A new variable step-size fractional lower-order moment algorithm for non-Gaussian interference environments

Abstract

A new variable step-size fractional lower-order moment (FLOM) algorithm is proposed for non-Gaussian interference suppression. The Normalized FLOM algorithm is a generalization of the normalized least mean square (NLMS) algorithm with a lower order p ≤ 2 and p = 2 corresponding to the NLMS algorithm. With a fixed step size, the smaller the order p, the faster the convergence of the NFLOM algorithm, but the higher the excess mean square errors (MSE) after convergence. The proposed VSS-FLOM provides a compromise solution to the two conflicting goals of low excess MSE and fast convergence. Simulation results for a system identification application in Gaussian and compound K interference environments show that the proposed VSS-FLOM algorithm outperforms the fixed step-size NFLOM and VSS-LMS algorithms in non-Gaussian interference environments.