Actual measurements have shown that in many cases, the wireless channel is a channel with sparse. In order to use the bayesian channel estimation algorithm to estimate channel, using sparse characteristic of the channel at the same time, the paper improved the search method of support set in SABMP algorithm .In the process of dominate search support set introduces the ideas of viterbi decoding path, eventually reduce the computational complexity of the algorithm. Theoretical analysis and simulation results verify the effectiveness of the improved algorithm.