A modified FXLMS algorithm for active impulsive noise control

Nowadays, numerous approaches have been introduced in literature for active noise control (ANC) systems. Surveys stated that filtered - x - Least Mean Square (FxLMS) algorithm appears best choice for controller in adaptive filter. The researchers have the tendency to improve its performance by enhancing the existing algorithm modification of step size as well as the structure of the existing method In this paper, a soft threshold based FxLMS algorithm for impulsive noise is proposed. Here threshold selection is based on Stein Unbiased Risk Estimation (SURE). This new algorithm improves the stability by fast convergence in presence of heavy tail impulsive noise. To demonstrate the performance of proposed method, simulation is conducted using MATLAB compared with Akthar´s algorithm.