Communication Toolbox

Communication
in
MATLAB

Communication ToolBox

Communication
Simulink
block-Set

Provide Ready-to-use tools to design, analysis and simulate

communication systems, both for analog and digital

Communication Toolbox
Communication -Simulink

Source and sink in communication systems

Source coding
Error Correction Coding
Modulation and Demodulation
Signal Multiple Access
Transmitting and Receiving filters
Channels
2

Information
source

Transmitter

Channel

Receiver

Information
sink
3

MATLAB functions
The MATLAB functions are in
sub-directory comm

help comm
..
help amod
help qam

Signal Sources and Display Functions.

biterr

- Bit error number and bit error rate

computation.

eyescat

- Eye-pattern diagram or scatter plot.

randint

- Random integer matrix generator.

randbit

- Binary noise generator.

symerr
rate

- Symbol error number and symbol error

5
computation.

Random Signals Genrator

Provides

Gaussian

Uniform
Rayleigh

Poisson
Rice
6

Random Signals Genrator

The

wgn

White Gaussian Noise

function generates random matrices using a white

Gaussian noise distribution.

The power of the noise can be specify in either dB (decibels), dBm, or linear
units.
Either real or complex noise can generated.

White Gaussian Noise - example

For, the command below generates a column vector of length 50
containing real white Gaussian noise whose power is 2 dB. The
function assumes that the load impedance is 1 Ohm.

y1 = wgn(50,1,2)
sum(y1.^2)/length(y1)

y4 = wgn(200,1,2,60,linear)
y4 = wgn(20000,1,2,60)
sum(y4.^2)/length(y4)

To generate complex white Gaussian noise whose

power is 2 watts, across a load of 60 Ohms.

y2 = wgn(50,1,2,60,'complex','linear');
or
y3 = wgn(50,1,2,60,'linear','complex');8

Random Symbol Matrices

randsrc

The
function generates random matrices whose entries
are chosen independently from an alphabet that is specify, with a
distribution that you specify.
A special case generates bipolar matrices.
For example, to generates a 5-by-4 matrix whose entries are
independently chosen and uniformly distributed in the set {1,3,5}.
a=

randsrc

(5,4,[1,3,5])

a=
3

If we want

1 to be twice (as likely to occur as either 3 or 5),

then we use the command to prescribe the skewed distribution.

b = randsrc(5,4,[1,3,5; .5,.25,.25])
b=
3

50%

10

Random Integer Matrices

The randint function generates random integer matrices
whose entries are according the specified range.
A special case generates random binary matrices.
For example, the command below generates a 5-by- 4 matrix
containing random integers between 2 and 10.
c=

randint(5,4,[2,10])

c=
2

5 10

7 10

10

4 10

randsrc(5,4,[2])

randint(5,4,2)

ans =

ans =

2
2
2
2
2

2
2
2
2
2

2
2
2
2
2

2
2
2
2
2

0
0
0
0
0

1
0
0
0
1

1
0
0
1
0

1
1
1
1
0

11

Random Bit Error Patterns

The randerr function generates matrices whose entries are either 0
or 1. However, its options are rather different from those of
randint, since randerr is meant for testing error-control coding.
Example, the command below generates a 5-by-4 binary matrix
having the property that each row contains exactly one 1.

randerr (5,4)

f=
f=

12

if you want to perturb each codeword by introducing one error

with probability 0.4 and two errors with probability 0.6, then the
command below should replace the one above.
% Each row has one '1' with probability 0.4, otherwise two '1's
g = randerr(5,4,[1,2; 0.4,0.6])
g=
0

Note : The probability matrix that is the third argument of

randerr affects only the number of 1s in each row, not their
placement.

13

Example
generating an equiprobable binary 100-element column vector
using any of the commands below.
The three commands produce different numerical outputs, but use
the same distribution.
binarymatrix1 = randsrc(100,1,[0 1] ); % Possible values are 0,1.

binarymatrix2 = randint(100,1 ,2 ); % Two possible values

binarymatrix3 = randerr(100,1,[0 1;.5 .5] ); % No 1s, or one 1

Third argument

14

Error Rates
Comparing messages before and after transmission
for evaluating the quality of a communication system
design (or the evaluate algorithm performances).
Counting bit errors is different from counting symbol errors.
In either the bit-or symbol-counting case,

Number of errors (bit or symbols)

error rate= Total transmitted number (of bits or symbols).

15

Error Rates - (continue)

Bits
The

biterr function computes the number of bit errors and

the bit error rate.

Symbols
The

symerr function computes the number of symbol

errors and the symbol error rate.

