Vanet, The Key Issue Of The Security Of The System

VANET is one of the core technologies to realize intelligent transportation systems through offering an efficient information sharing platform for vehicles. It provides drivers and passengers with technical supports in the efficiency and security of the road traffic, the reduction of traffic accidents, the improvement of ride comfort, as well as the safety of their lives and properties. Therefore, researching, developing and enforcing the VANET system are conductive to solving the current problems of the traffic. Definitely, the security issues of the VANET system should also be researched systematically in depth.The thesis, therefore, researches the key issues of VANET security. After investigating the VANET system characteristics, environments and resources, the thesis provides the promising solutions to the issues as follows:1) Construction of the vehicle node reputation model:By analyzing the difficulties and drawbacks during applying the traditional reputation model in VANET, and combining with abundant applications of VANET, we design an application specified reputation model. Taking full advantages of the various data types of vehicle nodes, it merges various types of information sources through the establishment of a Bayesian network to ensure its accuracy and authority. Moreover, under the trusted AUs'evaluation of the vehicle nodes, it will adapt to the highly dynamic VANET quickly and enhance its ability against attacks. In addition, the security analysis and simulation experiments indicate that in comparison with other reputation models, VASRep takes both higher performance and is more suitable for the VANET system.2) Trusted computing of the reputation:By computing and managing the reputation through the trusted computing technologies, it can not only realize the security and high efficient computing of reputation but protect the privacy of the node's context data. It is more suitable for the VANET system that each node sustains its own reputation data locally.3) Quantified risk adaptive access control model for message dissemination: In the view of VANET's dynamic performance and the fact that traditional access control model tends to be more rigid and difficult to be controlled when facing the mutable environment and the expanding application type, the thesis proposes a framework of quantified risk adaptive message dissemination control (QSAMIC) by the quantification of risk of processing. The framework leverages the theories of fuzzy sets and fuzzy logic to resolve the multi-value problems, fuzzy boundary problems and multiple policies and rules results combing problems in decision based on quantitative risk. The thesis also composes flexible risk mitigation mechanisms to ensure that the model takes system security as a prerequisite and then enlarge the information sharing. The QSAMIC can constantly control the risk under the acceptable risk level during the whole session of access, and thus to improve the security of the VANET system more efficiently. Furthermore, flexible access control model is more adaptable to the complicated and volatile situation of the VANET system.4) Specify and enforce the QSAMIC:Based on the eXtensible Access Control Markup Language (XACML), the theis designs a risk adaptive policy language extended from XACML and propose a framework to enforce the policies, including the data flow, decision algorithm, and the functions of the enhanced PEP.5) Sybil attack detection:The Sybil attack is a typical attack mainly against digital ID management in an open environment, and it would cause a severe damage to the VANET system security. So it is very important to detect this attack and take defensive measures for enhancing the security of the VANET system. The thesis proposes the SADSIV algorithm which compares the differences of neighbor vehicle nodes'digital signature vectors and makes a classification of Sybil node set. The evaluation to the the efficiency and robustness through simulation experiments shows that the SADSIV algorithm is promising.