Research And Implementation Of Privacy Protection In Internet Of Vehicles

The Internet of Vehicles(IoV)is a Mobile Internet that can coordinate vehicles,roads,base stations,central clouds,and pedestrians.It has unique advantages in preventing problems of frequent traffic accidents,optimizing the allocation and deployment of traffic resources,and improving the driving experience of car owners.The exertion of these advantages depends on the shared information of vehicles in the communication environment of the IoV.The traffic management center first summarizes the beacon messages broadcasted regularly by vehicles over a period of time containing their own driving conditions and the nearby real-time traffic condition information collected by vehicles.Then traffic management center processes this information by means of data analysis to grasp the latest trends of traffic conditions,and then makes the best traffic decision or provides users with other high-quality services related to road conditions,vehicles and the environment.However,vehicles communicates based on open wireless channels in the IoV,which is vulnerable to various networking attacks and security threats,thus the security of message exchanges cannot be guaranteed.Moreover,these messages often contain sensitive information such as vehicle identities and location that are closely related to user privacy.Once disclosed,it will cause users’ concerns about privacy and reduce users’ enthusiasm to participate in the data sharing in IoV.Therefore,it is necessary to solve the possible privacy disclosure problems of the IoV.Therefore,this thesis focuses on the privacy protection issue in the communication environment of the IoV.Based on the thorough investigation and analysis of the state-of-the-art privacy protection techniques of the IoV,this thesis conducts an in-depth study from two aspects of the IoV:the identity privacy and the location privacy.The main contents of work are as follows:(1)Aiming at identity privacy protection,this thesis proposes a distributed anonymous authentication scheme for the IoV based on zeroknowledge proof and blockchain techniques.First of all,in order to solve the problems of the heavy burden and the single point of failure of the trusted center in the traditional centralized authentication scheme,the proposed scheme introduces the blockchain into the traditional IoV communication model,and uses the blockchain to store vehicle authentication information to achieve distributed authentication.Secondly,the scheme proposes an authentication algorithm based on zero-knowledge proof and Merkle tree for the initial authentication of registered vehicles to RSU.The vehicle only needs to provide the RSU with zero knowledge identity credential in the authentication process,which can not only avoid the problem of certificate management,but also prevent the semi-trusted RSU from obtaining information useful for revealing the vehicle identity.After the initial authentication,the authentication information of a vehicle has been stored on the blockchain.When it enters subsequently into a new area and needs to be re-authenticated with the RSU,the RSU only needs to call the smart contract on the blockchain to authenticate the legitimacy of the vehicle’s identity.This realizes the distributed authentication function.In addition,the proposed scheme introduces the concept of regional service management organization and regional secret value to solve the problem of the difficulty in revoking malicious vehicles in traditional identity-based signature authentication schemes.Finally,the security analysis to the scheme shows that it can meet the security requirements.At the same time,this thesis implements the proposed authentication scheme,and evaluates the experimental results.The evaluation results show that the overhead and authentication efficiency of the scheme can meet the requirements of the IoV.(2)Aiming at location privacy protection,the thesis proposes a personalized location privacy protection scheme based on deep reinforcement learning and differential privacy.First,in order to solve the problem that the existing IoV location privacy protection schemes have not considered the personalized privacy requirements of different users,this thesis proposes a new personalized location sensitivity level assignment scheme,which not only allows users to determine their own sensitive location sets by building a semantic location classification tree,but also allocate corresponding sensitivity levels to different stop points by extracting stop point features from a user’s historical trajectory data set,where the sensitivity level of the location determines the privacy protection strength that needs to be provided for the location.Secondly,this scheme proposes a personalized location privacy protection algorithm based on deep reinforcement learning and differential privacy.This algorithm can help users find an rational privacy budget allocation strategy to add noise and generate interference locations,and balance the relationship between the protection of sensitive semantic locations and the loss of quality of service.Finally,the proposed location privacy protection scheme is implemented and experimented on the real trajectory dataset.The the experimental results are evaluated.The evaluation results show that the proposed scheme can realize personalized location privacy protection,and can give higher privacy protection strength to more sensitive semantic locations,thereby reducing the risk of location semantic information leakage.At the same time,compared with the traditional location privacy protection scheme based on differential privacy,the location privacy protection scheme proposed in the thesis can better resist Bayesian inference attack.