16

Example: Computing Error Rates

Dugma2.m
The script below uses the symerr function to compute the symbol
error rates for a noisy linear block code. After artificially adding
noise to the encoded message, it compares the resulting noisy code
to the original code. Then it decodes and compares the decoded
message to the original one.

17

m = 3; n = 2^m-1; k = n-m; % Prepare to use Hamming code.

msg = randint(k*200,1,2);

% 200 messages of k bits each

code = encode(msg,n,k,'hamming');
codenoisy = rem(code+(rand(n*200,1)>.95),2); % Add noise.
% Decode and correct some errors.
newmsg = decode(codenoisy,n,k,'hamming');
% Compute and display symbol error rates.
[codenum,coderate] = symerr(code,codenoisy);
[msgnum,msgrate] = symerr(msg,newmsg);
disp(['Error rate in the received code: ',num2str(coderate)])
disp(['Error rate after decoding: ',num2str(msgrate)])

18

Comparison of Symbol Error Rate and Bit Error

Rate

In the example above, the symbol errors and bit errors are the
same because each symbol is a bit. The commands below
illustrate the difference between symbol errors and bit errors in
other situations.
a = [1 2 3]'; b = [1 4 4]';
format rat % Display fractions instead of decimals.
[snum,srate] = symerr(a,b)
snum =
2
srate =
2/3

19

a = [1 2 3]'; b = [1 4 4]'; a=(001,010,011) b=(001,100,100)

[bnum,brate] = biterr(a,b)
bnum =
5
brate =
5/9

bnum is five because the second entries differ in two bits and the
third entries differ in three bits. brate is 5/9 since the total number
of bits is nine. The total number of bits is, by definition, the
number of entries in a or b times the maximum number of bits
among all entries of a and b

20
Dugma2.m

Display devices
Provides display devices that make it easier to
analyze the performance of communication system
Display devices

Eye-Pattern
diagram

Scatter
plot

Error Rate
computation

21

Eye Diagrams
An eye diagram is a tool for studying the effects of intersymbol
interference and other channel impairments in digital transmission.
To construct an eye diagram, 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.
One way to use an eye diagram is to look for the place where the
"eye" is most widely opened, and use that point as the decision point
when sampling to recover a digital message.

22

How to draw an Eye diagram

To produce an eye diagram from a signal, use the eyediagram function.

EYEDIAGRAM(X, N, PERIOD, OFFSET, PLOTSTRING)

generates an eye diagram of X with N samples per

trace. N must be an integer greater than 1. X can be a real or complex
vector, or a two-column matrix with real signal in the first column and
imaginary signal in the second column. If X is a real vector, EYEDIAGRAM
generates one eye diagram. If X is a two-column matrix or a complex
vector, EYEDIAGRAM generates two eye diagrams, one for the real (in-phase)
PERIOD is used to determine the horizontal axis limits. PERIOD
must be a positive number. The horizontal axis limits are -PERIOD/2 and
+PERIOD/2. The default value of PERIOD is 1.
OFFSET determines which points are centered on the horizontal axis
starting with the (OFFSET+1)st point and every Nth point thereafter. OFFSET
must be a nonnegative integer in the range 0 <= OFFSET < N. The default

PLOTSTRING can be any of the strings used in the PLOT function value for OFFSET is 0.
.
eye_diagram.m23

Scatter Plots
A scatter plot of a signal shows the signal's value at a given decision point. In the
best case. The decision point should be at the time when the eye of the signal's eye
diagram is the most widely open.
To produce a scatter plot from a signal,
use the scatterplot function.
SCATTERPLOT(X, N, OFFSET, PLOTSTRING)
generates a scatter plot of X. X can be a real or complex vector, or
a two-column matrix with real signal in the first column and
imaginary signal in the second column.
SCATTERPLOT(X, N) generates a scatter plot of X with decimation factor N.
Every Nth point in X is plotted, starting with the first value. The default
for N is 1.
OFFSET is the number of samples skipped at the beginning of X before
plotting. The default value for OFFSET is zero.
PLOTSTRING can be any of the strings used in the PLOT function. The default value for
PLOTSTRING is 'b.'.

24
scatterplots.m

Signal Generator
Random Signal Generator

Source Coding
commands
Error-Control Coding
Lower-Level Functions
for ECC & GF

Modulation
&
Demodulation
25

Source Coding

Part 2

Source coding, also known as quantization or signal formatting

a way of processing data in order to reduce redundancy or prepare
it for later processing.
Analog-to-digital conversion and data compression are two
categories of source coding.

