main.asv

A simple code for understanding lms nlms and wiener filters for beginners

clear all
N=2;
u = rand(3000,1);
W = [1 2]';
d = filter(flipud(W),1,u);

[row,col]=size(u);
r = xcorr(u,N);                         %  Calculate (k)th autocorrelation and accumulate
p = xcorr(d,u,N);                       %  Calculate (k)th cross correlation and accumulate

R = toeplitz(r(N+1:2*(N+1)-1))/(row*col)  %  (L x L) input autocorrelation matrix
P = p(N+1:2*(N+1)-1)/(row*col) 

winit=[0;0];
Jmin=mean(d.*d) - p'*inv(R)*p;
mu=0.01;
wo=R\p;

w1=[(wo(1)-7):0.1:(wo(1)+7)];
w2=[(wo(2)-7):0.1:(wo(2)+7)];

Jm = jmat(w1,w2,R,p,Jmin);
[w,J]=sd(mu,winit,N,R,p,Jmin);

[e,wm]=lms(mu,N,u,d);

figure(2)
contour(w1,w2,Jm',50);xlabel('w1');ylabel('w2');
hold on
plot(w(1,:),w(2,:),'-O');grid
plot(wm(1,:),wm(2,:),'-x');grid
hold off

%figure(2)
%plot(J);grid on;xlabel('No. of Itterations');ylabel('MSE');title('Steepest Descent');


%figure(3)
%plot(Jn);xlabel('No. of Itterations');ylabel('MSE');title('Newtons Method');