An Improved Maximum Correntropy Criterion Algorithm with Flexible Zero Attractor for Sparse Channel Estimation in Mixed Gaussian Noise Environment

A maximum correntropy criterion (MCC) algorithm with a flexible zero attractor is proposed, whose estimation performance is investigated for estimating a multi-path channel in mixed Gaussian environment. The proposed algorithm is denoted as flexible zero attraction MCC (FZA-MCC), in which the flexible zero attractor (FZA) is implemented by an approximation parameter adjustment function. The estimation behavior of the proposed FZA-MCC algorithm is carried out to evaluate a wide bandwidth multi-path channel in mixture Gaussian environment. The obtained results demonstrated that the FZA-MCC algorithm converges faster and achieves lower estimation error than the MCC and its related sparse algorithms.