Lab Tasks/Assignments

ADSP Lab

Submitted to: Submitted by:

Mr. Rajesh Mehra Gaurav Soni
Assistant Professor M.E.-Modular-ECE
ECE 2011 Batch

1
 INDEX

S.No. Name of Experiment Page No.

To understand the basics of MATLAB

1 4

To write MATLAB code for random number generation and plot it.
2 7

To write MATLAB code for waveform addition and plot it.

3 8

To write MATLAB code for folding a sequence and plot it.

4 10

To write MATLAB code to plot x=a^n*u(n).

5 11

To write MATLAB code to generate the impulse sequence at n0 samples

6 lying between n1 and n2. 13

To write MATLAB code to generate complex exponential

7 x=exp(complex(sigma,w).*n). 14

To write MATLAB code for FIR filter design.

8 16

To write MATLAB code for IIR Filter design.

9 23

To write MATLAB code to upsample an input sequence by a factor L=3

10 with a frequency of 0.042 Hz. 29

2
 To write MATLAB code to downsample an input sequence by a factor
11 M=3 with a frequency of 0.042 Hz. 30

To write MATLAB code for estimating the cost of filter.

12 31

To write MATLAB code to design an adaptive filter to extract a desired

13 signal from noise corrupted signal by cancelling the noise. 32

ADDITIONAL TASKS/PROGRAMS
14 34

3
EXPERIMENT-1

AIM: To understand the basics of MATLAB

Exercises:

>> a=[1 2 3 4 5 6]

a=

1 2 3 4 5 6

>> b=a+2

b=

3 4 5 6 7 8

>> plot(b)

7.5

6.5

5.5

4.5

3.5

3
1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

4
>> bar(b)

0
1 2 3 4 5 6

>> xlabel('sample #')

>> ylabel('pounds')

5
pounds

0
1 2 3 4 5 6
s a m p le #

5
>> plot(b,'*')

7.5

6.5

5.5

4.5

3.5

3
1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

>> axis([0 10 0 10])

10

0
0 1 2 3 4 5 6 7 8 9 10

6
EXPERIMENT-2

AIM: To write MATLAB code for random number generation and plot it.

Code:

>> n=input('enter the required number of samples:=');

enter the required number of samples:=10
>> x=rand(1,n);
>> stem(x)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0
1 2 3 4 5 6 7 8 9 10

7
EXPERIMENT-3

AIM: To write MATLAB code for waveform addition and plot it.

Code:

>> fs=100;
>> t=[1:100]/fs;
>> a1=input('enter value of amplitude a1:=');
enter value of amplitude a1:=5
>> a2=input('enter value of amplitude a2:=');
enter value of amplitude a2:=10
>> a3=input('enter value of amplitude a3:=');
enter value of amplitude a3:=15
>> s1=a1*sin(2*pi*t*5);
>> s2=a1*sin(2*pi*t*15);
>> s3=a1*sin(2*pi*t*30);
>> subplot(2,2,1);
>> plot(s1);
>> xlabel('sine waveform with frequency=5 hz');
>> ylabel('amplitude');
>> subplot(2,2,2);
>> plot(s2);
>> xlabel('sine waveform with frequency=15 hz');
>> ylabel('amplitude');
>> subplot(2,2,3);
>> plot(s3);
>> xlabel('sine waveform with frequency=30 hz');
>> ylabel('amplitude');
>> subplot(2,2,4);
>> s4=s1+s2+s3;
>> plot(s4);
>> xlabel('addition of three sine waves');
>> ylabel('amplitude');

8
 5 5
amplitude

amplitude
0 0

-5 -5
0 50 100 0 50 100
sine waveform with frequency=5 hz sine waveform with frequency=15 hz

5 20

10
amplitude

amplitude

0 0

-10

-5 -20
0 50 100 0 50 100
sine waveform with frequency=30 hz addition of three sine waves

9
EXPERIMENT-4

AIM: To write MATLAB code for folding a sequence and plot it.

