An efficient algorithm for impulsive active noise control using maximum correntropy with conjugate gradient

Abstract

Active noise control (ANC) has attracted more and more attention because of its low cost, simple structure and excellent noise reduction effect in low-frequency environment. However, the existing algorithms for ANC systems have some nonignorable issues in the impulse noise environment which will lead to a poor noise reduction effect. To solve this problem, a filtered-x maximum correntropy criterion with conjugate gradient (FxMCC-CG) algorithm is proposed in this paper. The modified CG method is used to iteratively update the filter weight coefficients, which significantly improves the convergence speed and robustness of the proposed FxMCC-CG algorithm. At the same time, to enhance the performance of FxMCC-CG algorithm in an impulsive noise environment, we come up with an online adaptive adjustment method of the kernel size of the maximum correntropy criterion (MCC) according to the sample variance of the reference signal. The simulation results prove that the proposed FxMCC-CG algorithm can maintain good performance under various noise environments, and its noise reduction ability and convergence speed are better than some existing algorithms.