Improved Impulsive Noise Suppression Method: Joint Myriad Detection and Gaussian Fitting Robust Local Weighted Smoothing

Impulsive noise is a common impediment in many wireless communication systems, which prevents the system from error-free transmission. To this end, this paper aims to investigate the clipping and robust locally weighted regression (RLOESS) to mitigate the adverse effects of impulsive noise. To improve the impulsive noise suppression performance, a Myriad detection - Gaussian fitting robust local weighted regression smoothing (M-GLOESS) algorithm is designed. The proposed M-GLOESS finds the outliers with the help of the Myriad filter and realizes a better impulsive noise suppression performance by introducing a Gaussian fitting robust correction coefficient diagonal matrix. Simulation results are presented to verify the effectiveness of the proposed M-GLOESS, which is robust to the impulsive noise and has better performance than the traditional algorithms.