Code:

>> n=[-5:5];
>> x=[10:-1:0];
>> subplot(2,1,1);
>> stem(n,x);
>> title('the original sequence');
>> xlabel('samples n');
>> ylabel('x(n)');
>> y=fliplr(x);
>> subplot(2,1,2);
>> stem(n,y);
>> title('the folded sequence x(n)');
>> xlabel('samples n');
>> ylabel('x(-n)');

the original sequence

10

6
x(n)

0
-5 -4 -3 -2 -1 0 1 2 3 4 5
sam ples n
the folded sequence x(n)
10

5
x(-n)

0
-5 -4 -3 -2 -1 0 1 2 3 4 5
sam ples n

10
EXPERIMENT-5

AIM: To write MATLAB code to plot x=a^n*u(n).

Code:

>> ns=input('enter the ending point of the sequence if it is not from n=0:=');
enter the ending point of the sequence if it is not from n=0:=10

>> ne=input('enter the ending point of the sequence:=');

enter the ending point of the sequence:=20

>> a=input('enter the value of a:=');

enter the value of a:=1

>> n=[ns:ne];
>> u=[(n-ns)>=0];
>> subplot(2,2,1);
>> stem(n,u);
>> title('sample n');
>> ylabel('u(n)');
>> e=a*exp(n);
>> subplot(2,2,2);
>> stem(n,e);
>> title('exponential signal e=(a^n)');
>> xlabel('sample n');
>> ylabel('e=(a^n)');
>> m=u.*e;
>> subplot(2,1,2);
>> stem(n,m);
>> title('multiplied signal m=u*e');
>> xlabel('samples n');
>> ylabel('m=a*exp(n).*u');

11
 n
sample n x 10exponential signal e=(a )
8

1 6

e=(an)
u(n) 0.5
2

0 0
10 15 20 10 15 20
sample n
x 10
8 multiplied signal m=u*e
5

4
m=a*exp(n).*u

0
10 11 12 13 14 15 16 17 18 19 20
samples n

12
EXPERIMENT-6

AIM: To write MATLAB code to generate the impulse sequence at n0 samples

lying between n1 and n2.

Code:

>> ns=input('enter the ending point of the sequence if it is not from n=0:=');
enter the ending point of the sequence if it is not from n=0:=5
>> ne=input('enter the ending point of the sequence:=');
enter the ending point of the sequence:=12
>> n=[ns:ne];
>> u=[(n-ns)>=0];
>> stem(n,u);

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0
5 6 7 8 9 10 11 12

13
EXPERIMENT-7

AIM: To write MATLAB code to generate complex exponential

x=exp(complex(sigma,w).*n)

Code:

>> n=input('enter the value of n:=');

enter the value of n:=5

>> sigma=input('enter the real value sigma:=');

enter the real value sigma:=2

>> w=input('enter the imaginary value w:=');

enter the imaginary value w:=3

>> x=exp(complex(sigma,w).*n)

x=

-1.6733e+004 +1.4324e+004i

>> n=[2 4;1 3]

n=

2 4
1 3

>> x=exp(complex(sigma,w).*n)

x=

1.0e+003 *

0.0524 - 0.0153i 2.5155 - 1.5995i

-0.0073 + 0.0010i -0.3676 + 0.1663i

14
>> plot(x,n)
Warning: Imaginary parts of complex X and/or Y arguments ignored

3.5

2.5

1.5

1
-500 0 500 1000 1500 2000 2500 3000

15
EXPERIMENT-8

AIM: To write MATLAB code for FIR filter design

Code:

a.Design of a low pass filter with cutoff frequency 0.6*pi radians per samples
(with filter order specified)

>> fc=0.6;
>> N=20;
>> Hf=fdesign.lowpass('N,fc',N,fc);
>> hd(1)=design(Hf,'window','window',@hamming);
>> hd(2)=design(Hf,'window','window',{@chebwin,50});
>> hfvt=fvtool(hd);
>> legend(hfvt,'hamming window design','dolph chebyshev window design');

Magnitude Response (dB)

hamming window design

-10
dolph chebyshev window design

-20

-30
Magnitude (dB)

-40

-50

-60

-70

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

16
b.(Without specifying filter order)

>> fp=0.6;
>> fst=0.7725;
>> fc=(fp+fst)/2;
>> Ap=0.01;
>> Ast=90;
>> setspecs(Hf,'fp,fst,Ap,Ast',fp,fst,Ap,Ast);
>> Hd(4)=design(Hf,'Kaiserwin');
>> hfvt=fvtool(Hd(4));

Magnitude Response (dB)

-20

-40
Magnitude (dB)

