An active impulsive noise control algorithm with self-p-normalized method

We study methods to overcome the limitation of active noise control (ANC) systems against impulsive noise in this paper. A new filter-x least mean square adaptive algorithm and its sign error edition are proposed. The proposed algorithms use a self-normalization method that combines self-normalization and p-norm based on the presented algorithms. The self-p-normalization method reuses the filtered reference signal to generate a non-linear function to transform the reference signal and takes the p-norm into account to cope with the problem, which is caused by the sparsity of signal power and accelerate the convergence speed. Furthermore, computer simulations and experiments were carried out. The result shows that the proposed self-p-normalized algorithms have a more significant effect on suppressing the over updating weights of the conventional algorithms against impulse. In particular, the proposed algorithms get higher convergence speed without sacrificing the performance of steady-state error. (C) 2021 Elsevier Ltd. All rights reserved.