LMS convergence criterion

Hi,
I want to perform the convergence stability of my LMS algo. but I don't
know how.
I use the error, the variance of the coefficients and a time (a step time
to compute the variance). I know that the last one isn't very accurate.
Anybody can help me.

Thank you in advance.
Jo



		
This message was sent using the Comp.DSP web interface on
www.DSPRelated.com

For LMS convergence, the update parameter "mu" should be less than
2/lambda_max, where lambda_max is the maximum eigen value of the
autocorrelation matrix of the input.

I suggest that you lookup "Adaptive Filter Theory" by Haykin.

-Philonoist

Johan wrote:
> Hi,
> I want to perform the convergence stability of my LMS algo. but I don't
> know how.
> I use the error, the variance of the coefficients and a time (a step time
> to compute the variance). I know that the last one isn't very accurate.
> Anybody can help me.
>
> Thank you in advance.
> Jo
>
>
>
>
> This message was sent using the Comp.DSP web interface on
> www.DSPRelated.com