On a class of subband adaptive filters with inherent decorrelating properties

This paper describes and analyzes a class of subband adaptive filters (SAFs) that stems from various approaches of applying subband and multirate techniques in adaptive filtering. This class of SAFs has a unique weight-control mechanism, whereby subband error signals are used to adapt a fullband tap-weight vector. In this paper, we show that the recursive tap-weight adaptation can be expressed as an optimization process which approximates the steepest-descent and Newton´s methods in minimizing two fundamentally different error criteria. Theoretical analysis reveals the inherent decorrelating properties of the adaptive algorithm, which manifest themselves in different forms as integral elements of the weight-control mechanism.