Simulation-based schedulability assessment for real-time systems

Real-time systems not only require functional correctness, but also specific timing properties. Correct timing is especially challenging for hard real-time systems such as in medicine, avionics, and space industries, where missing a deadline can lead to catastrophic failure. A number of theories tackled this issue to determine whether a set of tasks running on a given architecture meets its timing constraints. One technique is schedulability analysis, which can provide guarantees for the timing behavior for a set of tasks. However, the use of schedulability tests involve an intrinsic amount of pessimism, which greatly reduces the number of configurations that can be considered as schedulable. This removes potentially promising system configurations from the task allocation optimization process, thereby reducing the quality of the final result. The aim of this paper is to overcome this limitation in the context of heterogeneous multiprocessor architectures. We propose a simulation-based approach to assess solutions discarded by a schedulability test, and include them in the optimization process. We tested our method on the optimization of the communication cost of a set of tasks scheduled on a quad core architecture, showing an improvement of up 11% when compared to the use of a schedulability test.