Automating ecu identification for vehicle security

Michael Jaynes, Ram Dantu, Roland Varriale, Nathaniel Evans

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

24 Scopus citations

Abstract

The field of vehicular cybersecurity has received considerable media and research attention in the past few years. Given the increasingly connected aspect of consumer automobiles, along with the inherent danger of these machines, there has been a call for experienced security researchers to contribute towards the vehicle security domain. The proprietary nature of Controller Area Network (CAN) bus messages, however, creates a barrier of entry for those unfamiliar, due to the need to identify what the messages on a given vehicle's bus are broadcasting. This work aims to automate the process of correlating CAN bus messages with specific Electronic Control Unit (ECU) functions in a new vehicle, by creating a machine learning classifier that has been trained on a dataset of multiple vehicles from different manufacturers. The results show that accurate classification is possible, and that some ECUs that broadcast similar vehicle dynamics broadcast similar CAN messages.

Original languageEnglish
Title of host publicationProceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages632-635
Number of pages4
ISBN (Electronic)9781509061662
DOIs
StatePublished - Jan 31 2017
Externally publishedYes
Event15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016 - Anaheim, United States
Duration: Dec 18 2016Dec 20 2016

Publication series

NameProceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016

Conference

Conference15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016
Country/TerritoryUnited States
CityAnaheim
Period12/18/1612/20/16

Keywords

  • Automotive security
  • CAN bus
  • Classification
  • Machine learning

Fingerprint

Dive into the research topics of 'Automating ecu identification for vehicle security'. Together they form a unique fingerprint.

Cite this