Model-Based Function Mapping and Bandwidth Reservation for Mixed-Critical Adaptive Systems
Résumé
The new AutoSAR adaptive platform makes mixedcritical automotive systems able to adapt at runtime in response to hardware and software failures. However, mapping functions of these adaptive automotive systems to the hardware nodes and reserving their bandwidth are still major challenges. To tackle such major challenges, in this paper, we propose a model-based approach for the automatic identification of functions mapping and bandwidth reservation for the different configurations of an adaptive software system. The functions mapping and bandwidth reservation are performed by a Tabu search-based strategy. In this strategy, there is a temporal isolation between critical and non-critical functions. Thus, overruns of non-critical functions do not affect timing guarantees of critical functions, and the quality of service for the non-critical functions is maximized.