Ensuring Trustworthy Automated Road Vehicles: A Software Integrity Validation Approach

This paper presents a cascading scheme designed to measure and validate the software integrity of interconnected and automated road vehicles. With modern vehicular architectures decoupling hardware from specific functions and supporting frequent updates, we require proactive means to detect software manipulations, especially as compromised control units pose a severe risk to passenger safety. Therefore, we must ensure the benignity of critical software components that impact the vehicle's dynamics while allowing for rapid adaptions and user customizations. Our scheme requires vehicles to obtain proof of trustworthiness before performing specific actions, such as driving autonomously or joining a vehicle fleet. To achieve this, we employ remote attestation techniques to compute an integrity identifier verified and validated by a trusted third party. We specifically design the process of integrity measurement for hierarchically organized networks of control units, as is the case in domain-based and zonal architectures. For evaluation, we utilize a recently presented zonal architecture deployed in four fully automated road vehicles and demonstrate that our scheme incurs an acceptable overhead.