pith. sign in

arxiv: 2606.08225 · v1 · pith:IKQQ54DOnew · submitted 2026-06-06 · 📡 eess.SP

A Double Proportionate Sparse Adaptive Filter for Impulsive Noise Environments

classification 📡 eess.SP
keywords impulsiveadaptivenoisesparsealgorithmsdoubledp-saffilter
0
0 comments X
read the original abstract

Sparse adaptive filters and impulsive noise robust algorithms have largely been developed along separate tracks, leaving a gap when both properties are needed simultaneously. This letter proposes the double proportionate sparse adaptive filter (DP-SAF), which closes this gap within a single $\mathcal{O}(M)$ update. Two independent diagonal gain matrices are introduced; one scales the adaptation step proportionately to coefficient magnitudes, and the other applies a magnitude-dependent zero-attraction that is strongest for inactive taps. A sign-error update provides robustness against impulsive corruptions. Both gain matrices are derived from a minimum-norm optimization framework. Simulations under a Bernoulli impulsive noise model show that DP-SAF consistently achieves a better steady-state MSD than the competing algorithms while matching or exceeding their convergence speeds.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.