Exp1: Computation of Linear and Circular convolution

Aim: To compute linear convolution and circular convolution of two signal by executing
MATLAB program
Equipment and Software Requirement:
1. PC: windows OS,
2. Software: MATLAB

Program 1:
close all; clear all; clc;
x1=[1,2,3,1]
x2= [-1,2,0,-1]
% ploting x1
n1= 0:3;
stem(n1,x1);
% ploting x2
n2= 0:3;
stem(n2,x2);
% Calculating linear convolution between x1 and x2
x3= conv(x1,x2);
% ploting plotting convolution between x1 and x2 i.e x3
n3= 0:6;
stem(n3,x3)
% Displaying x3
disp(x3)
x1=[1,2,3,1]
x2= [-1,2,0,-1]
% ploting x1
n1= 0:3;
subplot(3,1,1);
stem(n1,x1);
% ploting x2
n2= 0:3;
subplot(3,1,2);
stem(n2,x2);
% Calculating linear convolution between x1 and x2
x3= conv(x1,x2)
% ploting plotting convolution between x1 and x2 i.e x3
n3= 0:6;
subplot(3,1,3);
stem(n3,x3)
% Displaying x3
disp(x3)
Output/Observation:
x1 =

1 2 3 1

x2 =

-1 2 0 -1

x3 =

-1 0 1 4 0 -3 -1

Plots:
Program 2
% Finding Circular convolution of two signals
close all; clear all; clc;
x1=[1,-1,1,1]
x2= [2,3,1,4]
% ploting x1
n1= 0:3;
subplot(3,1,1);
stem(n1,x1)
% ploting x2
n2= 0:3;
subplot(3,1,2);
stem(n2,x2)
% Calculating Circular convolution between x1 and x2
x3= cconv(x1,x2,4)
% ploting plotting circular convolution between x1 and x2 i.e x3
n3= 0:3;
subplot(3,1,3);
stem(n3,x3)
% Displaying x3

Output/Observation:
x1 =

1 -1 1 1

x2 =

2 3 1 4

x3 =

2 6 4 8

Plots:
Result:
 Exp2: Spectrum analysis of different signals using DFT/ FFT.
Aim: To implement MATLAB program to obtain magnitude spectrum of signal by DFT/FFT
Equipment and Software Requirement:
1. PC: windows OS,
2. Software: MATLAB

Program:
clc;clear all; close all;
% Finding Length of signal
x=[1 0 1 0 1 0 1 0];
N=length(x);
e=exp(1);
% Computing DFT of signal
X=zeros(N,1); % array of with N zeros
for k=0:N-1
sum=0;
for n=0:N-1
sum= sum+x(n+1)*e^(-i*2*pi*n*k/N);
end
X(k+1)=sum;
end

%plotting Magnitude of DFT of x

k=0:7;
subplot(3,1,1)
stem(k,abs(X))

%plotting phase of DFT of x

subplot(3,1,2)
stem(k,angle(X))
% calculating DFT by FFT algorithm
X2=fft(x);
% plotting magnitude X2
subplot(3,1,3)
stem(k,abs(X2))
Output/Observation:
X=

4.0000 + 0.0000i
-0.0000 - 0.0000i
0.0000 - 0.0000i
0.0000 - 0.0000i
4.0000 + 0.0000i
-0.0000 - 0.0000i
0.0000 - 0.0000i
-0.0000 - 0.0000i

Plots:
Result:
Exp3: Design IIR filter for the given specifications using impulse invariant
and bilinear transformation technique.
Aim: to design IIR filter for the given specifications using impulse invariant method and
bilinear transformation method

Equipment and Software Requirement:

1. PC: windows OS,
2. Software: MATLAB

Program 1:
% IIR filter by Impulse invariance method
clc;clear all;close all;
fp=input('enter passband edge frequency:');
fs= input('enter stopband edge frequency:');
rp=input('Attenuation at passband edge frequency:');
rs=input('Attenuation at stopband edge frequency:');
FS=input('sampling rate orSampling frequency:');

%calculating Order N, Angular cutoff frequencies

Op=2*pi*fp;
Os=2*pi*fs;
[N,Oc]= buttord(Op,Os,rp,rs, 's');

% Finding Transfer function H(s) coefficient of analog butterworth filter

[b,a] = butter(N,Oc,'s');

% Finding Transfer function H(z) coefficient of digital butterworth filter

[bz,az]=impinvar(b,a,FS);

