Scheduling Mandatory-Optional Real-Time Tasks in Homogeneous Multi-Core Systems with Energy Constraints Using Bio-Inspired Meta-Heuristics
In this paper we present meta-heuristics to solve the energy aware reward based scheduling of real-time tasks with mandatory and optional parts in homogeneous multi-core processors. The problem is NP-Hard. An objective function to maximize the performance of the system considering the execution of o...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Graz University of Technology
2019-04-01
|
Series: | Journal of Universal Computer Science |
Subjects: | |
Online Access: | https://lib.jucs.org/article/22604/download/pdf/ |
id |
doaj-bfe24794638149ac90e82cfc91c23174 |
---|---|
record_format |
Article |
spelling |
doaj-bfe24794638149ac90e82cfc91c231742021-06-23T07:57:23ZengGraz University of TechnologyJournal of Universal Computer Science0948-69682019-04-0125439041710.3217/jucs-025-04-039022604Scheduling Mandatory-Optional Real-Time Tasks in Homogeneous Multi-Core Systems with Energy Constraints Using Bio-Inspired Meta-HeuristicsMatias Micheletto0Rodrigo Santos1Javier Orozco2Universidad Nacional del SurUniversidad Nacional del SurUniversidad Nacional del SurIn this paper we present meta-heuristics to solve the energy aware reward based scheduling of real-time tasks with mandatory and optional parts in homogeneous multi-core processors. The problem is NP-Hard. An objective function to maximize the performance of the system considering the execution of optional parts, the benefits of slowing down the processor and a penalty for changing the operation power-mode is introduced together with a set of constraints that guarantee the real-time performance of the system. The meta-heuristics are the bio-inspired methods Particle Swarm Optimization and Genetic Algorithm. Experiments are made to evaluate the proposed algorithms using a set of synthetic systems of tasks. As these have been used previously with an Integer Lineal Programming approach, the results are compared and show that the solutions obtained with bio-inspired methods are within the Pareto frontier and obtained in less time. Finally, precedence related tasks systems are analyzed and the meta-heuristics proposed are extended to solve also this kind of systems. The evaluation is made by solving a traditional example of the real-time precedence related tasks systems on multiprocessors. The solutions obtained through the methods proposed in this paper are good and show that the methods are competitive. In all cases, the solutions are similar to the ones provided by other methods but obtained in less time and with fewer iterations.https://lib.jucs.org/article/22604/download/pdf/reward base schedulingmulticore systemsenergy |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Matias Micheletto Rodrigo Santos Javier Orozco |
spellingShingle |
Matias Micheletto Rodrigo Santos Javier Orozco Scheduling Mandatory-Optional Real-Time Tasks in Homogeneous Multi-Core Systems with Energy Constraints Using Bio-Inspired Meta-Heuristics Journal of Universal Computer Science reward base scheduling multicore systems energy |
author_facet |
Matias Micheletto Rodrigo Santos Javier Orozco |
author_sort |
Matias Micheletto |
title |
Scheduling Mandatory-Optional Real-Time Tasks in Homogeneous Multi-Core Systems with Energy Constraints Using Bio-Inspired Meta-Heuristics |
title_short |
Scheduling Mandatory-Optional Real-Time Tasks in Homogeneous Multi-Core Systems with Energy Constraints Using Bio-Inspired Meta-Heuristics |
title_full |
Scheduling Mandatory-Optional Real-Time Tasks in Homogeneous Multi-Core Systems with Energy Constraints Using Bio-Inspired Meta-Heuristics |
title_fullStr |
Scheduling Mandatory-Optional Real-Time Tasks in Homogeneous Multi-Core Systems with Energy Constraints Using Bio-Inspired Meta-Heuristics |
title_full_unstemmed |
Scheduling Mandatory-Optional Real-Time Tasks in Homogeneous Multi-Core Systems with Energy Constraints Using Bio-Inspired Meta-Heuristics |
title_sort |
scheduling mandatory-optional real-time tasks in homogeneous multi-core systems with energy constraints using bio-inspired meta-heuristics |
publisher |
Graz University of Technology |
series |
Journal of Universal Computer Science |
issn |
0948-6968 |
publishDate |
2019-04-01 |
description |
In this paper we present meta-heuristics to solve the energy aware reward based scheduling of real-time tasks with mandatory and optional parts in homogeneous multi-core processors. The problem is NP-Hard. An objective function to maximize the performance of the system considering the execution of optional parts, the benefits of slowing down the processor and a penalty for changing the operation power-mode is introduced together with a set of constraints that guarantee the real-time performance of the system. The meta-heuristics are the bio-inspired methods Particle Swarm Optimization and Genetic Algorithm. Experiments are made to evaluate the proposed algorithms using a set of synthetic systems of tasks. As these have been used previously with an Integer Lineal Programming approach, the results are compared and show that the solutions obtained with bio-inspired methods are within the Pareto frontier and obtained in less time. Finally, precedence related tasks systems are analyzed and the meta-heuristics proposed are extended to solve also this kind of systems. The evaluation is made by solving a traditional example of the real-time precedence related tasks systems on multiprocessors. The solutions obtained through the methods proposed in this paper are good and show that the methods are competitive. In all cases, the solutions are similar to the ones provided by other methods but obtained in less time and with fewer iterations. |
topic |
reward base scheduling multicore systems energy |
url |
https://lib.jucs.org/article/22604/download/pdf/ |
work_keys_str_mv |
AT matiasmicheletto schedulingmandatoryoptionalrealtimetasksinhomogeneousmulticoresystemswithenergyconstraintsusingbioinspiredmetaheuristics AT rodrigosantos schedulingmandatoryoptionalrealtimetasksinhomogeneousmulticoresystemswithenergyconstraintsusingbioinspiredmetaheuristics AT javierorozco schedulingmandatoryoptionalrealtimetasksinhomogeneousmulticoresystemswithenergyconstraintsusingbioinspiredmetaheuristics |
_version_ |
1721362426266386432 |