Abstract: The shallow sea active sonar detection performance is mainly affected by multi-path expansion and reverberation interference.This paper deduces the dual extended channel matching method based on the channel grid model and proposes a channel matching detection algorithm.The proposed algorithm uses the Least Squares-Matching Pursuit(LS-MP) algorithm to estimate the channel sparse response and removes the redundant part according to the different distribution of reverberation and target channel.Then,copy signal is reconstructed through the estimated channel response,and utilized to calculate correlation with the received signal to detect the echo.Numerical simulation results show that the proposed algorithm can effectively suppress the dual-extended channel loss,improve the Signal To Reverberation Ratio(SRR) processing gain and the signal detection performance of the sonar system.The sea trial data processing results show that the proposed algorithm enhances the SRR processing gain by about 7-12 dB compared with matched filtering and 6 dB compared with adaptive correlator.