-60

-80

-100

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

17
c.Optimal Filter Design

>> Hd(5)=design(Hf,'equiripple');
>> hfvt=fvtool(Hd(4:5));
>> legend(hfvt,'Kaiser Window Design,Filter order 68','Equiripple Design,Filter Order 53')
Magnitude Response (dB)

Kaiser Window Design,Filter order 68

Equiripple Design,Filter Order 53
-20

-40
Magnitude (dB)

-60

-80

-100

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

d.Controlling filter parameters

>> N=20;
>> setspecs(Hf,'N,fc,Ap,Ast',N,fc,Ap,Ast);
>> Hd(6)=design(Hf,'equiripple');
>> hfvt=fvtool(Hd(5:6));
>> legend(hfvt,'Equiripple design,53 coefficients','Equiripple design,20 coefficients')

18
 Magnitude Response (dB)

-10

-20
Equiripple design,53 coefficients
Equiripple design,20 coefficients
-30

-40
Magnitude (dB)

-50

-60

-70

-80

-90

-100

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

e.Transition Width Control

>> setspecs(Hf,'N,fp,fst',N,fp,fst);
>> Hd(7)=design(Hf,'Equiripple');
>> measure(Hd(7))

ans =

Sampling Frequency : N/A (normalized frequency)

Passband Edge : 0.6
3-dB Point : 0.65698
6-dB Point : 0.68512
Stopband Edge : 0.7725
Passband Ripple : 0.33005 dB
Stopband Atten. : 34.4345 dB
Transition Width : 0.1725

>> hfvt=fvtool(Hd(5),Hd(7));
>> legend(hfvt,'Equiripple design,53 coefficients','Equiripple design,20 coefficients')

19
 Magnitude Response (dB)

-10

-20
Equiripple design,53 coefficients
Equiripple design,20 coefficients
-30

-40
Magnitude (dB)

-50

-60

-70

-80

-90

-100

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

f.Stop Band Attenuation Control

>> Hd(8)=design(Hf,'Equiripple','Wstop',5);
>> measure(Hd(8))

ans =

Sampling Frequency : N/A (normalized frequency)

Passband Edge : 0.6
3-dB Point : 0.64658
6-dB Point : 0.67277
Stopband Edge : 0.7725
Passband Ripple : 0.50871 dB
Stopband Atten. : 44.628 dB
Transition Width : 0.1725

>> hfvt=fvtool(Hd(7:8));

20
 >> legend(hfvt,'passband weight=1,Stopband weight=1','Passband weight=1,Stopband
weight=5')

Magnitude Response (dB)

-5 passband weight=1,Stopband weight=1

Passband weight=1,Stopband weight=5
-10

-15

-20
Magnitude (dB)

-25

-30

-35

-40

-45

-50

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

g.No control over passband ripples

>> setspecs(Hf,'N,fp,fst,Ast',N,fp,fst,Ast);
>> Hd(9)=design(Hf,'Equiripple');
>> hfvt=fvtool(Hd(8:9));
>> legend(hfvt,'Equiripple design using weights','Equripple design constraining the stopband')

21
 Magnitude Response (dB)

-10

-20
Equiripple design using w eights
Equripple design constraining the stopband
-30

-40
Magnitude (dB)

-50

-60

-70

-80

-90

-100

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

22
EXPERIMENT-9

AIM: To write MATLAB code for IIR Filter design

Code:

a.Butterworth Filter with cutoff frequency 0.4Pi

>> N=8;
>> f3db=0.4;
>> d=fdesign.lowpass('N,f3db',N,f3db);
>> Hbutter=design(d,'butter');
>> hfvt=fvtool([Hbutter]);
Magnitude Response (dB)

-20

-40

-60

-80
Magnitude (dB)

-100

-120

-140

-160

-180

-200

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

b.Chebyshev type 1

>> Ap=0.5;
>> setspecs(d,'N,f3db,Ap',N,f3db,Ap);
>> Hcheby1=design(d,'cheby1');
>> hfvt=fvtool([Hbutter,Hcheby1],'color','white');
>> legend(hfvt,'Butterworth','Chebyshev Type 1');

23
 Magnitude Response (dB)

-50

-100
Magnitude (dB)

-150

Butterworth
Chebyshev Type 1

-200

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

c.Chebyshev type II

>> Ast=80;
>> setspecs(d,'N,f3db,Ast',N,f3db,Ast);
>> Hcheby2=design(d,'Cheby2');
>> set(hfvt,'filters',[Hbutter;Hcheby2]);
>> legend(hfvt,'Butterworth','Chebyshev Type II');

