Research On Reputation Value Evaluation And Trusted Storage For Crowdsourcing Nodes In Internet Of Vehicles

With the large-scale commercialization of 5th Generation Mobile Communication Technology(5G),its low latency,high density and high reliability network features provide communication foundation for the development of Internet of Vehicles(Io V).And with the abundance of on-board equipment,the improvement of storage and computing capacity,the function of vehicle is richer and richer.In order to make full use of vehicle performance,we can combine the traditional concept of Crowdsourcing with Io V to form Vehicular Crowdsourcing Networks(VCNs).So vehicles can obtain or provide services through the platform in a low-cost and flexible way,such as advertising services,route planning,emergency reporting,and so on.However,malicious behaviors of vehicular crowdsourcing nodes in the network,such as false service or misinformation,may affect or even damage VCNs.Therefore,in order to ensure its service quality,the use of reputation management can effectively restrain the behavior of vehicles.In order to realize the reputation management of VCNs’ vehicular crowdsourcing nodes,this thesis mainly studies the reputation evaluation and reputation storage from two perspectives.Firstly,this thesis designed a reputation management framework for crowdsourcing nodes in the VCNs.Then,according to the framework of reputation value management,a reputation evaluation scheme is designed.Finally,based on blockchain technology,the consensus protocol is optimized,and the research of trusted storage is carried out.The main research contents of this thesis are as follows:(1)Based on the research background of VCNs,in order to prevent the harm of malicious nodes or malicious behaviors to the system,it is necessary to realize effective reputation management.Therefore,a reputation management framework based on vehicular crowdsourcing nodes is proposed,which mainly includes a two-layer structure of RSU layer and vehicle layer to realize reputation management together.The framework clearly defines the main responsibilities of each layer,the entity objects it contains and the overall workflow.(2)Based on the reputation management framework,a reputation evaluation scheme is designed.The scheme selectively accepts the evaluation task at the vehicle end according to the similarity of attributes of the messages to be evaluated,so as to improve the accuracy of the reputation evaluation from the source.Trusted Execution Environment(TEE)is used on the RSU side to ensure the reliability of the calculation process,and the intermediate reputation value is calculated according to Bayesian inference.Update the global reputation value by taking into account the vehicle’s role attributes(as message sender and evaluator)and historical reputation values.Moreover,the complexity of the proposed scheme is analyzed,and the efficiency of the proposed scheme is verified by experiments.(3)Based on the blockchain technology,combined with the background of reputation management,a blockchain-based trust storage scheme is proposed.In order to solve the impact of the consensus protocol on the updating efficiency of the trusted storage scheme,the thesis studies and optimizes the consensus protocol,and proposes the FSBFT(Fast Speculative Byzantine Fault Tolerance)consensus protocol to meet the performance requirements of VCNs.Then,a storage scheme based on blockchain is designed.First,the RSU selects nodes according to the reputation threshold,and then the selected nodes update together the reputation values stored in TEE to the blockchain according to the optimized consensus protocol FSBFT.Finally,by comparing various consensus protocols,it is verified that the optimized consensus protocol FSBFT has lower delay and higher throughput.