Efficient Feasibility Analysis for Real-Time Systems with EDF Scheduling

TL;DR

This paper introduces new fast exact feasibility tests for uniprocessor EDF scheduling, enabling efficient evaluation of task sets and unifying existing schedulability methods into an integrated theory.

Contribution

The paper develops novel algorithms for rapid exact feasibility testing in EDF scheduling, unifying previous bounds and tests into a comprehensive schedulability framework.

Findings

New algorithms evaluate accepted task sets in similar time as old tests.

Feasible task sets previously not accepted are now efficiently tested.

Existing bounds and algorithms are derived from the new tests, unifying schedulability analysis.

Abstract

This paper presents new fast exact feasibility tests for uniprocessor real-time systems using preemptive EDF scheduling. Task sets which are accepted by previously described sufficient tests will be evaluated in nearly the same time as with the old tests by the new algorithms. Many task sets are not accepted by the earlier tests despite them beeing feasible. These task sets will be evaluated by the new algorithms a lot faster than with known exact feasibility tests. Therefore it is possible to use them for many applications for which only sufficient test are suitable. Additionally this paper shows that the best previous known sufficient test, the best known feasibility bound and the best known approximation algorithm can be derived from these new tests. In result this leads to an integrated schedulability theory for EDF.