A Subband Adaptive Filtering for Distributed Active Noise Control Systems

This study proposes an extension of the subband adaptive filtering technique to distributed active noise control systems and the details of the algorithm are presented. The computational requirements and convergence performance are examined for a network with a ring topology without constraints in communication. The proposed algorithm divides the signal of each node into several subbands to reduce the dynamic range of the spectrum and consequently accelerate the convergence speed. Since the subband decomposition contains a decimation process, the computational complexity is reduced as the number of subbands is increased. Simulations are conducted to compare different algorithms and an 8-node network is studied in terms of noise reduction performance. The results demonstrate the efficacy of the proposed algorithm for broadband active noise control and the convergence speed could be improved compared to the distributed multiple error filtered-x (DMEFxLMS) algorithm.