A Review of Advances in Subband Adaptive Filtering

Subband adaptive filtering (SAF) generally employs multirate filter banks for signal decomposition and reconstruction. This technique allows for fast convergence and reduced computational complexity through use of the robust least mean squares (LMS) algorithm in acoustic environments. However, the performance of subband adaptive schemes is often degraded by artifacts introduced by the insertion of filter banks in the signal path. Recently, several schemes have been proposed to reduce the effect of one or more of these artifacts. This article presents an overview of the development and trends in this particular area of digital signal processing. Comparison tables are given to assess the performance of the main subband adaptive structures found in literature. It also presents simulation results obtained from some cases of adaptive noise cancellation configurations as examples of these SAF systems. The effect of aliasing insertion is demonstrated for various settings of filterbanks and the influence of filter bank optimization on the performance is depicted as mean square error (MSE) plots.