Title :
Enhanced-Convergence Normalized LMS Algorithm[DSP Tips & Tricks]
fDate :
5/1/2009 12:00:00 AM
Abstract :
Least mean square (LMS) algorithms have found great utility in many adaptive filtering applications. This article shows how the traditional constraints placed on the update gain of normalized LMS algorithms are overly restrictive. We present relaxed update gain constraints that significantly improve normalized LMS algorithm convergence speed.
Keywords :
adaptive filters; least mean squares methods; adaptive filtering applications; enhanced-convergence normalized LMS algorithm; least mean square algorithms; Adaptive filters; Algorithm design and analysis; Convergence; Error correction; Filtering algorithms; Gain; Least squares approximation; Signal design; Signal processing algorithms; Upper bound;
Journal_Title :
Signal Processing Magazine, IEEE
DOI :
10.1109/MSP.2009.932168
