Title :
On the convergence analysis of the transform domain normalized LMS and related M-estimate algorithms
Author :
Chan, S.C. ; Zhou, Y.
Author_Institution :
Dept. of Electr. & Electron. Eng., Univ. of Hong Kong, Hong Kong
fDate :
Nov. 30 2008-Dec. 3 2008
Abstract :
In this paper, we study the convergence performance of the transform domain normalized least mean square (TDNLMS) algorithm and its robust version, the TD normalized least mean M-estimate (TDNLMM) algorithm, which is derived from robust M-estimation and has the improved performance over their conventional TDNLMS counterpart in impulsive noise environment. Using the Pricepsilas theorem and its extension, and by introducing new special integral functions, related expectations can be evaluated so as to obtain decoupled difference equations describing the mean and mean square behaviors of these algorithms. The analytical results are in good agreement with computer simulation results.
Keywords :
Gaussian noise; adaptive filters; convergence of numerical methods; difference equations; integral equations; least mean squares methods; Pricepsilas theorem; TD normalized least mean M-estimate algorithm; convergence analysis; decoupled difference equations; integral functions; mean square behaviors; transform domain normalized least mean square; Adaptive filters; Additive noise; Algorithm design and analysis; Convergence; Discrete Fourier transforms; Discrete wavelet transforms; Least squares approximation; Noise robustness; Performance analysis; Working environment noise;
Conference_Titel :
Circuits and Systems, 2008. APCCAS 2008. IEEE Asia Pacific Conference on
Conference_Location :
Macao
Print_ISBN :
978-1-4244-2341-5
Electronic_ISBN :
978-1-4244-2342-2
DOI :
10.1109/APCCAS.2008.4745996
