This project aims to address the challenge of reverse engineering CAN bus for security purposes, especially when a third part lacks a DBC file and is unaware of the protocol. Traditional methods fail to identify the exact bits controlling specific actions due to the vast search space, making them unsuitable for reverse engineering and analysis of in-vehicle networks. To overcome this, we propose an efficient algorithm that significantly reduces the complexity of the search space. This algorithm has been successfully implemented on a Raspberry Pi system, with part of the source code released for public use.
Our key contributions include:
- An efficient algorithm that significantly reduces the complexity of searching for control bits.
- Implementation of the algorithm on a Raspberry Pi system, with partial source code available for developers.
- Precise mapping of CAN signals to vehicle control actions, enhancing security detection capabilities.
- Provides an efficient and robust reverse engineering solution for unique ID and payload combinations, enabling better security analysis for in-vehicle networks.