26

Source Coding

Part 2

Source coding divides into two basic procedures

source encoding

source decoding

converts a source signal

into a digital signal
using a quantization method.
The symbols in the resulting signal
are nonnegative integers
in some finite range.

recovers the original

information from the
source coded signal.
27

Representing Quantization Parameters

Scalar quantization is a process that maps all inputs within a specified range to
a common value.

Mapping process
Input

Common
value

It maps inputs in a different range of values to a different common value.

scalar quantization digitizes an analog signal

28

Representing Quantization Parameters

. Two parameters determine a quantization.
a partition and a codebook

Partitions
A quantization partition defines several contiguous, nonoverlapping ranges of
values within the set of real numbers.
For example, if the partition separates the real number line into the four sets:

1.{x : x 0}
2.{x : 0 x 1}
3.{x : 1 x 3}
4.{x : 3 x}

2
0

3
1

4
3

then you can represent the partition as the three-element vector

partition = [0,1,3];
Notice that the length of the partition vector is one less than the number of
partition intervals.

29

Codebooks
A codebook tells the quantizer which common value to assign to inputs that
fall into each range of the partition. Represent a codebook as a vector whose
length is the same as the number of partition intervals. For example, the
vector
codebook = [-1, 0.5, 2, 3];
is one possible codebook for the partition [0,1,3].

3
2
0.5
-1
1

2
0

3
1

4
3

30

Compand - Source code mu-law or A-law compressor or expander

dpcmdeco - Decode using differential pulse code modulation

dpcmenco - Encode using differential pulse code modulation

dpcmopt - Optimize differential pulse code modulation parameters
lloyds - Optimize quantization parameters using the Lloyd algorithm
quantiz- Produce a quantization index and a quantized output value

31

Source encoding
Partition(1)

Signal

indx

Partition(2)

+
Quantization decode

Partition(N)

quant

Codebook(indx)

+distor
(.)2

Distortion
computation

1/M

+
Memory

32

Quantizing a Signal
How to use partition and codebook parameters in
the quantiz function ?

quantiz
Produce a quantization index and a quantized output value
Syntax
index = quantiz(sig,partition);
[index,quants] = quantiz(sig,partition,codebook);
[index,quants,distor] = quantiz(sig,partition,codebook);

33

Scalar Quantization Example 1

Dugma5.m
The code below shows how the quantiz function uses
partition and codebook to map a real vector, sample, to a
new vector, quantized, whose entries are either -1, 0.5, 2,
or 3.
-1
partition = [0,1,3];

]( 0.5 ](
0
1

]( 3
3

codebook = [-1, 0.5, 2, 3];

sample = [-2.4, -1, -.2, 0, .2, 1, 1.2, 1.9, 2, 2.9, 3, 3.5, 5];
[index,quantized] = quantiz(sample,partition,codebook);
quantized

34

Dugma5.m

Scalar Quantization Example

2
Dugma6.m
This example illustrates the nature of scalar quantization. After
quantizing a sampled sine wave, it plots the original and quantized
signals. The plot contrasts the x's that make up the sine curve with
the dots that make up the quantized signal. The vertical coordinate
of each dot is a value in the vector codebook.
t = [0:.1:2*pi];

% Times at which to sample the sine function

sig = sin(t);

% Original signal, a sine wave

partition = [-1:.2:1];

% Length 11, to represent 12 intervals

codebook = [-1.2:.2:1];

% Length 12, one entry for each interval

[index,quants] = quantiz(sig,partition,codebook); % Quantize.

plot(t,sig,'x',t,quants,'.')
axis([-.2 7 -1.2 1.2])

35

Dugma6.m

partition = [0,1,3];
codebook = [-1, 0.5, 2, 3];
samp = [-2.4, -1, -.2, 0, .2, 1, 1.2, 1.9, 2, 2.9, 3, 3.5, 5];
[index,quantized] = quantiz(samp,partition,codebook);
quantized
quantized =
Columns 1 through 6
-1.0000 -1.0000 -1.0000 -1.0000 0.5000 0.5000
Columns 7 through 12
2.0000 2.0000 2.0000 2.0000 2.0000 3.0000
Column 13
3.0000
Determining Which Interval Each Input Is in
partition = [3,4,5,6,7,8,9];
index = quantiz([2 9 8],partition)
index =
0
6
5

36

Optimizing Quantization Parameters

Quantization distorts a signal. You can lessen the distortion by choosing appropriate partition
and codebook parameters. One way to produce partition and codebook parameters easily is to
optimize them according to a set of so-called training data.

lloyds
Optimize quantization parameters using the Lloyd algorithm
Syntax
[partition,codebook] = lloyds(trainingset,initcodebook);
Description
[partition,codebook] = lloyds(trainingset,initcodebook) optimizes the
scalar quantization parameters partition and codebook for the training data in the
vector trainingset. initcodebook, a vector of length at least 2, is the initial guess
of the codebook values. The output codebook is a vector of the same length as
initcodebook. The output partition is a vector whose length is one less than the
length of codebook.

