Multichannel Adaptive Active Noise Control Taking Account of Channel Changes

This paper is concerned with adaptive active noise control of a multichannel system which involves a multiple number of primary sources, reference microphones, secondary control sources and canceling error microphones. First, on the assumption that the secondary path response matrices are known a priori, we give a structure of a multi-channel adaptive feedforward controller and propose a stability guaranteed robust adaptive algorithm with variable step size for updating the FIR type of controllers. Secondly, in cases of unknown secondary path matrices, we also present two identification-based adaptive algorithms for realizing controllers preserving the stability of the adaptive system. The effectiveness is examined by comparing the proposed algorithms with ordinary filtered-x algorithm through active noise control experiments.