Towards Self-organization in Automotive Embedded Systems

Gereon Weiss, Marc Zeller, Dirk Eilers, Rudi Knorr

Abstract: The rising complexity of upcoming embedded computing systems cannot be managed by traditional methodologies. Hence, with Autonomic Computing and Organic Computing new paradigms of self-organizing systems have been introduced. The automotive sector as an application domain of embedded systems also has to face the challenge of growing complexity. Thus, in this paper we show a potential evolution of automotive electronic systems to (partly) self-organizing systems and identify the nowadays missing capabilities. We chose the infotainment domain for exemplarily demonstrating this evolution by introducing new enhancements for self-organization. These extensions are evaluated in a case study of a typical vehicle infotainment system based on the MOST (Media Oriented Systems Transport) technology. The implementation and evaluation show that the newly introduced techniques work well in a realistic scenario. Thereby, we validate that an evolution of present statically designed automotive electronic systems to self-organized systems is feasible.