University of North Alabama
Over the years, the sporadic task model has proven remarkably useful for the modeling of recurring processes that occur in hard-real-time systems. The multiprocessor scheduling of collections of real-time jobs is considered. Sufficient conditions are derived for determining whether a specified system meets all deadlines when scheduled by a static-priority algorithm. These conditions are used to obtain efficient schedulability tests for sporadic task systems scheduled using the popular deadline-monotonic real-time scheduling algorithm. Resource-augmentation bounds are provided that quantify the tightness of these schedulability tests.