Context-Aware Adaptation in DySCAS

Abstract

DySCAS is a dynamically self-configuring middleware for automotive control systems. The addition of autonomic, context-aware dynamic configuration to automotive control systems brings a potential for a wide range of benefits in terms of robustness, flexibility, upgrading etc. However, the automotive systems represent a particularly challenging domain for the deployment of autonomics concepts, having a combination of real-time performance constraints, severe resource limitations, safety-critical aspects and cost pressures. For these reasons current systems are statically configured. This paper describes the dynamic run-time configuration aspects of DySCAS and focuses on the extent to which context-aware adaptation has been achieved in DySCAS, and the ways in which the various design and implementation challenges are met.