Static-priority scheduling on multiprocessors
Björn Andersson, Sanjoy Baruah, J. Jönsson
Abstract
Björn Andersson, Sanjoy Baruah, J. Jönsson
Abstract
The preemptive scheduling of systems of periodic tasks on a platform comprised of several identical processors is considered. A scheduling algorithm is proposed for static-priority scheduling of such systems; this algorithm is a simple extension of the uniprocessor rate-monotonic scheduling algorithm. It is proven that this algorithm successfully schedules any periodic task system with a worst-case utilization no more than a third the capacity of the multiprocessor platform. It is also shown that no static-priority multiprocessor scheduling algorithm (partitioned or global) can guarantee schedulability for a periodic task set with a utilization higher than one half the capacity of the multiprocessor platform.
OpenAlex reports 323 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The preemptive scheduling of systems of periodic tasks on a platform comprised of several identical processors is considered. A scheduling algorithm is proposed for static-priority scheduling of such systems; this algorithm is a simple extension of the uniprocessor rate-monotonic scheduling algorithm. It is proven that this algorithm successfully schedules any periodic task system with a worst-case utilization no more than a third the capacity of the multiprocessor platform. It is also shown that no static-priority multiprocessor scheduling algorithm (partitioned or global) can guarantee schedulability for a periodic task set with a utilization higher than one half the capacity of the multiprocessor platform.
Key concepts: Uniprocessor system, Computer science, Multiprocessor scheduling, Parallel computing, Earliest deadline first scheduling, Rate-monotonic scheduling, Dynamic priority scheduling, Multiprocessing