Title :
A step-size control for adaptive filters using a sign-based estimation of the normalized excess mean-squared error
Author :
Lippuner, Dani ; Kaelin, August N.
Author_Institution :
Signal & Inf. Process. Lab., Swiss Fed. Inst. of Technol., Zurich, Switzerland
fDate :
6/23/1905 12:00:00 AM
Abstract :
A new estimation measure for the normalized excess MSE (mean-squared error) in adaptive system identification is presented in this paper. This measure, which is based on the signs of the filter update terms, allows one to directly derive an appropriate step size of the adaptive FIR filter. Its usefulness is demonstrated in-a step-size control for the LMS (least-mean-square) algorithm
Keywords :
FIR filters; acoustic signal processing; adaptive filters; echo suppression; least mean squares methods; LMS algorithm; acoustic echo cancellation; adaptive FIR filter; adaptive system identification; estimation measure; filter update terms; normalized excess MSE; sign-based estimation; step-size control; Adaptive filters; Computer errors; Covariance matrix; Finite impulse response filter; Information processing; Laboratories; Least squares approximation; Noise measurement; Signal processing; Size measurement;
Conference_Titel :
Electronics, Circuits and Systems, 2001. ICECS 2001. The 8th IEEE International Conference on
Print_ISBN :
0-7803-7057-0
DOI :
10.1109/ICECS.2001.957423
