Dear all,
My question is about estimation I found in many algorithms.
x(n)=b*x(n)+(1-b)*x(n-1);-----> (DCT and DFT LMS, power normalized-NLMS)
and sometimes its:
x(n)=b1*x(n)+(1-b1)*x(n-1) if x(n)>x(n-1);
x(n)=b2*x(n)+(1-b2)*x(n-1);

I'm working in echo cancellation...
I want to know how derive it and the effects on my x(n) signal...
References are welcome...
thank you

On Mar 5, 11:45=A0am, "mmoctar" <mmoc...@gmail.com> wrote:
> Dear all,
> My question is about estimation I found in many algorithms.
> x(n)=3Db*x(n)+(1-b)*x(n-1);-----> (DCT and DFT LMS, power normalized-NLMS=
)
> and sometimes its:
> x(n)=3Db1*x(n)+(1-b1)*x(n-1) if x(n)>x(n-1);
> x(n)=3Db2*x(n)+(1-b2)*x(n-1);
>
> I'm working in echo cancellation...
> I want to know how derive it and the effects on my x(n) signal...
> References are welcome...
> thank you

Adaptive Filter Theory
Simon Haykin:
http://www.amazon.com/Adaptive-Filter-Theory-Simon-Haykin/dp/0130901261/ref=
=3Dpd_bbs_1?ie=3DUTF8&s=3Dbooks&qid=3D1236274152&sr=3D8-1