37

Example: Optimizing Scalar Quantization Parameters

Dugma7.m
The lloyds function optimizes the partition and codebook
according to the Lloyd algorithm.
t = [0:.1:2*pi];
sig = sin(t);
partition = [-1:.2:1];
codebook = [-1.2:.2:1];
% Now optimize, using codebook as an initial guess.
[partition2,codebook2] = lloyds(sig,codebook);
[index,quants,distor] = quantiz(sig,partition,codebook);
[index2,quant2,distor2] = quantiz(sig,partition2,codebook2);
% Compare mean square distortions from initial and optimized
[distor, distor2] % parameters.

38

Dugma 7.m

Companding a
Signal
In certain applications, such as speech processing, it is common
to use a logarithm computation, called a compressor, before
quantizing. The inverse operation of a compressor is called an
expander. The combination of a compressor and expander is
called a compander.

The compand function supports two kinds of companders: -law

and A-law companders.

39

-law

compressor

Algorithm
For a given signal x, the output of the -law compressor is

V log 1

sgn(x)
y
log 1
where
V is the maximum value of the signal x,
is the -law parameter of the compander,
log is the natural logarithm
sgn is the signum function (sign in MATLAB).

40

A-law

compressor

The output of the A-law compressor is

A x

sgn(x)

1 log( A)

Ax

V 1 log

sgn(x)
1 log( A)

where
A is the A-law parameter of the compander
other elements are as in the -law case.

V
0 x
A
V
x V
A

41

compand
Source code mu-law or A-law compressor or expander
Syntax

out = compand(in,Mu,maxim);
out = compand(in,Mu,maxim,'mu/compressor');
out = compand(in,Mu,maxim,'mu/expander');
out = compand(in,A,maxim,'A/compressor');
out = compand(in,A,maxim,'A/expander');

42

Examples compressors and expanders

.

compressed = compand(1:5,87.6,5,'a/compressor')
compressed =
3.5296 4.1629 4.5333 4.7961 5.0000
expanded = compand(compressed,87.6,5,'a/expander')
expanded =
1.0000 2.0000 3.0000 4.0000 5.0000

43

mu = 255;

% Parameter for mu-law compander

sig = -4:.1:4;
sig = exp(sig);

% Exponential signal to quantize

V = max(sig);
% 1. Quantize using equal-length intervals and no compander.
[index,quants,distor] = quantiz(sig,0:floor(V),0:ceil(V));
% 2. use same partition and codebook, but compress
% before quantizing and expand afterwards.
compsig = compand(sig,mu,V,'mu/compressor');
[index,quants] = quantiz(compsig,0:floor(V),0:ceil(V));
newsig = compand(quants,mu,max(quants),'mu/expander');
distor2 = sum((newsig-sig).^2)/length(sig);
[distor, distor2]

% Display both mean square distortions.

44

The entire code will be under the name A -Law Compander as *.m file

Implementing Differential Pulse

Code Modulation (DPCM)
DPCM - Differential Pulse code Modulation

Signal

Source encoded index indx(k)

e(k)
+

Quantization
source encode

quantized y(k)
+
Predictor

xpredicted(k)
45

DPCM - Differential Pulse code Modulation - decoding method

Source encoded
index indx(k)

Quantization
source encode

quantized y(k)
+
Predictor

46

dpcmenco
Encode using differential pulse code modulation
Syntax
indx = dpcmenco(sig,codebook,partition,predictor)

Description
encode the vector sig.

partition is a vector whose entries give the endpoints

of the partition intervals.

codebook, a vector whose

length exceeds the length of partition by one, prescribes
a value for each partition in the quantization.

predictor specifies the predictive transfer function.

If the transfer function has predictive order M,
then predictor has length M+1 and an initial entry of 0.

47

dpcmdeco
Decode using differential pulse code modulation
Syntax
sig = dpcmdeco(indx,codebook,predictor);
[sig,quanterror] = dpcmdeco(indx,codebook,predictor);

Description
sig = dpcmdeco(indx,codebook,predictor)
implements differential pulse code demodulation to decode
the vector

indx.

codebook represents the predictive-error

The vector
quantization codebook.

The vector predictor specifies the predictive transfer function.

