一种CAN总线恶意帧检测方法.docVIP

  An Algorithm for Detection of Malicious Messages on CAN Buses Ling Congli, Feng Dongqin** 5 10 15 20 25 30 35 (State Key Laboratory of Industrial Control Technology, Institute of Cyber-Systems and Control, Zhejiang University, HangZhou 310086) Abstract: Control systems are encountering increasing security threats; as one of the common systems, CAN control system is very facile to be attacked. Aiming at the improvement of CAN control system security, an algorithm for detection of malicious CAN messages is given, and it has been implemented in the CANoe simulation environment. The result shows that this algorithm has powerful detection function and valuable practice signification. Keywords: CAN bus; detection algorithm; malicious messages; CANoe 0 Introduction Safety is the degree to which accidental harm is prevented, reduced, and properly reacted to; and security is the degree to which malicious harm is prevented, reduced, and properly reacted to [1] CAN control systems have long been used in areas like automobile, shipbuilding, industrial automation, aerospace, medical equipment, industrial equipment etc. Recently, ever-increasing general protocols, hardware and software have been the dominant parts of CAN control systems. Meanwhile, in present competitive markets, isolated control system networks are being inter-connected. Due to connecting these networks, and introducing IT components into the control systems, security problems arise. CAN protocol have been designed by Robert Bosch in 1986 for automotive applications as a method for enabling robust serial communication[2].The inherited vulnerabilities of such a protocol constitute the main threats of CAN control system. Some security problems in embedded systems are introduced[3]. Vulnerabilities of the automotive CAN control system are analyzed and verified by experiments[4].A simulation of sniffer and replay attack on CAN buses was carried out[5].The security of automobile systems was summarized[6]. It po