An Analysis Platform for the Information Security of In-Vehicle Networks Connected with External Networks

Abstract

Abstract. Most in-vehicle units, which are used for various services such as remote diagnosis, information gathering, telematics, car navigation with communications, etc., are connected to both an external network and an automotive network, referred to as an in-vehicle network. The information security of the in-vehicle network is an important concern. In this paper, we describe the analysis platform that we designed and developed for protection mechanisms to analyze and verify the information security of an in-vehicle network connected with external networks. The platform provides features to analyze messages of the in-vehicle network, to evaluate the effects of attacks on the in-vehicle network, and to verify protection mechanisms for the in-vehicle network. First, we developed a message analysis platform subsystem that analyzes messages of the in-vehicle network and performs send and receive processes in the in-vehicle network. Next, we designed an attack evaluation platform subsystem that evaluates attack mechanisms from servers or attackers on the Internet. We also developed a prototype (MoVIS) of the subsystem. We confirmed the effectiveness of the attack evaluation platform and MoVIS in an actual automobile, which was subjected to spoofing and DoS attacks to the in-vehicle network via the Internet. Finally, we discuss the requirements of the protection verification platform, which provides protection mechanisms and protection features based on the experimental results and the characteristics of the in-vehicle network.

Techniques for utilizing classification towards securing automotive controller area network and machine learning towards the reverse engineering of CAN messages

“…The attacks demonstrated by these works [38,9,52,1,18,96,90] can be broken down into the following categories:…”