48

predictor = [0 1]; % y(k)=x(k-1)

Dugma8.m

partition = [-1:.1:.9];
codebook = [-1:.1:1];
t = [0:pi/50:2*pi];
x = sawtooth(3*t); % Original signal
% Quantize x using DPCM.
encodedx = dpcmenco(x,codebook,partition,predictor);
% Try to recover x from the modulated signal.
decodedx = dpcmdeco(encodedx,codebook,predictor);
plot(t,x,t,decodedx,'--')
distor = sum((x-decodedx).^2)/length(x) % Mean square error
The entire code will be under the name DPCM Encoding and
Decoding as *.m file

49

Dugma 8.m

dpcmopt
Optimize differential pulse code modulation parameters
Syntax
predictor = dpcmopt(trainingset,ord);
[predictor,codebook,partition] = dpcmopt(trainingset,ord,len);
[predictor,codebook,partition] =
dpcmopt(trainingset,ord,initcodebook);

50

Example: Comparing Optimized and

Nonoptimized DPCM Parameters

Dugma9.m
compared to
Dugma8.m

t = [0:pi/50:2*pi];
x = sawtooth(3*t);

% Original signal

initcodebook = [-1:.1:1]; % Initial guess at codebook

% Optimize parameters, using initial codebook and order 1.
[predictor,codebook,partition] = dpcmopt(x,1,initcodebook);
% Quantize x using DPCM.
encodedx = dpcmenco(x,codebook,partition,predictor);
% Try to recover x from the modulated signal.
decodedx = dpcmdeco(encodedx,codebook,predictor);
distor = sum((x-decodedx).^2)/length(x) % Mean square error

51

Modulation Features of the Toolbox - continue

Matlab supports for analog and digital signals

Modulation methods for analog signals

Frequency
modulation (FM)

Quadrature
Amplitude
Modulation(QAM)

Amplitude
modulation (AM)

Single side band

suppressed
carrier (SSB)

Double sideband
transmission
carrier (DSB-TC)

Phase
modulation (PM)

Double sideband
suppressed carrier
(DSB-SC)

52

Modulation Features of the Toolbox - continue

Matlab - digital signals

Modulation methods for digital signals
Amplitude shift
keying (ASK)

Quadrature
Amplitude (QAM)

Frequency shift
keying (FSK)

Phase shift
keying (PSK)

53

Two option for representing

Modulation/demodulation process
Modulation/Demodulation options

Passband
Analog
mod/dem
amod
ademod

Digital
mod/dem
dmod
ddemod.
Modmap

Baseband
Analog
mod/dem
amodce
ademodce

Digital
mod/dem
Modmap
demodmap
ddemodce
dmodce

Complex envelope 54

Supported Modulation Tasks

Functions in the toolbox can accomplish these tasks:
1. Modulate a signal using one of the mentioned
techniques
2. Demodulate a signal using one of the mentioned
techniques
3. Map a digital signal to an analog signal, before
modulation
4. Demap an analog signal to a digital signal, after
demodulation
5. Map, demap, and plot constellations for QASK
modulation

55

x(t)

Baseband
presentation

0f

cos C t

A 1 x(t ) cos
C

t
A

Passband
presentation

56

Modulation/Demodulation.
ademod

- Analog demodulation computation.

ademodce - Demodulate from a complex envelope of

analog modulated signal.
amod

- Analog modulation.

amodce

- Complex envelope of analog modulation.

apkconst

- Compute/plot ASK/PSK constellation.

ddemod

- Demodulate a digital modulated signal.

ddemodce - Demodulate from complex envelope of

digital modulated signal.

57

Modulation/Demodulation-continue
demodmap

- Digital demodulation demapping.

dmod

- Digital modulation.

dmodce

- Complex envelope of digital modulation.

modmap

- Digital modulation mapping.

qaskdeco

- Decode a two-dimensional code to one

dimension message.

qaskenco

- Encode a message to two-dimensional

form for QASK use.

58

Analog Modulation
Techniques

59

amod

Passband modulation

Analog passband modulator

Syntax
y = amod(x,Fc,Fs,'amdsb-sc');
y = amod(x,Fc,Fs,'amdsb-tc',offset);
y = amod(x,Fc,Fs,'amssb/opt');
y = amod(x,Fc,Fs,'amssb/opt',num,den);
y = amod(x,Fc,Fs,'amssb/opt',hilbertflag);
y = amod(x,Fc,Fs,'qam');
y = amod(x,Fc,Fs,'fm',deviation);
y = amod(x,Fc,Fs,'pm',deviation);
y = amod(x,Fc,[Fs phase],...);
[y,t] = amod(...);

60

amod

Passband modulation

Analog passband modulator

Syntax
y = amod(x,Fc,Fs,'amdsb-sc');
The syntax y = amod(x,Fc, ,[Fs phase],...)
modulates the received signal that x represents.
Fc is the carrier frequency in Hertz,
Fs is the sampling rate in Hertz.
(1/Fs represents the time interval between two consecutive
samples in x.).

