Optimal step-size of pseudo affine projection algorithm

摘要

A new statistical analysis model is used to analyze the pseudo affine projection (PAP) algorithm with the optimal step-size in this paper. By setting the weight error to be zero in the direction of the adaptive weight update, the optimal step-size is obtained. Two assumptions for the direction vector are proposed, which simplify the analysis of the convergence behavior for the optimal step-size PAP algorithm. Under this condition, deterministic recursive equations for the weight error and for the mean-square error are derived in the direction of the adaptive weight update. Simulation results are provided to corroborate the analytical results.