24
 Magnitude Response (dB)

-20

-40

-60
Butterworth
Chebyshev Type II
-80
Magnitude (dB)

-100

-120

-140

-160

-180

-200

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

d.Elliptic Filter

>> setspecs(d,'N,f3db,Ap,Ast',N,f3db,Ap,Ast);
>> Hellip=design(d,'Ellip');
>> set(hfvt,'Filters',[Hbutter,Hcheby1,Hcheby2,Hellip]);
>> axis([0 1 -90 2])
>> legend(hfvt,'Butterworth','Chebyshev Type I','Chebyshev Type II','Elliptic');
>>

25
 Magnitude Response (dB)

-10

-20 Butterw orth

Chebyshev Type I
Chebyshev Type II
-30 Elliptic

-40
Magnitude (dB)

-50

-60

-70

-80

-90
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized Frequency (×π rad/sample)

After Zoom

>> axis([0 0.44 -5 0])

Magnitude Response (dB)

0

-0.5

-1

Butterw orth
Chebyshev Type I
-1.5
Chebyshev Type II
Elliptic

-2
Magnitude (dB)

-2.5

-3

-3.5

-4

-4.5

-5
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Normalized Frequency (×π rad/sample)

26
 e.Group delay analysis

>> set(hfvt,'Analysis','grpdelay');

Group Delay

45

40

35

30
Group delay (in samples)

Butterw orth
25
Chebyshev Type I
Chebyshev Type II

20 Elliptic

15

10

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

f.All IIR Filters

>> fp=0.1;fst=0.3;Ap=1;Ast=60;
>> setspecs(d,'fp,fst,Ap,Ast',fp,fst,Ap,Ast);
>> Hd=design(d,'alliir');
>> set(hfvt,'Filters',Hd,'Analysis','magnitude','Designmask','On');
>> axis([0 1 -70 2])
>> legend(hfvt,'Butterworth','Chebyshev Type I','Chebyshev Type II','Elliptic');

27
 Magnitude Response (dB)

-10

-20

-30 Butterworth
Magnitude (dB)

Chebyshev Type I
Chebyshev Type II
Elliptic
-40

-50

-60

-70
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized Frequency (×π rad/sample)

g.finding order of a filter

>> order(Hd)

ans =

7 5 5 4

28
EXPERIMENT-10

AIM: To write MATLAB code to upsample an input sequence by a factor L=3

with a frequency of 0.042 Hz.

Code:

>> N=input('Input length = ');

Input length = 5
>> L=input('Up sampling Factor = ');
Up sampling Factor = 3
>> fo=input('input signal frequency = ');
input signal frequency = 0.042
>> n=0:N-1;
>> x=sin(2*pi*fo*n);
>> subplot(2,1,1);
>> stem(x);
>> y=upsample(x,3);
>> subplot(2,1,2);
>> stem(y);

0.8

0.6

0.4

0.2

0
1 1.5 2 2.5 3 3.5 4 4.5 5

0.8

0.6

0.4

0.2

0
0 5 10 15

29
EXPERIMENT-11

AIM: To write MATLAB code to downsample an input sequence by a factor

M=3 with a frequency of 0.042 Hz.

Code:

>> N=input('Input length = ');

Input length = 20
>> M=input('Down sampling Factor = ');
Down sampling Factor = 3
>> fo=input('input signal frequency = ');
input signal frequency = 0.042
>> n=0:N-1;
>> x=sin(2*pi*fo*n);
>> subplot(2,1,1);
stem(x)
>> y=downsample(x,3);
>> subplot(2,1,2);
>> stem(y)

0.5

-0.5

-1
0 2 4 6 8 10 12 14 16 18 20

0.5

-0.5

-1
1 2 3 4 5 6 7

30
EXPERIMENT-12

AIM: To write MATLAB code for estimating the cost of filter.

