Research And Implementation Of Network Vulnerability Detection Technology For Intelligent Vehicle

With the development of network technology,security incidents in the Internet have become very common.Similarly,in the automotive field,as people become more and more demanding for the networking and intelligence of automobiles,automobiles are gradually exposed to external networks from a closed environment,which means that it will face greater security risk.Therefore,research on automotive network security issues in recent years is particularly urgent and important.At present,automobile security threats mainly exist in the bellowing seven major modules of the automobile,which includes TSP(Telematics Service Provider),APP,T-Box(Telematics Box),IVI(In-Vehicle Infotainment),CAN bus,ECU and V2V(Vehicle to Vehicle)communication.In order to avoid the automobile's vulnerabilities being used by criminals to cause huge losses to personal safety and property,it is necessary to detect network vulnerabilities in the automobile module in advance to prevent security problems.To this end,this thesis studies the automotive network security issues and implements the detection of automotive network vulnerabilities.The main research work of the thesis is as follows:The thesis analyzes the network threats faced by the main modules of the automobile,and divides it into the in-vehicle network and the au out-vehicle network according to its network architecture.TSP and in-vehicle IVI are the components based on the traditional Ethernet architecture.The vulnerability detection method makes use of the network to collect information and initiate active attacks.The vulnerability detection method for the in-vehicle network mainly uses fuzzy test to conduct vulnerability mining.Study the protocol of the in-vehicle-network.The security threat analysis of CAN bus protocol,UDS protocol and DOIP protocol is carried out to find the vulnerability of the protocol,and the types of vulnerabilities that may be caused by the variation of different fields are given.The thesis presents a framework of intelligent vehicle network vulnerability detection based on fuzzy testing,and designs and implements a network vulnerability detection system suitable for vehicle intranet,vehicle IVI and TSP.The data model and state model are designed for the in-vehicle communication protocol.The existing fuzzy test mutation strategy is analyzed,and a weight-based fuzzy test improvement strategy is proposed.Specifically,a field with a high weight value has more chances of mutation.In this way,the quality of the synthesized test case is improved,thereby achieving the most abnormal response triggered by the least number of test cases,and increasing the probability of successful vulnerability mining.An abnormal monitoring mechanism for the interior ECU is proposed.The in-vehicle node and fuzzy test engine data I/O are realized.In addition,a unique mutated data set is set for the UDS protocol,and the types of test cases are increased to achieve the purpose of triggering different vulnerabilities.The intelligent vehicle network vulnerability detection systemproposed in this thesis successfully detected three vulnerabilities in the automobile ECU and TSP,indicating that the fuzzy test technology is effective for detecting the vulnerability of the automobile network.