Static Analysis of Controller Area Network Communication for Attack Detection

Abstract

The vulnerability of modern cars increases due to their multiple connections to the environment, which offers the possibility of remote attacks in the worst case with fatal outcome. The controller area network (CAN) is still highly used and includes no security features, so intrusion detection systems are a promising approach to secure the communication. The proposed method monitors the CAN communication and uses static checks to differentiate between normal and attack traffic. This enables reliable and comprehensible attack detection and achieves a detection rate up to 100%, generating zero false alarms for the investigated data sets.