Diffusion L0-norm constraint improved proportionate LMS algorithm for sparse distributed estimation

TL;DR

This paper introduces the L0-IPLMS algorithm, which combines diffusion proportionate LMS and sparsity constraints to enhance sparse system estimation, demonstrating superior performance in simulations.

Contribution

The paper proposes a novel diffusion L0-norm constrained LMS algorithm that integrates benefits of prior methods for improved sparse system estimation.

Findings

L0-IPLMS outperforms existing diffusion LMS algorithms in simulations.

The proposed method effectively exploits system sparsity.

Simulation results confirm enhanced convergence and accuracy.

Abstract

To exploit the sparsity of the considered system, the diffusion proportionate-type least mean square (PtLMS) algorithms assign different gains to each tap in the convergence stage while the diffusion sparsity-constrained LMS (ScLMS) algorithms pull the components towards zeros in the steady-state stage. In this paper, by minimizing a differentiable cost function that utilizes the Riemannian distance between the updated and previous weight vectors as well as the L0 norm of the weighted updated weight vector, we propose a diffusion L0-norm constraint improved proportionate LMS (L0-IPLMS) algorithm, which combines the benefits of the diffusion PtLMS and diffusion ScLMS algorithms and performs the best performance among them. Simulations in a system identification context confirm the improvement of the proposed algorithm.