Scheduling real-time systems with end-to-end timing constraints using the distributed pinwheel model

Abstract

Real-time distributed applications have timing constraints on tasks running on several processors. To design real-time systems with end-to-end performance requirements, we need to have algorithms to schedule and to coordinate tasks on different processor nodes. An end-to-end scheduling approach based on the pinwheel scheduling model is presented for distributed real-time systems, We show how tasks on different nodes may be transformed to have periods consisting of only harmonic numbers. With harmonic periods, we can use a polynomial-time algorithm to find the start and finish times of each task on each node. Phase alignment algorithms are then applied to adjust the phases between schedules of neighboring nodes so that the overall end-to-end delay is minimized. Using the distributed pinwheel model, schedules on different nodes are synchronized to lessen the delays. For many real-time systems, this approach provides a predictable performance and a short end-to-end delay.