Active Coefficient Detection Maximum Correntropy Criterion Algorithm for Sparse Channel Estimation Under Non-Gaussian Environments

In this paper, a kind of active coefficient detection (ACD)-based maximum correntropy criterion (MCC) algorithm is proposed to estimate a sparse multi-path channel under the non-Gaussian environments. The proposed ACD-based MCC algorithms are realized by developing an active coefficient detection mechanism, which can distinguish the active taps within the sparse channels and find out the position and the number of active taps. Therefore, only the active taps coefficient is updated while the trivial channel coefficients are set to be zeros. Various computer simulation experiments are carried out to investigate the performance of the proposed ACD-based MCC algorithms under different impulsive noises. The achieved simulation results prove that the proposed ACD-based MCC algorithms are effective and outperform the previous adaptive filtering algorithms for the sparse channel estimation with regard to both the convergence and the estimation error.