Reverse Engineering the CAN Bus: Vulnerability Analysis in the Tesla Model 3

Reverse Engineering the CAN Bus: Vulnerability Analysis in the Tesla Model 3

연구 분야: Analysis

학회: World Congress in Computer Science, Computer Engineering & Applied Computing

Most current research in automotive security involves attacking the infotainment system, the in-dashboard main computer interface in newer-model cars, including Tesla and other electric vehicles. Because it interconnects with multiple technologies including Bluetooth, USB, 4G/5G, satellite, and has access to many sensors and functions in high-tech cars, the in-dash computer seems to hold the largest attack surface for potential vulnerabilities. But the oldest technology tying the automobile’s systems together is also the most vulnerable, the CAN bus. The CAN 2.0 bus standard, published in 1991, is a plaintext protocol and sufficiently dated that there is no modern security in its design. This can be exploited easily with physical access to the interior of the vehicle, with no user interaction: No social engineering, no links to click, no passwords to enter. Just capturing, sifting through, and modifying a simple protocol is all that is needed to control parts of the dashboard display, and, independently, the functioning of the Tesla’s systems. The Tesla dashboard can be made to report the car is in park, but the parking brake is released. The Tesla display can show the car is moving 7 MPH in reverse, but the Tesla is stationary, and the parking brake is applied. The Tesla’s dashboard can report the hazard lights are active, while the lights on the exterior of the car are dark, or vice versa. With further research, much more malicious actions can be achieved with Teslas and other modern vehicles via CAN bus reverse engineering and exploitation.