Forming resource-sharing coalitions: a distributed resource allocation mechanism for self-interested agents in computational grids

  • Authors:
  • Linli He;Thomas R. Ioerger

  • Affiliations:
  • Texas A&M University, TX;Texas A&M University, TX

  • Venue:
  • Proceedings of the 2005 ACM symposium on Applied computing
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

Designing efficient resource allocation mechanism for computational grids is extremely challenging because the effective agents in computational grids are inherently self-interested due to their different ownerships. Providing incentive for agents to share their resource with others is the key to make computational grids realistic. The global efficiency should be generated through the interactions among agents from the bottom up. In game theory, forming coalition is such a cooperative game among self-interested agents. We develop a distributed resource allocation mechanism for computational grids by forming resource-sharing coalitions among self-interested agents through automated multiparty negotiation. This mechanism is based on a task-oriented mechanism for measuring the economic value of computational resource usage. The simulation results show that the self-interests of agents in computational grids have considerable impact on the decisions of each agent about how to allocate their resource to appropriate tasks.