关键词: 车联网安全, 内生安全, 拟态防御, 强化学习, 信息系统安全

Abstract: With the rapid development of automotive intelligence, onboard system changes the landscape of vehicle behavior automation. Various firmware and hardware devices can interact or exchange information with the onboard intelligent system. The Internet of vehicles carries the automatic control of software, ECU and hardware via the onboard intelligent system. Instate providing users with daytoday driving functionality, the onboard system been evolved and increase its complexity. There is no clear boundary between system security and functional safety. This paper gives an overview of the onboard intelligent system of the Internet of vehicles based on experimental modeling. It also emphasizes that under the scenario of the Internet of vehicles, the vulnerability and system failure of the intelligent vehicle system will directly affect the functional safety, which means it can threaten the safety of passengers. Therefore, the onboard system security of the Internet of vehicles becomes more and more important. This paper discusses the relationship between system security and functional safety in the Internet of vehicles based on an existing issue. In order to locate the actual system security in the Internet of vehicles, the existing defense indicates that the importance to find a balance point between vehicle performance and system security within the limited resource, this paper proposed a method about prereinforcement learning defense mechanism based on pseudo defense.Key words Internet of vehicles security; endogenous security; mimicry defense; reinforcement learning; information system security

Key words: Internet of vehicles security, endogenous security, mimicry defense, reinforcement learning, information system security