Online reverse engineering of CAN data

Autor: Toon Bogaerts, Peter Hellinckx, Siegfried Mercelis, Jens de Hoog, Wim Casteels
Jazyk: Dutch; Flemish
Rok vydání: 2020
Předmět:
Zdroj: Internet of Things
ISSN: 2542-6605
Popis: These days, most vehicles are equipped with the Controller Area Network (CAN) messaging system. This system enable numerous on-board sensors to share useful data with each other. Clearly, this data can be of high value for other applications within Internet of Things. However, the interpretation of that data is devious due to the CAN protocol. Hence, to give sense to the acquired data, reverse engineering is needed to provide insights of the sensor data of the vehicle. In terms of reverse engineering, this is already an ongoing topic in research. However, no methods already exist to perform this reverse engineering process in an online way. Therefore, this paper presents novel methods to detect and identify signals, both for continuous and discrete signals. The reverse engineering process for continuous signals deals with search space reduction and evaluating correlations between requested signals and obtained references. This paper also introduces a new methodology to generate reference signals by transferring knowledge of different vehicles. In terms of reverse engineering discrete signals, this paper presents a novel method that utilises the user during the process. The results show that our presented methods have a high potential value, but generalisation over a wider variety of vehicles and situations is needed. However, this paper paves a way to a more scalable and usable solution for reverse engineering in future applications.
Databáze: OpenAIRE
Externí odkaz: