Hyeokchan Kwon, Sokjoon Lee and Byung-ho Chung, Electronics and Telecommunications Research Institute, Republic of Korea

In this paper, we present wifi fingerprint-based approach to securing the connected vehicle against wireless attack. In current connected vehicles such as Tesla EV, Mitsubishi outlander PHEV etc., there is a wi-fi access point on the vehicle to connect to the mobile device which has telematics apps installed. And generally the wi-fi access point is managed by the head unit system in the vehicle. Currently, the headunit in the vehicle utilizes white-list that contain MAC addresses of the pre-registered (i.e authorized) device. However, the white-list based mechanism cannot detect the device that forges its MAC address with authorized one. This paper presents security mechanism to detect rogue telematics device that has a spoofed (i.e, forged) MAC by analysing wi-fi fingerprint. We generate wi-fi fingerprint by analysing radio frequency features such as error vector magnitude (EVM), frequency offset, I/Q offset, sync correlation and so on. And we also utilizing distance information for improving detection ratio. The prototype of the proposed mechanism is implemented in this work, and we provide experimental results.

Connected Vehicle Security, Wireless Attack, Wi-Fi Fingerprint, Telematics Device Authentication