes:

?
i. BPSK.
ii. QPSK.
iii. BASK.
124
 Review Question
Ans
(1) PCM data rate = 8,000 x 7 = 56,000 bps
(2) 224 data bits require 224 bits / 56 kbps = 4 x 10-3 s
During this time, no. of bits sent including sync bits = 224 +

32 = 256. Hence, PCM bit rate Rb = 256 / 4 x 10-3 = 64 kbps

Therefore, baseband BW, B = (Rb / 2)(1 + r)
= (64 x 103 / 2)(1.2) = 38.4 kHz
3(i)For BPSK, k = 1
BT = 2B / k = 2 x 38.4 kHz / 1 = 76.8 kHz
(ii) For QPSK, k = 2
BT = 2B / k = B = 38.4 kHz
125
 World’s Most Popular Modulation
• Gaussian minimum shift keying
• Gaussian filtered form of FSK
• Sum of I and Q results in a constant amplitude circle

Figure 4.39 – Other Modulations

126
 Minimum Shift Keying (MSK)
• MSK can be derived from OQPSK using half-sine
pulses.

• The MSK signal is defined by:

s(t) = I(t) sin(πt/2T) cos(2π ft) Q(t) sin(πt/2T) sin(2π ft)

127
Minimum Shift Keying (MSK) (cont’d)

128
Minimum Shift Keying (MSK) (cont’d)

129
 Minimum Shift Keying (MSK) (cont’d)

• MSK makes the phase change linear and limited to

±90°.

130
 Filter Alteration of the Signal
01 00
I - Signal
Low Pass Filter
To IQ
Q - Signal Modulator
10 11 Low Pass Filter

Modulation Mapping

Figure 4.40 – Effect of Filtering

131
 Common Filter Types
• Gaussian

Root Raised Cosine

• Raised cosine
Raised Cosine

• Root raised cosine Gaussian

• Setting up the demodulation

process requires knowledge
of what filter is used

Figure 4.41 – Type of Filters

132
 Roll-Off Effect
• Usually α is between 0.2 – α = 0.0 α = 1.0
0.5
BW = (1+α) *Symbol rate

• α determines the
bandwidth

Figure 4.42 – Roll-Off Effect

133
Effect of Baseband Filtering (QPSK)

135
 Symbol Rate and Bit Rate
• The modulation type determines the number of bits per symbol.
– BPSK 1 bit/symbol
– DBPSK 1 bit/symbol
– QPSK 2 bits/symbol
– p/4 DQPSK 2 bits/symbol
– DQPSK 2 bits/symbol
– 8PSK 3 bits/symbol
– 16QAM 4 bits/symbol
– 64QAM 5 bits/symbol
– 256QAM 6 bits/symbol
• For a fixed symbol rate, having more bits will provide a faster
transfer rate.
• Setting up a WCA requires you to know the modulation type and
symbol rate, not the bit rate.

136
 Others
• 32QAM • HPSK
– ADSL, etc. – Hybrid phase shift keying
– Also known as Orthogonal
• 256QAM
Complex Quadrature Phase
– Microwave communication Shift Keying (OCQPSK)
– Some cable modems – Used in the CDMA2000®
(1xRTT) reverse link
• 1024QAM
– Still experimental • VSB
• OQPSK – Vestigial side band
– Offset QPSK – 8VSB, 16VSB
– Used to avoid zero crossings – US digital broadcast TV

• DQPSK

137
 Summary
• Digital modulation is cheaper, faster, more accurate,
more efficient, and more secure.
• Higher order modulation is used for greater
transmission rates in the same spectrum occupancy.
• Higher order modulation is more susceptible to noise.
• Baseband filters are used to control spectrum.
• Vector signal analyzers such as Keysight VSA 89600
can be used to evaluate modulation quality.

141
 Putting IQ on the Carrier Wave
I Ch. sig
• Two mixers in phase quadrature I(t)
• I (in-phase) & Q (quadrature) signals I signal
cos(  t)
• Use I and Q signals to map to any Carrier ( )
S(t)
point in signal space (amplitude and
phase) 90 o

• Use to create higher order sin(  t)

modulation schemes Q signal

• e.g. 64-QAM (64 amplitude and Q(t)

phase states) Q Ch. sig
sin(  t)
S(t) = I(t)cos(t) – Q(t)sin(t) = A(t)[cos (t + (t))]
g
Mag M  I 2  Q2 Ma
Phase
cos(  t)
Q
Phase   atan 
I
142
Separating the Components: The Receiver
S(t) = I(t)cos(t) – Q(t)sin(t)

The composite signal is separated by multiplying (mixing) sin(t) and cos

(t), the resulting sin2 and cos2 terms become
[I(t) or Q(t)] × ½[1 ± cos(2t)] terms – the 2t’s are removed by the LPF.

LPF I(t)

S(t)

cos(t) +90º

LPF Q(t)

143
 Transmitter Basics

Mod
Processing/ I I
Encode
A/D Compression/
Symbols
Error Correction
Q Q

IF RF


Convert to digital if necessary (signal coding)

DSP, channel coding (compression, error correction)

Map to I and Q

Modulation shaping filter

Modulate

Convert to RF

Filter, amplify, send to antenna

Figure 4.45 – Block Diagram for the Transmitter

144
 Receiver Basics

I I Adaption/
Decode Process/
AGC Demod Q Q D/A
Bits Decompression

RF IF


Convert to IF

Filter

Carrier, symbol recovery

Demodulate

Decode Bits

Convert to analog (if necessary)
Figure 4.46 – Block Diagram for the Receiver

145
 References
[1] Leon W. Couch II, “Digital and Analog Communication Systems”,
Pearson Education 7th Edition., N.J., 2007.
[2] Bernard Sklar, “Digital Communications : Fundamentals and
Applications”, Prentice Hall, 2nd Edition, 2001.
[3] B. P. Lathi, “Modern Digital and Analog Communication Systems”,
Oxford University Press 3rd Edition., N.Y., 1998.
[4] Digital Modulation in Communications Systems — An Introduction,
Keysight Application Note 1298, Literature Number 5965-7160E.
[5] Simon Haykin, “Communication Systems”, John Wiley & Sons, 4th
Edition., N.Y., 2001

146