Research On Trust Management In Opportunistic Networks

Along with the emergence of a large number of intelligent devices that areprovided with short-range wireless communication capabilities, and the development ofcomputer and network technology, the Opportunistic Networks have been developingrapidly. Opportunistic Networks has become an important part of the future Internetof Things and pervasive computing, and it is also the theoretical basis ofOpportunistic Compute and Mobile Compute, and its application scope is more andmore widely.But the disoperative nodes or malicious nodes in Opportunistic Networks canmake whole network performance dropped substantially even paralysis. And theunconnected or self-organization of Opportunistic Networks makes existing trustmanagement strategies can’t used in Opportunistic Networks. This article in view ofthe characteristics of Opportunistic Networks proposed a bayesian-based trustmanagement strategy. In the trust management of bayesian-based, through thebayesian method to calculate the direct trust, and consider the node preference,through the normalization method synthesis direct and indirect trust to the finally trustvalue, according to the trust value of the node to achieve the purpose of access control.In the proposed method, trust does not decay over time and complete trust updateafter the node interactive according to certain interactive cycle, saves the time ofcollection before the interaction time for trust computation and trust recommended.But in Bayesian-based method, the ambiguity of trust is equivalent to randomness andnode evaluation is often not deterministic, so this paper proposes the trustmanagement strategy based on the theory of fuzzy. Fuzzy-based trust managementstrategy evaluates nodes’ trust from multiple attribute, such as file quality,transmission rate, response time, etc. Through the fuzzy membership function toevaluates the evaluation of each attribute and Fuzzifies them into single factorevaluation vector, to fuzzy comprehensive process the trust vector of evaluation get comprehensive trust evaluation, and defuzzify the eventually comprehensive fuzzytrust into trust value through the gravity method, thereby get the final evaluation ofnodes.Simulate the trust management strategy with the Opportunistic Networkssimulation software ONE. The proposed trust management strategies is for theOpportunistic Networks typical forwarding algorithm Epidemic、SprayAndWait andMaxProp, that is to contrast to the proposed trust management method on networkperformance. The simulation results show that: add the trust management strategyforwarding algorithm in Opportunistic Networks with malicious nodes can get higherforwarding efficiency of the network and lower average network delay.