A secure recommendation-based reputation system for node incentive cooperation

As the selfish behaviors in wireless networks, nodes will significantly degrade the performance of the network, effective cooperation incentive of nodes has become a hot issue in cooperative communication. A reputation system can stimulate cooperation, but the recommendation-based trust model may make network performance decline, due to fake recommendation, convergence of iteration and node redemption. Hence, a global trust model based on secure recommendation trust is proposed to relieve the malicious influence of fake recommendation on the accuracy of recommendation trust. Mathematic analysis and simulation results reveal that the proposed trust model has convergence characteristic, and the reputation system with this model can more effectively stimulate node cooperation and performs better in terms of PDR (packet successful delivery ratio) and MNPD (mean number of packets dropped). What´s more, the topology transform-based recommendation trust algorithm makes reputation system more suitable for large scale networks.