The Research on Analyzing the Number of Signal Definitions of In-Vehicle CAN Networks

Abstract:
In recent years, automotive cyber security has attracted widespread attention as vehicles continue to suffer from cyber attacks. Some of these attacks aim to penetrate into in-vehicle CAN networks such that the underlying vehicle could be controlled. Intrusion detection systems are then proposed for in-vehicle networks. However, the application of related security methods and tools is hindered by the non-uniform format of proprietary message definitions of in-vehicle networks. To solve the above problem, this paper proposes a method for identifying and analyzing the number of bus protocol signal definitions. The method performs classification by extracting frequency features and finally obtains the number of signal definitions. The experimental results show that the method can effectively identify the number of signal definitions, which is important and promising for reverse analysis of bus messages in subsequent security research.