Using system-level timing analysis for the evaluation and synthesis of automotive architectures

Emerging technologies allow the implementation of advanced car features enhancing the safety and the comfort of the driver. These complex functions are distributed among several ECUs, implemented by multiple tasks executed on each processor, and are characterized by non-functional requirements, including timing constraints. The design of the physical architecture and the placement of tasks and messages must be performed in accordance with the constraints and optimizing the performance of the functions. We show how schedulability analysis can be used in the development of complex automotive systems to find the architectures that can best support the target application in a what-if iterative process, and we address the opportunities for the synthesis of architecture configurations. A case study of an experimental vehicle shows the applicability of the approach.