Set-membership NLMS algorithm based on bias-compensated and regression noise variance estimation for noisy inputs

TL;DR

This paper introduces a bias-compensated set-membership NLMS algorithm that effectively reduces input noise effects using a novel noise variance estimation method, improving system identification accuracy.

Contribution

The paper presents a new bias-compensated set-membership NLMS algorithm with an innovative noise variance estimation technique for noisy input scenarios.

Findings

Low misalignment in system identification with noisy inputs

Effective noise mitigation through bias compensation

Improved robustness over traditional NLMS algorithms

Abstract

The bias-compensated set-membership normalised LMS (BCSMNLMS) algorithm is proposed based on the concept of set-membership filtering, which incorporates the bias-compensation technique to mitigate the negative effect of noisy inputs. Moreover, an efficient regression noise variance estimation method is developed by taking the iterative-shrinkage method. Simulations in the context of system identification demonstrate that the misalignment of the proposed BCSM-NLMS algorithm is low for noisy inputs.