Feasibility Analysis of Sporadic Real-Time Multiprocessor Task Systems

TL;DR

This paper introduces the first algorithm to test the feasibility of sporadic real-time task systems on multiprocessors, addressing key open problems in multiprocessor scheduling and comparing discrete and continuous scheduling approaches.

Contribution

It provides the first feasibility testing algorithm for sporadic tasks on multiprocessors and explores related notions like schedulability and online feasibility, also comparing discrete and continuous schedules.

Findings

First feasibility testing algorithm for sporadic tasks on multiprocessors

Discrete-time schedules are as powerful as continuous-time schedules

Addresses open problems in multiprocessor real-time scheduling

Abstract

We give the first algorithm for testing the feasibility of a system of sporadic real-time tasks on a set of identical processors, solving one major open problem in the area of multiprocessor real-time scheduling [S.K. Baruah and K. Pruhs, Journal of Scheduling, 2009]. We also investigate the related notion of schedulability and a notion that we call online feasibility. Finally, we show that discrete-time schedules are as powerful as continuous-time schedules, which answers another open question in the above mentioned survey.