Switching‐based variable step‐size approach for partial update LMS algorithms

Abstract

A switching-based variable step-size approach for partial update least-mean-square (LMS) algorithms is proposed. During the initial stage, the errors are dominated by the mismatch between the tap weights of the adaptive filter and its ideal weights. In this reported work, the error signals are correlated with the linear interpolated error signals to gain a large step-size; on the other hand, during the steady state, the errors mainly come from the additive noise. Therefore, the correlation is switched to the other mode so that the effect of noise can be eliminated. This can be done by correlating the error signals with a delayed version of the error signals, hence a small step-size is obtained during the late stages. Simulation results show that when only a half of all taps are updated in one iteration, the proposed method significantly enhances the convergence rate, especially for the ‘data independent’ partial update LMS algorithms.