X-LMS to generate CSLMS algorithm

[rira18]

Hi,

I'm interested to generate the Constrained Stability Least Mean Square (CSLMS) algorithm for ECG application. It's possible to use X-LMS to generate CSLMS algorithm? or I have to use Matlab? if it's the case please if someone can help me to obtain the algorithm, I deeply appreciate it because I didn't know waht to do ( I'm a beginner in the processing signal)

Thank you very much

Best regards

Rira

[Blayne_K]

Hello Rira,

I don't believe labview has any explicitly named CSLMS functions, but there are a great deal of pre-written signal analysis functions and advanced fitting algorithms. Do you have any idea of what the algorithm foundations of the Constrained Stability Least Mean Square algorithm is based upon? In most cases one can build up such functionality from subvi components if one has the time to do so.

Best, 

Blayne Kettlewell

[rira18]

Hi,

Thanks a lot to help me, I really appreciate it

The algorithm isn't complicated and it's more performan than the LMS and NLMS, the main change is the coefficient of the filter (weights) w(n+1)

The one relative to LMS is :    w(n+1)= w(n) +µ x(n) e(n) ).

To know more information about it you can see the attached file below (one page) which explain the difference between the algorithm related Filters.

Thanks

Best regards

Rira

[Blayne_K]

Hello Rira,

The best that I could suggest to you is to look through NI's curve fitting VI's whereabouts you could perform a linear, polynomial, ect fit and adjust the method to either least square, or least absolute residual and then programmatically set the weight inputs defined by your control system. If these VI's do not meet your specific need you could try using the formula node in Labview and write out the mathematical expressions that you would like to have performed on your data and run that in your VI to perform the XLMS functionality.

Regards,

Blayne Kettlewell