New FxLMAT based Algorithms for Active Control of Impulsive Noise

In the presence of non-Gaussian impulsive noise (IN) with a heavy tail, active control (ANC) algorithms often encounter stability problems. While adaptive filters based on higher-order error power principle have shown improved filtering capability compared to least mean square family for IN, however, performance filtered-x absolute third (FxLMAT) algorithm tends degrade under high impulses. To address this issue, paper proposes three modifications enhance FxLMAT IN. improve stability, first alteration i.e. variable step size (VSSFxLMAT)algorithm is suggested that incorporates energy input and signal but has slow convergence. its convergence, second modification filtered x robust normalized (FxRNLMAT) presented still lacks robustness. Therefore, modified RNLMAT (MFxRNLMAT) devised, which relatively stable when encountered noise. With comparable computational complexity, proposed MFxRNLMAT gives better robustness convergence speed than all variants cos hyperbolic algorithm, algorithm.