A trust based cooperative model for congestion in ad hoc networks

Ad hoc network is flexibly formed by a group of peer nodes and depends on the irresponsible cooperation among them, so it is vulnerable to various attacks. Thus, trust is introduced to promote nodes to transmit messages. Nodes with higher trust values are more easily selected for transmission which results in unbalanced load distribution. Therefore, packets from different sources are more likely to meet others at the same nodes. Consequently, congestion forms and trust values of congestion nodes decrease. Learning from social life, trust values decrease fast and increase slowly. Then congestion nodes encounter false trust fluctuation, resulting in that source nodes lose the best candidates to transmit packets and the performance of the whole network degrades. Thus, a trust based cooperative model for congestion is proposed to prevent false trust fluctuation and improve successful transmission rate. Simulation results show that congestion easily happens in busy networks and is largely alleviated by our method.