V2C: A Trust-Based Vehicle to Cloud Anomaly Detection Framework for Automotive Systems

Abstract

Vehicles have become connected in many ways. They communicate with the cloud and will use Vehicle-to-Everything (V2X) communication to exchange warning messages and perform cooperative actions such as platooning. Vehicles have already been attacked and will become even more attractive targets due to their increasing connectivity, the amount of data they produce and their importance to our society. It is therefore crucial to provide cyber security measures to prevent and limit the impact of attacks. As it is problematic for a vehicle to reliably assess its own state when it is compromised, we investigate how vehicle trust can be used to identify compromised vehicles and how fleet-wide attacks can be detected at an early stage using cloud data. In our proposed V2C Anomaly Detection framework, peer vehicles assess each other based on their perceived behavior in traffic and V2X-enabled interactions, and upload these assessments to the cloud for analysis. This framework consists of four modules. For each module we define functional demands, interfaces and evaluate solutions proposed in literature allowing manufacturers and fleet owners to choose appropriate techniques. We detail attack scenarios where this type of framework is particularly useful in detecting and identifying potential attacks and failing software and hardware. Furthermore, we describe what basic vehicle data the cloud analysis can be based upon.