Stochastic Model for the Mean Weight Evolution of the IAF-PNLMS Algorithm

Abstract

This correspondence studies the adaptive weight evolution of the individual-activation-factor proportionate normalized least-mean-square (IAF-PNLMS) algorithm. For such, the modeling methodology used considers that the gain matrix is time varying and the input signal is not restricted to be white. A model is obtained that predicts the algorithm mean weight behavior for both transient and steady-state phases. Through simulation results, the accuracy of the proposed model is verified. In addition, the approach developed here is general and can be applied to other PNLMS-type algorithms.