The third input argument is a two-element vector

(instead of a scalar. )

Fs, is the sampling rate

The second entry, phase, is the initial phase of the
carrier signal, measured in radians.

61

Passband modulation
DSB-SC

Double Side Band - DSB -SC

(Suppressed Carrier).
y t A m t cos(2 f c t )

m(t)
message

Amplifier

A cos c t
62

amod
Analog passband modulator

Syntax
y = amod(x,Fc,Fs,'amdsb-sc');
y = amod(x,Fc,Fs,'amdsb-tc',offset);
y = amod(x,Fc,Fs,'amssb/opt');
y = amod(x,Fc,Fs,'amssb/opt',num,den);
= amod(x,Fc,Fs,'amssb/opt',hilbertflag);
=y amod(x,Fc,Fs,'amdsb-tc',offset);
y = amod(x,Fc,Fs,'qam');
y = amod(x,Fc,Fs,'fm',deviation);
Offset -min(min(x))
y = amod(x,Fc,Fs,'pm',deviation);
y = amod(x,Fc,[Fs phase],...);
[y,t] = amod(...);

63

Ademod
The function ademod performs analog passband demodulation.
Syntax
z = ademod(y,Fc,Fs,'amdsb-tc',offset,num,den);
z = ademod(y,Fc,Fs,'amdsb-tc/costas',offset,num,den);
z = ademod(y,Fc,Fs,'amdsb-sc',num,den);
z = ademod(y,Fc,Fs,'amdsb-sc/costas',num,den);
z = ademod(y,Fc,Fs,'amssb',num,den);
z = ademod(y,Fc,Fs,'qam',num,den);
z = ademod(y,Fc,Fs,'fm',num,den,vcoconst);
z = ademod(y,Fc,Fs,'pm',num,den,vcoconst);
z = ademod(y,Fc,[Fs phase],...);
64

Passband modulation
DSB-SC

Double Side Band - DSB -SC

(demodulator).

y t A m t cos(2 f c t ) noise( t )
^

Low Pass filter

m t
message

2 A sin( 2 f c t c )
Local Oscillator

z = ademod(y,Fc,Fs,'amdsb-sc',num,den);
65

amod
Analog passband modulator
Syntax
y = amod(x,Fc,Fs,'amdsb-sc');
y = amod(x,Fc,Fs,'amdsb-tc',offset);
y = amod(x,Fc,Fs,'amssb/opt');
y = amod(x,Fc,Fs,'amssb/opt',num,den);
y = amod(x,Fc,Fs,'amssb/opt',hilbertflag);
y = amod(x,Fc,Fs,'qam');
y = amod(x,Fc,Fs,'fm',deviation);
y = amod(x,Fc,Fs,'pm',deviation);
y = amod(x,Fc,[Fs phase],...);
[y,t] = amod(...);

66

Passband modulation
AM-SSB
x(t)

SIDE
SIDEBAND
BAND
FILTER
FILTER

cos c1 t

x(t ) cos c t
C

DSB

cos c 2 t

H f

SSB

USB

LSB

LSB

f
c

USB

67

f t f t F 1 Sgn f t 1
*
* t
Hilbert Filter
j , 0
H jSgn
j , 0

f t

H 1 e

S t
SSB

68

SSB Hilbert Transmitter

s t cos c t

L.O

90

Hilebrt
filter

SSB

(t )

sin c t

s (t ) s(t ) sin c t
69

Ademod
The function ademod performs analog passband demodulation.
Syntax
z = ademod(y,Fc,Fs,'amdsb-tc',offset,num,den);
z = ademod(y,Fc,Fs,'amdsb-tc/costas',offset,num,den);
z = ademod(y,Fc,Fs,'amdsb-sc',num,den);
z = ademod(y,Fc,Fs,'amdsb-sc/costas',num,den);
z = ademod(y,Fc,Fs,'amssb',num,den);
z = ademod(y,Fc,Fs,'qam',num,den);
z = ademod(y,Fc,Fs,'fm',num,den,vcoconst);
z = ademod(y,Fc,Fs,'pm',num,den,vcoconst);
z = ademod(y,Fc,[Fs phase],...);
70