%Following syntax calculate Frequency response of H(z)

[Hw, f]= freqz(bz,az,500,FS);
% Gain vs frequency plot
gain_abs=abs(Hw);
subplot(2,1,1);
plot(f,gain_abs);
% Gain in dB vs frequency plot
gain_dB=20*log(abs(Hw));
subplot(2,1,2);
plot(f,gain_dB);
Output/Observation:
enter passband edge frequency:500
enter stopband edge frequency:1000
Attenuation at passband edge frequency:3
Attenuation at stopband edge frequency:40
sampling rate orSampling frequency:5000
>> bz

bz =

-0.0000 0.0000 0.0016 0.0056 0.0037 0.0005 0.0000 0

>> az

az =

1.0000 -4.1476 7.7482 -8.3442 5.5547 -2.2747 0.5287 -0.0537

Plots:
Program 2:
%IIR filter design by Bilinear Transformation method
clc;clear all;close all;
fp=input('enter passband edge frequency:');
fs= input('enter stopband edge frequency:');
rp=input('Attenuation at passband edge frequency:');
rs=input('Attenuation at stopband edge frequency:');
FS=input('sampling rate orSampling frequency:');

%calculating Order N, Angular cutoff frequencies

Op=2*FS*tan(pi*fp/FS); % pre warping frequencies
Os=2*FS*tan(pi*fs/FS); % pre warping frequencies
[N,Oc]= buttord(Op,Os,rp,rs, 's');
% Finding Transfer function H(s) coefficient of analog butterworth filter
[b,a] = butter(N,Oc,'s');

% Finding Transfer function H(z) coefficient of digital butterworth filter

[bz,az]=bilinear(b,a,FS); %use this command for bilinear transformation

%Following syntax calculate Frequency response of H(z)

[Hw, f]= freqz(bz,az,500,FS);
% Gain vs frequency plot
gain_abs=abs(Hw);
subplot(2,1,1);
plot(f,gain_abs);
% Gain in dB vs frequency plot
gain_dB=20*log(abs(Hw));
subplot(2,1,2);
plot(f,gain_dB);
Output/Observation:
enter passband edge frequency:500
enter stopband edge frequency:1000
Attenuation at passband edge frequency:3
Attenuation at stopband edge frequency:40
sampling rate orSampling frequency:5000
>> az

az =

1.0000 -3.4944 5.4250 -4.6889 2.3579 -0.6500 0.0764

>> bz
bz =

0.0004 0.0024 0.0061 0.0081 0.0061 0.0024 0.0004

Plots:

Result:
 Exp4: Design of FIR filter for the given specifications using frequency
sampling and windowing technique.

Aim: To design of FIR filter for the given specifications using frequency sampling and
windowing technique.

Equipment and Software Requirement:

1. PC: windows OS,
2. Software: MATLAB

Program:
% MATLAB program for windowing technique.
clc;
close all;
clear all;
e=exp(1);
%duration of FIR filter impulse response
m=input('enter the value of m=');
%cutt off frequency
wc=input('enter the value of wc=');
w=linspace(-pi,pi,1000)';

%Desired Frequency Response

Hd=zeros(length(w),1);
for i=1:length(w)
if(abs(w(i))<=wc)
Hd(i)=1*e^(-1i*w(i)*(m-1)/2);
else
Hd(i)=0;
end
end
%obtaining corresponding impulse response
hd=zeros(50,1);
for n=0:49
hd(n+1)=trapz(w,Hd.*(e.^(1i*w*n))/(2*pi));
end

%window function
wr= [ones(m,1);zeros(50-m,1)];
%wr(1:m,:)=(1-cos(2*pi*(0:m-1)/(m-1)))/2;
wr(1:m,:)=(0.54-0.46*cos(2*pi*(0:m-1)/(m-1)));

%impulse response of FIR filter

h=hd.*wr;
H=zeros(length(w),1);
for i=1:length(w)
sm_H=0;
for n=0:length(h)-1
sm_H=sm_H+h(n+1)*exp(-1i*w(i)*n);
end
H(i)=sm_H;
end

