A Trust Game Model and Algorithm for Cooperative Spectrum Sensing in Cognitive Radio Networks

The existence of malicious secondary users will bring damage to the performance of cooperative spectrum sensing and spectrum access in cognitive radio networks, so effectively curbing malicious behavior of secondary users become the key to success for cooperative spectrum sensing mechanism. Inspired by economics of repeated game theory, a trust game model and algorithm of distributed cooperative spectrum sensing oriented to malicious secondary users named FRTrust is proposed. In FRTrust, the reputation status is used to describe the performance of a secondary user in cooperative spectrum sensing process. It encourages secondary users to choose positive and honest behavior strategies for greater and long term benefits. Simultaneously, a ‘differentiation’ punishment mechanism is presented to respond malicious behaviors of secondary users. By this way, the ‘first offender’ secondary users and the ‘recidivist' secondary users can be distinguished effectively. If a secondary user departs from the normal times in its life period, it will be punished superimposed. Simulation results show that the FRTrust algorithm can encourage secondary users to participate in spectrum sensing in a cooperative attitude, improve the transaction success ratio of cooperative spectrum sensing, and guarantee the fairness and spectrum access performance for cognitive radio networks.