Abstract: Aiming at the problems of weak enengy of microseismic signals and low SNR a filtering algorithm based on Hankel matrix is introduced into microseismic data processing,it can attenuate random noise in microseismic data,and improve the picking accuracy of arrival of microseismic events;characteristic values of signals are automatically selected by using Akaike information criterion (AIC),so it can be applied to real time auto-processing of microseismic monitoring.The filtering method has better fidelity of the signal while attenuating the random noise,compared with the traditional methods.Synthetic data and real field data verify the validity and practicability of the method.