Economic-based resource allocation for reliable Grid-computing service based on Grid Bank

  • Authors:

  • Wei-Chang Yeh;Shang-Chia Wei

  • Affiliations:

  • Advanced Analytics Institute, School of Software, Faculty of Engineering and Information Technology, University of Technology Sydney, PO Box 123 Broadway, NSW 2007, Australia;Integration and Collaboration Laboratory, Department of Industrial Engineering and Engineering Management, National Tsing Hua University, P.O. Box 24-60, Hsinchu, 300, Taiwan, ROC

  • Venue:
  • Future Generation Computer Systems
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

The Grid-computing service united by numerous distributed and heterogeneous resources supplies various advanced and cumbersome problems with high-performance computing. Based on reciprocal transactions of the Grid Bank (Barmouta and Buyya, 2003 [16]) we present an economic-based resource allocation model to derive the service reliability of Grid-computing from cellular automata Monte-Carlo simulation (CA-MCS) for the service level agreement, and to evaluate total rental-time cost of Grid resources by virtual payment assessment for the free rider problem. Regarding the probability of the task completion, this paper converts the Grid system into the multi-state unreliable network in advance, and then the transforms network facilitates the CA-MCS to simulate the service reliability. To economize on total rental-time cost and ensure the Grid-computing service being dependable, this paper proposes a binary-code Genetic Algorithm (bGA) and an integer-code Particle Swarm Optimization (iPSO), in which both consider Elite-selected and Reborn (ER) mechanisms, to explore the best resource allocation in the light of cost-effectiveness and guaranteed reliability. Finally, the experimental results concerning optimal resource allocation of virtual Grid system have proven that the ER-bGA outperforms the ER-iPSO in terms of solution quality based on the test of statistical significance.