Subband Affine Projection Algorithm Using Variable Step Size

Abstract

In applications with highly correlated inputs or long length of filter, adaptive filters suffer from slow convergence and large steady-state error. Affine projection algorithms based on the subband structure and step size controlling are good solutions for these problems. In this paper, we propose a new subband affine projection algorithm with variable step size. Experimental results on highly correlated inputs produce faster convergence, lower misadjustment error, and smaller complexity than conventional methods based on the fullband structure.