A Multi-Sampled Multiband-Structured Subband Filtering Algorithm

Abstract

A multi-sampled multiband-structured sub-band adaptive filtering (MS-MSAF) algorithm is proposed in this paper. The key feature of the proposed algorithm lies in the expansion of the sub-sampled number to a common value. The properties of mean stability and mean-square deviation of the MS-MSAF algorithm are analyzed in the original time domain, and the theoretical analysis also leads to a way to accelerate the convergence speed of the proposed MS-MSAF algorithm. The obtained theoretical results are verified through computer simulations.