A Grid Resource Allocation Method Based on Iterative Combinatorial Auctions

摘要

Whereas grids enable the sharing, selection, query and aggregation of geographically distributed resources for solving large-scale problems, providing efficient resource allocation mechanism managing grid resources is a complex undertaking. In this work, we consider the problem of engineering agents that act as proxies for the procurement of computational and data resources. Since applications may require multiple grid services, mechanisms such as single-good auctions may not be appropriate for the management of these services, we propose a new scheduling algorithm based on iterative combinatorial auction. We provide the design of prices announcement, agent creation and resource situation submission, bidding rules, and temporary allocation rules. A simulation environment is established based on the Gridsim toolkit and the simulation experiments indicate that the algorithm needs less communication and meet the requirements of userpsilas QoS better than the conventional algorithms.