Quadrature Amplitude
Modulation
QAM
m I (t)

cos(2f c t c )
900

mQ (t)

y( t )

y( t ) m I ( t ) cos(2f c t c ) m Q ( t ) sin( 2f c t c )
y = amod(x,Fc,Fs,'qam');

71

Quadrature Amplitude
DeModulation
QAM

y( t )

LPF

m I (t)

LPF

mQ (t)

cos(2f c t c )
900

z = ademod(y,Fc,Fs,'qam',num,den);
72

73

 Mathematical expressions

FM,PM

y = amod(x,Fc,Fs,'pm',deviation);

cos c t 2 k c t c

y t

cos

t k c m d

PM
FM

y = amod(x,Fc,Fs,'fm',deviation);
74

Ademod

FM,PM

The function ademod performs analog passband demodulation.

Syntax
z = ademod(y,Fc,Fs,'amdsb-tc',offset,num,den);
z = ademod(y,Fc,Fs,'amdsb-tc/costas',offset,num,den);
z = ademod(y,Fc,Fs,'amdsb-sc',num,den);
z = ademod(y,Fc,Fs,'amdsb-sc/costas',num,den);
z = ademod(y,Fc,Fs,'amssb',num,den);
z = ademod(y,Fc,Fs,'qam',num,den);
z = ademod(y,Fc,Fs,'fm',num,den,vcoconst);
z = ademod(y,Fc,Fs,'pm',num,den,vcoconst);

75

FM Demodulator block diagram

FM

Phase
detector

LPF
VCO

z = ademod(y,Fc,Fs,'fm',num,den,vcoconst)
implements frequency demodulation.
The spectrum of the demodulated signal is between min(y) + Fc
and max(y) + Fc. The demodulation process uses a phase-locked loop
composed of a multiplier (as a phase detector), a lowpass filter, and a
voltage-controlled oscillator (VCO).
If Fs is a two-element vector, then its second
element is the initial phase of the VCO, in radians.
The optional argument vcoconst is a scalar that represents the
VCO constant in Hz/V.
76

PM Demodulator block diagram

PM

Phase
detector

LPF

1/s

VCO

z = ademod(y,Fc,Fs,'pm',num,den,vcoconst)
Implements phase demodulation. The demodulation
process uses a phase-locked loop (which acts as an
FM demodulator) cascaded with an integrator. The phase-locked
loop consists of a multiplier (as a phase detector), a lowpass
filter, and a
voltage-controlled oscillator (VCO).
If Fs is a two-element vector, then its second element is the
initial phase of the VCO, in radians.
The optional argument vcoconst is a scalar that represents the
77
input signal's sensitivity.

Example
Double- and Single-Sideband Comparison Example
% Sample the signal 100 times per second, for 2 seconds.
Fs = 100;
t = [0:2*Fs+1]'/Fs;
Fc = 10; % Carrier frequency
x = sin(2*pi*t); % Sinusoidal signal
% Modulate x using single- and double-sideband AM.
ydouble = amod(x,Fc,Fs,'amdsb-sc');
ysingle = amod(x,Fc,Fs,'amssb');
% Plot spectra of both modulated signals.
zdouble = fft(ydouble);
zdouble = abs(zdouble(1:length(zdouble)/2+1));
frqdouble = [0:length(zdouble)-1]*Fs/length(zdouble)/2;
plot(frqdouble,zdouble); % The plot on the left-hand side below
figure;
zsingle = fft(ysingle);
zsingle = abs(zsingle(1:length(zsingle)/2+1));
frqsingle = [0:length(zsingle)-1]*Fs/length(zsingle)/2;
plot(frqsingle,zsingle); % The plot on the right-hand side below

78

Example
Hilbert Filter Example
The next example uses a Hilbert filter in the time domain.
Fc = 25; % Carrier signal frequency
Fs = 100; % Sampling rate of signal
[numh,denh] = hilbiir(1/Fs,15/Fs,15); % Design Hilbert filter.
t = [0:1/Fs:5]'; % Times to sample the signal
x = cos(t); % Signal is a cosine wave.
y = amod(x,Fc,[Fs pi/4],'amssb',numh,denh); % Modulate,
% using a Hilbert filter in the time domain.
z = ademod(y,Fc,[Fs pi/4],'amssb'); % Demodulate.
plot(t,z) % Plot recovered signal.
%The resulting plot is on the left below. If you replace the sixth line above with
y = amod(x,Fc,[Fs pi/4],'amssb'); % Modulate,
hold on
plot(t,x,,r);