Code:

1.Butterworth Filter

>> N=8;
>> f3db=0.4;
>> d=fdesign.lowpass('N,f3db',N,f3db);
>> Hbutter=design(d,'butter');
>> hfvt=fvtool([Hbutter]);
>> cost(Hbutter)

ans =

Number of Multipliers : 17
Number of Adders : 16
Number of States :8
MultPerInputSample : 17
AddPerInputSample : 16

2.Chebyshev-1

>> Ap=0.5;
>> setspecs(d,'N,f3db,Ap',N,f3db,Ap);
>> Hcheby1=design(d,'cheby1');
>> hfvt=fvtool([Hbutter,Hcheby1],'color','white');
>> cost(Hcheby1)

ans =

Number of Multipliers : 17
Number of Adders : 16
Number of States :8
MultPerInputSample : 17
AddPerInputSample : 16

31
EXPERIMENT-13

AIM: To write MATLAB code to design an adaptive filter to extract a desired

signal from noise corrupted signal by cancelling the noise.

Code:

>> n=(1:1000)';
>> s=sin(0.075*pi*n);
>> v=0.8*randn(1000,1);
>> ar=[1,0.5];
>> v1=filter(1,ar,v);
>> x=s+v1;
>> ma=[1,-0.8,0.4,-0.2];
>> v2=filter(ma,1,v);
>> L=7;
>> hlms=adaptfilt.lms(7);
>> hnlms=adaptfilt.nlms(7);
>> [mumaxlms,mumaxmselms]=maxstep(hlms,x);
Warning: Step size is not in the range 0 < mu < mumaxmse/2:
Erratic behavior might result.
> In adaptfilt.lms.maxstep at 32
>> [mumaxlms,mumaxmsenlms]=maxstep(hnlms);
>> [ylms,elms]=filter(hlms,v2,x);
>> [ynlms,enlms]=filter(hnlms,v2,x);
>> bw=firwiener(L-1,v2,x);
>> yw=filter(bw,1,v2);
>> ew=x-yw;
>> plot(n(900:end),[ew(900:end),elms(900:end),enlms(900:end)]);
>> legend('wiener filter denoised sinusoid','NLMS denoised sinusoid','LMS
denoised sinusoid');
>> xlabel('time index (n)');
>> ylabel('amplitude');
>> [bw.' hlms.coefficients.' hnlms.coefficients.']

ans =

1.0131 0.8855 1.0347

0.3293 0.2237 0.2358
0.1321 0.1693 -0.1512
0.0811 0.0463 -0.0769
0.1745 0.2385 0.2180
0.1293 0.1568 0.1796
0.0647 0.0926 0.2457

32
 4
wiener filter denoised sinusoid
NLMS denoised sinusoid
LMS denoised sinusoid
3

1
amplitude

-1

-2

-3

-4
900 910 920 930 940 950 960 970 980 990 1000
time index (n)

33
ADD ON PROGRAMS

Matlab Basics

>>t=linspace(1,10,10)

t=

1 2 3 4 5 6 7 8 9 10

>> sin(t)

ans =

Columns 1 through 9

0.8415 0.9093 0.1411 -0.7568 -0.9589 -0.2794 0.6570 0.9894

0.4121

Column 10

-0.5440

>> x=t.*sin(t)

x=

Columns 1 through 9

0.8415 1.8186 0.4234 -3.0272 -4.7946 -1.6765 4.5989 7.9149

3.7091

Column 10

-5.4402

>> t.^2

ans =

1 4 9 16 25 36 49 64 81 100

34
>> z=sin(t.^2)

z=

Columns 1 through 9

0.8415 -0.7568 0.4121 -0.2879 -0.1324 -0.9918 -0.9538 0.9200 -

0.6299

Column 10

-0.5064

>> a=[1,2;3,4]

a=

1 2
3 4

>> nthroot(a,3)

ans =

1.0000 1.2599
1.4422 1.5874

35
Program to plot sine wave

x=linspace(0,2*pi,100);
>> y=sin(x);
>> plot(x,y)
>> title('plot created by gaurav')

plot created by gaurav

1

0.8

0.6

0.4

0.2

-0.2

-0.4

-0.6

-0.8

-1
0 1 2 3 4 5 6 7

36
Generation of delayed sine wave

>> t=0:pi/100:2*pi;
>> a=1;
>> y=a*sin(t);
>> y1=a*sin(t-0.25);
>> y1=a*sin(t-0.5);
>> y1=a*sin(t-0.25);
>> y2=a*sin(t-0.5);
>> y3=a*sin(t+0.25);
>> y4=a*sin(t+0.5);
>> plot(t,y,t,y1,t,y2,t,y3,t,y4);
>> grid on;
>> xlabel('time');
>> ylabel('amplitude');

delayed sine waves

1

0.8

0.6

0.4

0.2
amplitude

-0.2

-0.4

-0.6

-0.8

-1
0 1 2 3 4 5 6 7
time

37
FIR Filter

a. Design of a low pass filter with cutoff frequency 0.4*pi radians per samples
(with filter order specified)

