How Can I Make Correlation Matrix In Rls Algorithm

See more:

MatLab

Hi every one

I'm simulating RLS algorithm
my code is

w(1)=1;
lambda=0.99;
for i=2:length
    y(i)=x(i)*transpose(w(i-1));%output of filter
   % Correlation_matrix=corr(w(i),'rows','pairwise');
    u(i)=w(i-1)*x(i);
    k(i)=((u(i))/(lambda+transpose(x(i))*u(i)));%kalman
    e(i)=d(i)-y(i);%error
    w(i)=transpose(w(i-1))+k(i)*e(i);%weight of filter
end

In this algorithm we have two signal
1-main signal
2-noise signal
I mixed two signals and get "d" so d=main signal+noise signal and "x" is noise signal
and I should remove or reduce noise from "d" signal by above algorithm
in this line. u(i)=w(i-1)*x(i);
w(i-1) must be correlation matrix I don't know who can I compute it?

thanks in advance