| The booming development of Internet of Vehicle has brought people better transportation services while driving.However,due to the characteristics of open network,there are some security threats in Internet of Vehicle.Unregistered illegal vehicles pretending to be legitimate users enter the network to steal or falsify data;Legitimate registered vehicles send false messages due to faulty sensors or selfish intentions;Passive vehicles always do nothing.These malicious vehicles not only affect the normal operation of traffic,but also endanger people’s lives in serious cases.In order to prohibit and identify malicious vehicles more accurately and efficiently and ensure the authentication,integrity and reliability of information transmission.The following research is carried out in this thesis:(1)Based on blockchain technology,a malicious vehicle identification framework for the Internet of Vehicle is proposed,which is mainly composed of vehicle layer,RSU layer,KGC and TA and blockchain layer.The framework can realize secure identity authentication,reliable trust management and efficient trust value update.The framework clearly defines the entities and main functions involved in each layer,the overall workflow and the characteristics of the framework itself.(2)A certificateless aggregation signcryption scheme against forgery attacks is proposed.The purpose of the scheme is to prevent illegal vehicles from joining the network.Some certificateless aggregation signcryption schemes applicable to Internet of Vehicle still have the following security issues: Malicious vehicles can still forgery signcryption by replacing the public keys of legitimate vehicles.Multiple malicious vehicles generate invalid signcryption which can still pass the aggregation verification.In order to solve the above problems,this scheme analyzes and proves the problems in the existing schemes,and utilizes a secret key preimage protection technology and hash collision resistance to resist signature forgery attack and coalition attack.Based on the difficulty problem,confidentiality and unforgeability of the scheme are proved under the random oracle model.Finally,the performance analysis shows that the proposed scheme enhances the security without increasing the computing and communication cost.(3)A trust evaluation scheme based on activity is proposed.The purpose of the scheme is to remove malicious vehicles which send false information and passive vehicles from the network.Most trust evaluation schemes of Internet of Vehicle focus on identifying malicious vehicles that send false information in the network,but ignore the impact of negative vehicles on the accuracy of trust evaluation and network operation.In order to solve the above problems,this scheme utilizes the normalized tangent function to map the unit time number of vehicle nodes participating in information verification in the network to the active factor to reflect the activity degree of vehicles.The activity degree of vehicle is regarded as one of the factors affecting their comprehensive trust value,so as to promote the active participation of vehicles in information verification correctly.The vehicles send false information and passive vehicles will be removed from the network due to low comprehensive trust value.In this thesis,RSUs are responsible for the calculation and updating of the comprehensive trust value of vehicles.PBFT algorithm is improved in the process of RSUs consensus.Some trusted RSUs with idle resources are selected as consensus nodes,which improves the efficiency of consensus. |
