M-Estimate Based Normalized Subband Adaptive Filter Algorithm: Performance Analysis and Improvements

Abstract

This paper studies the mean and mean-square behaviors of the M-estimate based normalized subband adaptive filter algorithm (M-NSAF) with robustness against impulsive noise. Based on the contaminated-Gaussian noise model, the stability condition, transient and steady-state results for the algorithm are formulated analytically. The analysis results help us to better understand the M-NSAF performance in impulsive noise. To further obtain fast convergence and low steady-state estimation error, we derive a variable step size (VSS) M-NSAF algorithm. This VSS scheme is also generalized to proportionate M-NSAF variant for sparse systems. Computer simulations on system identification in impulsive noise and acoustic echo cancellation with double-talk are performed to demonstrate our theoretical analysis and the effectiveness of the proposed algorithms.

Study of Proximal Normalized Subband Adaptive Algorithm for Acoustic Echo Cancellation

“…9 by using a realistic speech as the input signal. In this scenario, we use the echo return loss enhancement (ERLE) as a performance index [79], defined as ERLE(n) = 10 log 10 (avg{d…”

A Novel Normalized Subband Adaptive Filter Algorithm Based on the Joint-Optimization Scheme

“…However, despite the lower computational complexity than the AP algorithm, the NSAF algorithms have better performance in terms of tracking speed owing to their self-whitening property. Moreover, some variants of the NSAF, such as sign subband adaptive filter (SSAF) and its variants [10]- [12], M-estimate NSAF [13], [14], sparsity-aware SSAF and NSAF [15], [16], and bias-compensated NSAF [17]- [19], have been proposed to improve performance.…”

Robust Recursive Least M-Estimate Adaptive Filter for the Identification of Low-Rank Acoustic Systems

“…It has demonstrated advantages of identifying long acoustic echo paths in terms of better tracking performance in white Gaussian noise and lower computational complexity. This algorithm, however, is sensitive to non-Gaussian noise that is common in acoustic environments, such as keyboard clicking sounds, noise yielded from footsteps, telephone ringing, door slams, objects dropping, firecrackers, and impulse-like noise produced by detection failures of doubletalk in acoustic echo cancellation, to name but a few [11]- [13].…”