>> fc=0.4;
>> N=25;
>> Hf=fdesign.lowpass('N,fc',N,fc);
>> hd(1)=design(Hf,'window','window',@hamming);
>> hd(2)=design(Hf,'window','window',{@chebwin,50});
>> hfvt=fvtool(hd);
>> legend(hfvt,'hamming window design','dolph chebyshev window design');

Magnitude Response (dB)

-10
hamming w indow design
dolph chebyshev w indow design
-20

-30
Magnitude (dB)

-40

-50

-60

-70

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

38
b.Without specifying filter order

>> fp=0.3;
>> fst=0.3725;
>> fc=(fp+fst)/2;
>> Ap=0.05;
>> Ast=100;
>> setspecs(Hf,'fp,fst,Ap,Ast',fp,fst,Ap,Ast);
>> Hd(4)=design(Hf,'Kaiserwin');
>> hfvt=fvtool(Hd(4));

Magnitude Response (dB)

-20

-40

-60
Magnitude (dB)

-80

-100

-120

-140

-160
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Normalized Frequency (×π rad/sample)

39
c.Optimal Filter Design

>> Hd(5)=design(Hf,'equiripple');
>> hfvt=fvtool(Hd(4:5));
>> legend(hfvt,'Kaiser Window Design,Filter order 68','Equiripple Design,Filter
Order 53')

Magnitude Response (dB)

-20
Kaiser Window Design,Filter order 68
Equiripple Design,Filter Order 53
-40

-60
Magnitude (dB)

-80

-100

-120

-140

-160
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Normalized Frequency (×π rad/sample)

40
d.Controlling filter parameters

>> N=10;
>> setspecs(Hf,'N,fc,Ap,Ast',N,fc,Ap,Ast);
>> Hd(6)=design(Hf,'equiripple');
>> hfvt=fvtool(Hd(5:6));
>> legend(hfvt,'Equiripple design,53 coefficients','Equiripple design,20
coefficients')

Magnitude Response (dB)

0 Equiripple design,53 coefficients

Equiripple design,20 coefficients

-20

-40
Magnitude (dB)

-60

-80

-100

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

41
e.Transition Width Control

>> setspecs(Hf,'N,fp,fst',N,fp,fst);
>> Hd(7)=design(Hf,'Equiripple');
>> measure(Hd(7))

ans =

Sampling Frequency : N/A (normalized frequency)

Passband Edge : 0.3
3-dB Point : 0.30949
6-dB Point : 0.33612
Stopband Edge : 0.3725
Passband Ripple : 3.9308 dB
Stopband Atten. : 13.0541 dB
Transition Width : 0.0725

>> hfvt=fvtool(Hd(5),Hd(7));
>> legend(hfvt,'Equiripple design,53 coefficients','Equiripple design,20
coefficients')

Magnitude Response (dB)

0 Equiripple design,53 coefficients

Equiripple design,20 coefficients

-20

-40
Magnitude (dB)

-60

-80

-100

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

42
f.Stop Band Attenuation Control

>> Hd(8)=design(Hf,'Equiripple','Wstop',50);
>> measure(Hd(8))

ans =

Sampling Frequency : N/A (normalized frequency)

Passband Edge : 0.3
3-dB Point : 0.20826
6-dB Point : 0.23965
Stopband Edge : 0.3725
Passband Ripple : 19.2864 dB
Stopband Atten. : 35.8683 dB
Transition Width : 0.0725

>> hfvt=fvtool(Hd(7:8));
>> legend(hfvt,'passband weight=1,Stopband weight=1','Passband
weight=1,Stopband weight=50')

Magnitude Response (dB)

0
passband w eight=1,Stopband w eight=1
Passband w eight=1,Stopband w eight=50

-10
Magnitude (dB)

-20

-30

-40

-50

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

43
g.No control over passband ripples

>> setspecs(Hf,'N,fp,fst,Ast',N,fp,fst,Ast);
>> Hd(9)=design(Hf,'Equiripple');
>> hfvt=fvtool(Hd(8:9));
>> legend(hfvt,'Equiripple design using weights','Equripple design constraining the
stopband')

