Loopy Belief Propagation Algorithm in Distributed Wireless Cooperative Spectrum Sensing

Considering the date fusion in distributed cognitive radio networks, this paper studies the loopy belief propagation (LBP) algorithm to make the data fusion in distributed network with loops, and also gives a weighted modified LBP. This paper considers the effect of a shadow fading in the wireless channel model, and also exists malicious nodes, which may give the false result. The key idea is to model the network of secondary users as Markov random fields and to apply the loopy belief propagation algorithm or the weighted loopy belief propagation algorithm to facilitate decision fusion in a distributed manner. The paper shown that weighted LBP can solve the deep shadow fading and malicious nodes more effectively. Finally, a simulation with MATLAB has been done and verifies the results.