Abstract: Offshore biological noise appears as impulse noise. This type of noise significantly reduces the performance of underwater acoustic communication systems. Combined with the sparse characteristics of underwater acoustic channels, a cascade filter algorithm based on maximum correntropy criterion is proposed to realize impulse noise elimination and sparse channel estimation. The first-stage adaptive filter completes noise elimination, and then the second-stage filter performs sparse channel estimation for the signals after denoising. The advantage of this method is no need to know the relevant information about signal and noise. Simulation results show that the proposed method is better than the single filter structure and the traditional channel estimation algorithm.