Finger Printing of Engine Control Units by Using Frequency Response for Secure In-Vehicle Communication

The Controller Area Network (CAN) is a primary protocol for in-vehicle communication between the engine control units. However, there are certain flaws in the architecture of this protocol which induce some inherent security vulnerabilities. This paper proposes a novel idea of authenticating the CAN nodes based on the unique frequency response of each transmitter. Electrical devices leave unique signatures in the transmitted signal, which can be used to fingerprint the source. This uniqueness is imparted because of the non-uniform nature and the imperfection of the material and the design of the device. Researchers have proposed different methods for parametric uniqueness in a physical signal. The method used for fingerprinting in this paper is based on the frequency response of the engine control unit (ECU). The calculated frequency response is then used to train a classifier. Experimental results in the proposed research reveal that from the identical data transmission from different ECUs in the network, it is possible to identify the transmitter (ECU) from the frequency response. This concept can be used to design an efficient Intrusion Detection System for the CAN bus security. It is particularly efficient in detecting the spoofing attacks. The performance and the robustness of the proposed method are also discussed in detail.