Magnitude Response (dB)

Equiripple design using w eights

-20 Equripple design constraining the stopband

-40
Magnitude (dB)

-60

-80

-100

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

44
IIR Filters

a.Butterworth Filter with cutoff frequency 0.8Pi

>> N=80;
>> f3db=0.8;
>> d=fdesign.lowpass('N,f3db',N,f3db);
>> Hbutter=design(d,'butter');
>>hfvt=fvtool([Hbutter]);

Magnitude Response (dB)

-100

-200

-300
Magnitude (dB)

-400

-500

-600

-700

-800
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Normalized Frequency (×π rad/sample)

45
b.Chebyshev type 1

>> Ap=50;
>> setspecs(d,'N,f3db,Ap',N,f3db,Ap);
>> Hcheby1=design(d,'cheby1');
hfvt=fvtool([Hbutter,Hcheby1],'color','white');
legend(hfvt,'Butterworth','Chebyshev Type 1');

Magnitude Response (dB)

-200

Butterw orth
-400
Chebyshev Type 1
Magnitude (dB)

-600

-800

-1000

-1200

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

46
c.Chebyshev type II

>> Ast=80;
>> setspecs(d,'N,f3db,Ast',N,f3db,Ast);
>> Hcheby2=design(d,'Cheby2');
>> set(hfvt,'filters',[Hbutter;Hcheby2]);
>> legend(hfvt,'Butterworth','Chebyshev Type II');

Magnitude Response (dB)

-100

-200
Butterw orth
Chebyshev Type II
-300
Magnitude (dB)

-400

-500

-600

-700

-800
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Normalized Frequency (×π rad/sample)

47
d.Elliptic Filter

>> setspecs(d,'N,f3db,Ap,Ast',N,f3db,Ap,Ast);
>> Hellip=design(d,'Ellip');
??? Error using ==> feval
Error using ==> fdesign.abstracttype.ellip at 19
The filter must be stable.

>> Ast=800;
>> setspecs(d,'N,f3db,Ast',N,f3db,Ast);
>> Hcheby2=design(d,'Cheby2');
set(hfvt,'filters',[Hbutter;Hcheby2]);
legend(hfvt,'Butterworth','Chebyshev Type II');

Magnitude Response (dB)

-100

-200
Butterw orth
Chebyshev Type II
-300
Magnitude (dB)

-400

-500

-600

-700

-800

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized Frequency (×π rad/sample)

>> setspecs(d,'N,f3db,Ap,Ast',N,f3db,Ap,Ast);
>> Hellip=design(d,'Ellip');
??? Error using ==> feval
Error using ==> fdesign.abstracttype.ellip at 19
Frequency value from the prototype filter must be between 0 and 1.

48
MATLAB code to design an adaptive filter to extract a desired signal from
noise corrupted signal by cancelling the noise.

1. With ar=[1,0.8];

>> n=(1:1000)';
>> s=sin(0.075*pi*n);
>> v=0.8*randn(1000,1);
>> ar=[1,0.8];
>> v1=filter(1,ar,v);
>> x=s+v1;
>> ma=[1,-0.8,0.4,-0.2];
>> v2=filter(ma,1,v);
>> L=7;
>> hlms=adaptfilt.lms(7);
>> hnlms=adaptfilt.nlms(7);
>> [mumaxlms,mumaxmselms]=maxstep(hlms,x);

Warning: Step size is not in the range 0 < mu < mumaxmse/2:

Erratic behavior might result.
> In adaptfilt.lms.maxstep at 32

>> [mumaxlms,mumaxmsenlms]=maxstep(hnlms);
>> [ylms,elms]=filter(hlms,v2,x);
>> [ynlms,enlms]=filter(hnlms,v2,x);
>> bw=firwiener(L-1,v2,x);
>> yw=filter(bw,1,v2);
>> ew=x-yw;
>> plot(n(900:end),[ew(900:end),elms(900:end),enlms(900:end)]);
>> legend('wiener filter denoised sinusoid','NLMS denoised sinusoid','LMS
denoised sinusoid');
>> xlabel('time index (n)');
>> ylabel('amplitude');
>> [bw.' hlms.coefficients.' hnlms.coefficients.']

