< Terug naar vorige pagina

Publicatie

Automatic reverse engineering of CAN bus data using machine learning techniques

Boekbijdrage - Boekabstract Conferentiebijdrage

The CAN (Controller Area Network) bus connects different Electronic Control Units (ECU) inside a vehicle. Valuable information about the state of the vehicle is present on this bus and is useful to track driver behaviour, the health of the vehicle, etc. However, the configuration of this bus is not publicly disclosed by the car manufacturers. Therefore, reverse engineering techniques need to be applied. Nevertheless, performing these techniques manually is cumbersome and time consuming. In this paper, we propose an automation of the analysis steps of reverse engineering in order to improve and facilitate the process. Two approaches of automation are discussed, namely an arithmetic approach and machine learning using classification. In conclusion, we show that reverse engineering of CAN bus traffic is at least partially possible by applying machine learning techniques and the performance of the classifiers increases when adding additional features to the analysis.
Boek: Proceedings of the 12th International Conference on P2P, Parallel, Grid, Cloud and Internet Computing (3PGCIC-2017) / Xhafa, Fatos [edit.]
Pagina's: 751 - 761
ISBN:978-3-319-69834-2
Jaar van publicatie:2018
Trefwoorden:P1 Proceeding
BOF-keylabel:ja
Authors from:Higher Education
Toegankelijkheid:Closed