Study of Robust Sparsity-Aware RLS algorithms with Jointly-Optimized Parameters for Impulsive Noise Environments

TL;DR

This paper introduces a unified sparsity-aware robust RLS algorithm for sparse system identification in impulsive noise, with joint parameter optimization leading to improved tracking and lower misadjustment.

Contribution

It proposes a generalized S-RRLS algorithm adaptable to various robustness and sparsity criteria, and introduces JO-S-RRLS with jointly optimized parameters for better performance.

Findings

Outperforms existing methods in impulsive noise environments

Exhibits low misadjustment and effective tracking of system changes

Demonstrates superior robustness and sparsity handling in simulations

Abstract

This paper proposes a unified sparsity-aware robust recursive least-squares RLS (S-RRLS) algorithm for the identification of sparse systems under impulsive noise. The proposed algorithm generalizes multiple algorithms only by replacing the specified criterion of robustness and sparsity-aware penalty. Furthermore, by jointly optimizing the forgetting factor and the sparsity penalty parameter, we develop the jointly-optimized S-RRLS (JO-S-RRLS) algorithm, which not only exhibits low misadjustment but also can track well sudden changes of a sparse system. Simulations in impulsive noise scenarios demonstrate that the proposed S-RRLS and JO-S-RRLS algorithms outperform existing techniques.