Abstract: In order to explore the application of CAN bus in automobile detection technology, an anomaly detection model of CAN bus based on information entropy and anomaly detection technology is constructed, and simulation system is used to simulate the attack of automobile, to analyze the feasibility of application in automobile CAN bus. It is found that the change of information entropy is basically stable in the process of increasing the time window in the analysis of information entropy change of on-board CAN bus, and finally approaches a fixed value. In the attack detection analysis of simulation experiments, it is found that the information entropy value of on-board CAN bus decreases significantly, and the anomaly detection system will issue an alarm, which has a good detection effect on replaying a large number of normal messages. The results show that the model can effectively prevent flooding attacks, a large number of replay attacks, and achieve good expected detection results. Although there are some shortcomings in the experimental process, it still provides a basis for the application of CAN bus detection technology in the automotive industry.
Keywords: CAN bus, information entropy, anomaly detection technology, simulation system
