A scalable approach for the description of dependencies in hard real-time systems

Abstract

During the design iterations of embedded systems, the schedulability analysis is an important method to verify whether the real-time constraints are satisfied. In order to achieve a wide acceptance in industrial companies, the analysis must be as accurate as possible and as fast as possible. The system context of the tasks has to be considered in order to accomplish an exact analysis. As this leads to longer analysis times, there is a tradeoff between accuracy and runtime. This paper introduces a general approach for the description of dependencies between tasks in distributed hard real-time systems that is able to scale the exactness and the runtime of the analysis. We will show how this concept can be applied to a real automotive application. © 2010 Springer-Verlag.