79

ademod uses a lowpass filter with sample time 1/Fs while demodulating,
in order to filter out the carrier signal.
To specify the lowpass filter, include num and den in the
list of input arguments. num and den are row vectors that
give the coefficients, in descending order, of the numerator
and denominator of the filter's transfer function.
Example:
butter is in the Signal Processing Toolbox.
[num,den] = butter(5,Fc*2/Fs);

80

z = ademod(y,Fc,Fs,'amdsb-tc',offset,num,den)
implements double-sideband amplitude demodulation.
offset is a vector whose kth entry is subtracted from the
kth signal after the demodulation.
z = ademod(y,Fc,Fs,'amdsb-tc/costas',offset,num,den) is the same as
the syntax above, except that the algorithm includes a Costas phase-locked loop.
z = ademod(y,Fc,Fs,'amdsb-sc',num,den)
implements double-sideband suppressed-carrier amplitude demodulation.
z = ademod(y,Fc,Fs,'amdsb-sc/costas',num,den) is the same as the syntax
above, except that the algorithm includes a Costas phase-locked loop.
z = ademod(y,Fc,Fs,'amssb',num,den) implements single-sideband
suppressed-carrier amplitude demodulation.

81

example
This example illustrates the use of the offset argument. Since the first ademod
command uses the same offset value of .3 that the amod command used, z1 is
similar to the original message signal. Since the second ademod command omits
offset, z2 has mean close to zero (not exactly zero because of roundoff error).
Fc = 25; % Carrier signal frequency
Fs = 100; % Sampling rate of signal
t = [0:1/Fs:5]'; % Times to sample the signals
x = [cos(t), sin(t)]; % Cosine signal and sine signal
y = amod(x,Fc,Fs,'amdsb-tc',.3); % Modulate
% and shift the values up by .3.
z1 = ademod(y,Fc,Fs,'amdsb-tc',.3); % Demodulate.
z2 = ademod(y,Fc,Fs,'amdsb-tc'); % Demodulate.
plot(t,z1,'b',t,z2,'r--') % Plot recovered signal.

82

example
Example: Varying the Filter's Cutoff Frequency

Fc = 25; % Carrier frequency

Fs = 100; % Signal sampling rate
t = [0:1/Fs:2]'; % Times to sample the signal
x = sawtooth(6*t,0); % Signal is a sawtooth.
y = amod(x,Fc,Fs,'amssb'); % Modulate.
F = Fc/Fs; % Change F to vary the filter's cutoff frequency.
[num,den] = butter(2,F); % Design Butterworth filter.
z = ademod(y,Fc,Fs,'amssb',num,den); % Demodulate and filter.
plot(t,x,'-',t,z,'--') % Plot original and recovered signals.

The code below modulates a sawtooth message signal, demodulates the resulting
signal using a Butterworth filter, and plots the original and recovered signals. Note
that the scaling in the butter function causes the cutoff frequency of the filter to be
F*Fs/2, not F itself.
83

Two option for representing

Modulation/demodulation process
Modulation/Demodulation options

Passband

Analog
mod/dem

Digital
mod/dem

Baseband

Analog
mod/dem

Digital
mod/dem

84

baseband modulation
Modulation
Method

x I (t)

DSB-SC

u(t)

SSB-SC

u(t)

DSB-TC

u(t) k

QAM
FM
PM

u I (t)

(t)

u(t)

u Q (t)

cos k vco u ( )d sin k vco u ( )d

cos k vco u ( )

sin k vco u ( )

85

amodce
Analog baseband modulator with out carrier
Syntax
y = amodce(x,Fs,'amdsb-tc',offset);
y = amodce(x,Fs,'amdsb-sc');
y = amodce(x,Fs,'amssb');
y = amodce(x,Fs,'amssb/time',num,den);
y = amodce(x,Fs,'amssb/time');
y = amodce(x,Fs,'qam');
y = amodce(x,Fs,'fm',deviation);
y = amodce(x,Fs,'pm',deviation);
y = amodce(x,[Fs phase],...);

86

The generic syntax y = amodce(x,Fs,...)

modulates the message signal that x
represents, and returns the modulated
signal's complex envelope.

87

For example, if t measures time in seconds, then the vector x below

is the result of sampling a frequency-one sine wave 100 times per
second for 2 seconds. The vector y represents the modulated signal.
The output shows that y is complex.

Fs = 100; % Sampling rate is 100 samples per second.

t = [0:1/Fs:2]'; % Sampling times for 2 seconds
x = sin(2*pi*t); % Representation of the signal
y = amodce(x,Fs,'pm'); % Modulate x to produce y.
whos
88