Analysis of the Efficacy of the Use of Inertial Measurement and Global Positioning System Data to Reverse Engineer Automotive CAN Bus Steering Signals

TL;DR

This paper demonstrates that combining inertial measurement and GPS data significantly enhances the reverse engineering of vehicle steering signals from CAN bus data, aiding vehicle control and security analysis.

Contribution

It introduces a novel method utilizing IMU and GPS data to improve reverse engineering of vehicle CAN signals without prior vehicle knowledge.

Findings

GPS data improves correlation for deceleration signals

IMU and GPS data together enhance CAN channel identification

Method supports vehicle control and cybersecurity applications

Abstract

Autonomous vehicle control is growing in availability for new vehicles and there is a potential need to retrofit older vehicles with this capability. Additionally, automotive cybersecurity has become a significant concern in recent years due to documented attacks on vehicles. As a result, researchers have been exploring reverse engineering techniques to automate vehicle control and improve vehicle security and threat analysis. In prior work, a vehicle's accelerator and brake pedal controller area network (CAN) channels were identified using reverse engineering techniques without prior knowledge of the vehicle. However, the correlation results for deceleration were lower than those for acceleration, which may be able to be improved by incorporating data from an additional telemetry device. In this paper, a method that uses IMU and GPS data to reverse-engineer a vehicle's steering wheel position CAN channels, without prior knowledge of the vehicle, is presented. Using GPS data is shown to greatly improve correlation values for deceleration, particularly for the brake pedal CAN channels. This work demonstrates the efficacy of using these data sources for automotive CAN reverse engineering. This has potential uses in automotive vehicle control and for improving vehicle security and threat analysis.