ans =

1.0074 0.8684 1.0036

0.0384 -0.0256 -0.0044
0.2879 0.2646 -0.0426
-0.0626 -0.0775 -0.1927
0.2824 0.3048 0.2858
0.0363 0.1388 0.2067
0.2160 0.1952 0.3369

49
 3
wiener filter denoised sinusoid
NLMS denoised sinusoid
LMS denoised sinusoid

0
amplitude

-1

-2

-3

-4
900 910 920 930 940 950 960 970 980 990 1000
time index (n)

2. With ar=[1,0.2];

>> n=(1:1000)';
>> s=sin(0.075*pi*n);
>> v=0.8*randn(1000,1);
>> ar=[1,0.2];
>> v1=filter(1,ar,v);
>> x=s+v1;
>> ma=[1,-0.8,0.4,-0.2];
>> v2=filter(ma,1,v);
>> L=7;
>> hlms=adaptfilt.lms(7);
>> hnlms=adaptfilt.nlms(7);
>> [mumaxlms,mumaxmselms]=maxstep(hlms,x);
Warning: Step size is not in the range 0 < mu < mumaxmse/2:
Erratic behavior might result.
> In adaptfilt.lms.maxstep at 32
>> [mumaxlms,mumaxmsenlms]=maxstep(hnlms);
>> [ylms,elms]=filter(hlms,v2,x);
>> [ynlms,enlms]=filter(hnlms,v2,x);
>> bw=firwiener(L-1,v2,x);

50
>> yw=filter(bw,1,v2);
>> ew=x-yw;
>> plot(n(900:end),[ew(900:end),elms(900:end),enlms(900:end)]);
>> legend('wiener filter denoised sinusoid','LMS denoised sinusoid','NLMS
denoised sinusoid');
>> xlabel('time index (n)');
>> ylabel('amplitude');
>> [bw.' hlms.coefficients.' hnlms.coefficients.']

ans =

1.0020 0.7985 0.7853

0.6063 -0.9678 0.4494
0.1318 -1.9994 -0.1909
0.0638 -1.7650 -0.0177
0.1330 -0.1956 -0.1847
0.1178 0.8161 -0.3353
0.0485 0.6470 -0.0202

50
wiener filter denoised sinusoid
LMS denoised sinusoid
NLMS denoised sinusoid

40

30
amplitude

20

10

-10
900 910 920 930 940 950 960 970 980 990 1000
time index (n)

51
3. With ar=[1,0.5];

>> n=(1:1000)';
>> s=sin(0.075*pi*n);
>> v=0.8*randn(1000,1);
>> ar=[1,0.5];
>> v1=filter(1,ar,v);
>> x=s+v1;
>> ma=[1,-0.8,0.4,-0.2];
>> v2=filter(ma,1,v);
>> L=7;
>> hlms=adaptfilt.lms(7);
>> hnlms=adaptfilt.nlms(7);
>> [mumaxlms,mumaxmselms]=maxstep(hlms,x);
Warning: Step size is not in the range 0 < mu < mumaxmse/2:
Erratic behavior might result.
> In adaptfilt.lms.maxstep at 32
>> [mumaxlms,mumaxmsenlms]=maxstep(hnlms);
>> [ylms,elms]=filter(hlms,v2,x);
>> [ynlms,enlms]=filter(hnlms,v2,x);
>> bw=firwiener(L-1,v2,x);
>> yw=filter(bw,1,v2);
>> ew=x-yw;
>> plot(n(900:end),[ew(900:end),elms(900:end),enlms(900:end)]);
>> legend('wiener filter denoised sinusoid','LMS denoised sinusoid','NLMS
denoised sinusoid');
>> xlabel('time index (n)');
>> ylabel('amplitude');
>> [bw.' hlms.coefficients.' hnlms.coefficients.']

ans =

0.9975 0.7548 0.6196

0.2983 0.1452 -0.3585
0.0930 -0.2884 -0.1028
0.0354 -0.4093 -0.4651
0.1215 -0.0258 -0.1977
0.0771 0.3733 0.0438
0.0333 0.1736 0.0575

52
 20
wiener filter denoised sinusoid
LMS denoised sinusoid
NLMS denoised sinusoid

10

0
amplitude

-10

-20

-30

-40
900 910 920 930 940 950 960 970 980 990 1000
time index (n)

53