The sea bottom reverberation is one of the important problems, which restrict the bottom targets detection performance with the high frequency imaging sonar. In this paper, a bottom reverberation suppression algorithm is presented to improve the performance of the high frequency side scan sonar for small target detection. In the paper, the statistics characteristic of the target echo and the bottom reverberation is introduced, with the experiment data processing. The target echo and the bottom reverberation could be separated with the difference of the statistics characteristic. And the Blind Source Separated method was applied to implement the signal separation. At last, the effective of this method was proved by the process of the experiment data.