A New Stochastic Analysis of Affine Projection Algorithm for Gaussian Inputs and Large Number of Coefficients

Sérgio J. M. de Almeida, José Carlos M. Bermudez

DOI: 10.14209/its.2002.471

Evento: 2002 International Telecommunications Symposium (ITS2002)

Keywords:

Abstract

"This paper presents a new analytical model for predicting the behavior of Affine Projection (AP) algorithm. The model is derived for autoregressive (AR) Gaussian inputs and for unity step size (fastest convergence). Deterministic recursive equations are determined for the mean weight and the mean square error behaviors for a large number of adaptive taps N, as compared to the order P of the algorithm. Simulation are presented which show excellent agreement between theory an simulations in steady-state, and fair to good agreement during transient. The model predictions for the transient phase improve as N/P increases. These caracteristics are of special interest in applications such as acoustic echo cancellation."

Download