A Hybrid Multiprocessor Task Scheduling Method Based on Immune Genetic Algorithm

Mostafa Rahimi Azghadi; Mohammad Reza Bonyadi; Sara Hashemi; Mohsen Ebrahimi Moghadam

2nd International ICST Workshop on Artificial Intelligence in Grid Computing

Research Article

A Hybrid Multiprocessor Task Scheduling Method Based on Immune Genetic Algorithm

Cite: BibTeX Plain Text

@INPROCEEDINGS{10.4108/ICST.QSHINE2008.4263,
    author={Mostafa Rahimi Azghadi and Mohammad Reza Bonyadi and Sara Hashemi and Mohsen Ebrahimi Moghadam},
    title={A Hybrid Multiprocessor Task Scheduling Method Based on Immune Genetic Algorithm},
    proceedings={2nd International ICST Workshop on Artificial Intelligence in Grid Computing},
    publisher={ACM},
    proceedings_a={AIGC},
    year={2010},
    month={5},
    keywords={},
    doi={10.4108/ICST.QSHINE2008.4263}
}

Mostafa Rahimi Azghadi
Mohammad Reza Bonyadi
Sara Hashemi
Mohsen Ebrahimi Moghadam
Year: 2010
A Hybrid Multiprocessor Task Scheduling Method Based on Immune Genetic Algorithm
AIGC
ICST
DOI: 10.4108/ICST.QSHINE2008.4263

Mostafa Rahimi Azghadi¹^,*, Mohammad Reza Bonyadi¹^,*, Sara Hashemi¹^,*, Mohsen Ebrahimi Moghadam¹^,*

1: Department of Electrical and Computer Engineering, Shahid Beheshti University; MC, Tehran, Iran

*Contact email: m_rahimi@std.sbu.ac.ir, m_bonyadi@std.sbu.ac.ir, sa.hashemi@mail.sbu.ac.ir, m_moghadam@sbu.ac.ir

Abstract

Multiprocessor task scheduling plays a fundamental role in parallel applications and distributed networks. All of the methods for this kind of scheduling are concerned with achieving optimal running time. In this way parallel execution of tasks on several processors based on precedence graph should be considered. In this study, first a new heuristic method has been introduced which improved the execution time of some precedence graphs. Furthermore, we presented a novel immune genetic approach for multiprocessor task scheduling problem. Finally, combination of the proposed heuristic and the genetic approach makes a new hybrid scheme which is better than other well known and recent methods.

Published: 2010-05-16
Publisher: ACM
Modified: 2010-05-16

: http://dx.doi.org/10.4108/ICST.QSHINE2008.4263