Title: Efficient job scheduling in cloud computing based on genetic algorithm

Authors: Shirin Hosseinzadeh Sahraei; Mohammad Mansour Riahi Kashani; Javad Rezazadeh; Reza Farahbakhsh

Addresses: Islamic Azad University, North Tehran Branch, Hakimiyeh, Shahid Babaee Highway, Tehran, Iran ' Islamic Azad University, North Tehran Branch, Hakimiyeh, Shahid Babaee Highway, Tehran, Iran ' Islamic Azad University, North Tehran Branch, Hakimiyeh, Shahid Babaee Highway, Tehran, Iran; University of Technology Sydney (UTS), 15 Broadway, Ultimo NSW 2007, Australia ' Institut Mines-Télécom, Télécom SudParis, CNRS Lab UMR5157, 9 rue Charles Fourier, 91011 Evry, France

Abstract: Scheduling in cloud is one of the challenging issues in resource management topic where the main question is how to manage time and cost in an optimised way. This study tackles the mentioned problem by managing time and cost through a genetic-based algorithm. The primary goal of this study is to manage jobs in a shorter time with lower cost and higher utilisation. Toward that end, we leverage the genetic algorithm solutions and a new model is proposed where jobs are created in genetic format. In the evaluation part of the model, different scenarios based on taking different fitness functions and format of the population are considered. We have analysed makespan, cost and utilisation in comparison to other two existing scheduling models (MAX-MIN and MIN-MIN). The results show considerable improvement in the cost, makespan and utilisation.

Keywords: cloud computing; job scheduling; genetic algorithm; cost; makespan; utilisation.

DOI: 10.1504/IJCNDS.2019.099968

International Journal of Communication Networks and Distributed Systems, 2019 Vol.22 No.4, pp.447 - 467

Received: 05 Sep 2017
Accepted: 07 Mar 2018
Published online: 03 Jun 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article