Energy Aware Scheduling Mandatory/Optional tasks in Multicore Real-Time Systems

ISABEL MENDEZ-DIAZ; PAULA ZABALA; RODRIGO SANTOS; JAVIER OROZCO

Montevideo

Workshop; VIII ALIO/EURO Workshop on Applied Combinatorial Optimization; 2014; 2014

Association of Latin-Iberoamerican Operational Research Societies y Association of European Operational Research Societies

Reward based scheduling of real-time systems ofperiodic, preemptable and independent tasks with mandatoryand optional parts in homogeneous multiprocessors with energyconsiderations is a problem that has not been analysed before.The problem is NP-Hard. In this paper, a restricted migrationschedule is adopted in which different jobs of the same task mayexecute in different processors and at different power-modes butno migration is allowed after the job has started its execution.An objective function to maximize the performance of the systemconsidering the execution of optional parts, the benefits of slowingdown the processor and a penalty for changing the operationfrequency is introduced together with a set of constraints thatguarantee the real-time performance of the system. Differentalgorithms are proposed to find a feasible schedule maximizingthe objective function and are compared using synthetic systemsof tasks generated following guidelines proposed in previouspapers.