Research On Hybrid Trust Evaluation Mechanism In Vehicular Social Networks

With the rapid development of Vehicular social networks,information security issues in Vehicular social networks had become research hot topic in recent years.At present,the research of information security mainly focuses on trust evaluation,authentication mechanism,privacy protection etc.The trust evaluation mechanism plays a vital role in the security of Vehicular social networks.However,the existing trust assessment methods of vehicle nodes mainly based on the historical communication behaviors of the node to evaluates,does not consider and mine the influence of social attributes on trust evaluation,so that the accuracy of the vehicle node trust evaluation needs to be improved.This paper proposed a trust evaluation method for Vehicular social networks,which can effectively suppress the spread of malicious information and improve the security performance of the network.The prime research work is as follows:(1)We established a community detection algorithm based on user interest similarity.The algorithm calculates the interest similarity between the social attributes of vehicle drivers,and introduces the interest similarity into the K-means community discovery algorithm to realize the community division of vehicle nodes.(2)A trust evaluation method based on social factors was proposed.In order to evaluate the trust value of nodes in different communities,social trust was divided into intracommunity trust and inter-community trust by combined with the community discovery algorithm designed in(1).Intra-community,three social factors of nodes interaction frequency,intimacy and similarity were introduced to calculate the social trust by mining the social attributes.Inter-community,nodes were evaluated based on community centrality and social distance.(3)We proposed a hybrid trust evaluation framework in Vehicular social network.The communication behaviors of the vehicle and the social behaviors of the driver was analyzed,constructed a communication trust model and a social trust model to accurate quantify the trust assessment of vehicle nodes.The proposed hybrid trust evaluation framework was compared and analyzed on the ONE simulation platform.The simulation results show that the hybrid trust evaluation mechanism proposed in this paper has a higher delivery rate,reduces the network overhead,enhances the accuracy of vehicle nodes trust evaluation and improves the performance of network transmission.