Improving schedulability of fixed-priority real-time systems using shapers

Abstract

In this paper, we introduce a technique for improving the schedulability of real-time embedded systems with fixed-priority scheduling. Our technique uses shapers to reduce the resource interference between higher-priority and lower-priority tasks, and thus enables more lower-priority tasks to be scheduled. We present a closed-form solution for the optimal greedy shaper for periodic tasks with jitter, as well as a schedulability condition for tasks in the presence of shapers. We also discuss two applications of greedy shapers: In compositional scheduling frameworks, shapers can help optimize the resource interfaces of real-time components, and in mixed-criticality systems, they can reduce deadline misses of low-criticality tasks while preserving schedulability of high-criticality tasks, even with lower priorities. We demonstrate the utility of our technique through an evaluation based on randomly generated workloads.