Efficient input-selective affine projection sign algorithm with enhanced convergence performance

2026-05-23

PRE

Bum Yong Park

Yoo, Jinwoo

Lee, Won Il

Shin, Jaewook *

DOI：10.1016/j.sigpro.2026.110599

Original

Original request for help

Save

Abstract

En 中文

This study introduces an enhanced affine projection sign algorithm that incorporates an input vector selection mechanism to achieve faster convergence. The proposed scheme selectively determines the contribution of each input vector based on a dual-threshold strategy derived from M-estimation and mean square deviation analysis. An efficient recursive formulation is further adopted to reduce computational effort without sacrificing accuracy. Simulation results demonstrate that the proposed approach achieves faster convergence and improved robustness compared to existing methods under impulsive noise environments.

Keywords:

Affine projection sign algorithm

Mean-square deviation

M-estimate

Input vector selection

Impulsive noise

Efficient input-selective affine projection sign algorithm with enhanced convergence performance

Abstract

Journal

Organization

Cited Papers

Citing Papers