Abstract:
In recent years, a lot of work has been done in the Active Noise Cancellation (ANC) system. Most researchers have worked on the Filtered x Least Mean Square (FxLMS) algorithm, which is extensively applied in ANC, due to its easy implementation. But FxLMS algorithm exhibit slow convergence and becomes in stable in presence of Impulsive noise. Therefore, many different variants of FxLMS have been reported in literature. In this thesis, we have investigated an alternative algorithm i.e; Least Mean Absolute Third (LMAT) algorithm in ANC. Furthermore, two variants of LMAT (Least Mean Absolute Third) with a data reuse (DR) algorithm for impulsive noise are suggested to mitigate the impulsive noise with and without Online Secondary Path Estimation (OSPM). Initially, the proposed algorithms i.e, the Data Reuse- Filtered x Robust Normalized LMAT (DR-FxRNLMAT) and Data Reuse- Normalized Step-Size Filtered x Robust Normalized LMAT (DR-NSSFxRNLMAT) are tested for high impulsive noise cancelation without OSPM. The proposed DR-NSSFxRNLMAT algorithm outperforms the existing algorithms as depicted by the rigorous simulations. Moreover, a detailed comparison of the proposed algorithms with already existing state-of-the-art techniques is also carried out. Active noise control algorithms have stability issues due to offline modeling of the secondary path (OSPM). Therefore, we have tested DR-Normalized Step-Size LMS (DR-NSSLMS) already reported in the literature for the first time in ANC domain with OSPM and our proposed DR-NSSFxRNLMAT with OSPM. The superior performance of the proposed DR-NSSFxRNLMAT algorithm has been validated regarding the fastest convergence, reduced steady-state error, and enhanced robustness of the ANC system with OSPM for impulsive noise.
