Embedded Virtualization Approaches for Ensuring Safety and Security within E/E Automotive Systems

Automotive E/E systems are subject to several conflicting non-functional requirements. Monetary costs, shorter technology and time-to-market cycles and rising restrictions to energy consumption build a subset of them. Furthermore, existing safety and security criteria must be met. Because of future game
changers, like highly automated driving or share-economy scenarios, the relevance to incorporate such non-functional requirements within the product development will increase.
Within the ARAMiS (Automotive, Railway and Avionic Multicore Systems) project funded by BMBF, different use-cases of the industry and scientific partners were analyzed. The project goal is to enable multicore systems across mobility domains and show different technical concept implementations and detail the achieved improvements over traditional systems.
In ARAMiS, a special focus is placed on embedded virtualization technologies for ensuring safety and security according to existing standards (ISO26262). For isolation purposes and flexible software relocation of Enterprise Systems, hypervisors are already an attractive and proven approach. Here, we analyze virtualization methods within real-time systems for hardware and software to enable embedded hypervisors for automotive needs.
... mehr
Within this paper, we present results arising from the work in the automotive domain within the ARAMiS research project. For that purpose, we have developed OEM specific demonstrator platforms for evaluation purposes to reflect typical automotive use-cases. These platforms are implemented on different hardware controllers, performance classes and mission scenarios. Specifically, we analyze the operation of mixed integrity systems.
As a result, the commonalities and technical challenges between demonstrator platforms are identified, and the suitability of virtualization technologies for embedded multicore controllers for automotive E/E systems is discussed. The results enable next generation scenarios for industry partners and provide guidance for future research activities.