IMAC: Impulsive-mitigation adaptive sparse channel estimation based on Gaussian-mixture model

Broadband frequency-selective fading channels usually have the inherent sparse nature. By exploiting the sparsity, adaptive sparse channel estimation (ASCE) methods, e.g., reweighted L1-norm least mean square (RL1-LMS), could bring a performance gain if additive noise satisfying Gaussian assumption. In real communication environments, however, channel estimation performance is often deteriorated by unexpected non-Gaussian noises which include conventional Gaussian noises and impulsive interferences. To design stable communication systems, hence, it is urgent to develop advanced channel estimation methods to remove the impulsive interference and to exploit channel sparsity simultaneously. In this paper, robust impulsive-mitigation adaptive sparse channel estimation (IMAC) method is proposed for solving aforementioned technical issues. Specifically, first of all, the nonGaussian noise model is described by Gaussian mixture model (GMM). Secondly, cost function of reweighted L1-norm penalized least absolute error standard (RL1-LAE) algorithm is constructed. Then, RL1-LAE algorithm is derived for realizing IMAC method. Finally, representative simulation results are provided to corroborate the studies.