i writed a code for adaptive channel equalizer with LMS algorithm but i donot recieve a desire answer for learning curve . i want to find problem of program . input signal of channel is a 16-QAM .

iterate=200;

N=1000;

M=8;

mu=0.01;

ms=zeros(N,1);

p=[0.1483;0.2595;-0.0318;-0.1536;-0.2118;0.1059;0.5295;0.7314];

for k=1:iterate

   j = sqrt(-1);

x = sign(randn(N,1)).*(2+sign(randn(N,1)))+ j*sign(randn(N,1)).*(2+sign(randn(N,1)));

    v=sqrt(0.001)*randn(N,1);

    d=filter(p,1,x)+v;

    u=d(1:N);

    w=zeros(M,1);

    e=zeros(N,1);

    y=zeros(N,1);

    uvect=zeros(M,1);

    for i=1:N

        uvect=[u(i);uvect(1:M-1)];

        y(i)=w'*uvect;

        e(i)=x(i)-y(i);

        w=(w+mu*uvect*conj(e(i)));

    end

    ms=ms+abs(e.*e);

end

mse1=ms/iterate;

semilogy(mse1);

ylabel('MSE');

xlabel('iteration number');

title('learning curve LMS');

Answer 1

[hadi zavi]

can anybody help me?