W_wf=zeros(length(w),1);
for i=1:length(w)
sm_W=0;
for n=0:length(wr)-1
sm_W=sm_W+wr(n+1)*exp(-1i*w(i)*n);
end
W_wf(i)=sm_W;
end
subplot(3,1,1);
n=0:49;
stem(n,hd);
title('hd(n)');
subplot(3,1,2);
n=0:49;
stem(n,wr);
title('window function');
subplot(3,1,3);
n=0:49;
stem(n,h)
title('h(n) impulse response of FIR filter');
figure,
subplot(3,1,1);
plot(w,abs(Hd));
title('Desired magnitude response');
subplot(3,1,2);
plot(w,abs(W_wf));
title('Fourier transform of window function');
subplot(3,1,3);
plot(w,abs(H));
title('Magnitude response of FIR filter');

Output/Observation:
enter the value of m=10
enter the value of wc=pi/3
>> real(h(1:m))
ans =
 -0.0707
-0.0455
0.0637
0.2122
0.3183
0.3183
0.2122
0.0637
-0.0455
-0.0707

Plots:
Program 2:
%MATLAB program by frequency sampling method
clc;
close all;
clear all;
e=exp(1);
fs=10000;
m=15;
w=((2*pi)/m)*(0:(m-1)/2);
mHd=[1 1 1 1 0 0 0 0];
Hrw=mHd;
k=0:7;
Gk=zeros(8,1);
for k=0:7
Gk(k+1)=((-1)^k)*Hrw(k+1);
end
Hd= mHd.*(e.^(-1j*w*(m-1)/2));
h=zeros(m,1);
h1=zeros(m,1);
for n=0:m-1
sm=0;
sm1=0;
for k=1:(m-1)/2
sm=sm+(2*real(Hd(k+1)*e.^(1i*(2*pi/m)*k*n)));
sm1=sm1+2*Gk(k+1)*cos(2*pi*k/m*(n+0.5));
end
h(n+1)=(sm+Hd(0+1))*1/m;
h1(n+1)=(sm1+Gk(0+1))/m;

end

subplot(3,1,1);
n=0:length(h)-1;
stem(n,h);
[Hf fr]=freqz(h,1,500,fs);
wfr=2*pi*fr/fs;
subplot(3,1,2);
plot(wfr,abs(Hf));
subplot(3,1,3);
plot(wfr,angle(Hf));
Output/Observation:
h=

-0.0498
0.0412
 0.0667
-0.0365
-0.1079
0.0341
0.3189
0.4667
0.3189
0.0341
-0.1079
-0.0365
0.0667
0.0412
-0.0498

Plots:
Result:
 Exp5: Study and comparison of different non-parametric spectral
estimation techniques
Aim: To compute spectral density of a given signal by using MATLAB program.

Equipment and Software Requirement:

1. PC: windows OS,
2. Software: MATLAB

Program:
clc;
clear all;
close all;
N=20;
A=input('enter the value of A');
B=input('enter the value of B');
n=0:N-1;
x=A*[ones(N/2,1);-1*ones(N/2,1)]+B*randn(N,1);

subplot(2,1,1);
stem(n,x);
[pxx,w]=periodogram(x);
subplot(2,1,2);
stem(w,pxx);
Output/Observation:
enter the value of A10
enter the value of B0.1

Plots:
Result:
 Exp 6: Design of Multirate LPF filters for the given specification
Aim: To design multirate LPF filters for the given specification.
Equipment and Software Requirement:
1. PC: windows OS
2. Software: MATLAB

Program:
clc;clear all; close all;
% Multirate filters with decimator and interpolator
Fs=8000;
n=0:99;
x=10*cos(2*pi*500/Fs*n)+5*cos(2*pi*2000/Fs*n)+0.1*rand(1,100);

%FIR filter specification

Fc=2500; %cutt-off of LPF
Fn=Fs/2;
Wc=Fc/Fn;
N=10; %Order of FIR filter
b_fir=fir1(N,Wc); %designing of FIR filter

% Filtering with downsampling (Decimator)

Ds=4;%downsampling rate
x_filter=filter(b_fir,1,x); %filtering by FIR filter
x_dc=downsample(x_filter, Ds);

% Filtering with upsampling (Interpolator)

Us=4; % upsampling rate
x_up=upsample(x_dc,Us); % upsampling
x_int=filter(b_fir,1,x_up); % filtering
subplot(3,1,1)
stem(n,x);
subplot(3,1,2)
stem(0:24,x_dc);
subplot(3,1,3)
stem(n,x_int);
figure
subplot(3,1,1)
stem(n,abs(fft(x)));
subplot(3,1,2)
stem(0:24,abs(fft(x_dc)));
subplot(3,1,3)
stem(n,abs(fft(x_int)));
Output/Observation:

Output Plots:
Result: