Title :
Comments on "A Recursive Least M-Estimate Algorithm for Robust Adaptive Filtering in Impulsive Noise: Fast Algorithm and Convergence Performance Analysis
Author :
Bershad, Neil J.
Author_Institution :
Henry Samueli Sch. of Eng., Univ. of California, Irvine, CA
Abstract :
A recent paper [S. C. Chan and Y. X. Zou, "A Recursive Least M-Estimate Algorithm for Robust Adaptive Filtering in Impulsive Noise: Fast Algorithm and Convergence Performance Analysis," IEEE Transactions on Signal Processing, vol. 57, no. 1, Jaunary 2008] studied the behavior of a recursive least M-estimate (RLM) adaptive filtering algorithm in an additive impulsive noise environment. The mean and mean-square behavior of the algorithm was analyzed using a joint Gaussian assumption for the input and the error signal. This note points out that this assumption contradicts the probability model for the impulsive noise [contaminated Gaussian (CG) noise]. Hence, the analytic results presented in Chan and Zou are of limited interest.
Keywords :
Gaussian processes; adaptive filters; convergence of numerical methods; least mean squares methods; probability; recursive estimation; transient response; convergence performance analysis; impulsive noise; joint Gaussian process; mean-square algorithm; probability model; recursive least m-estimate algorithm; robust adaptive filtering; Adaptive filters; Additive noise; Algorithm design and analysis; Convergence; Filtering algorithms; Gaussian noise; Noise robustness; Performance analysis; Signal processing algorithms; Working environment noise;
Journal_Title :
Signal Processing, IEEE Transactions on
DOI :
10.1109/TSP.2008.2007920
