Three-side Gaming Model for Resource Co-allocation in Grid Computing

Abstract

Co-allocation is a fundamental infrastructure to aggregate heterogeneous and distributed resources in grid environments. Although it has been studied extensively, co-allocation under the constraints to budget and deadline still remains an opening issue, which means that tradeoff between user QoS requirements and system performance should be agreed. In this paper, a novel agent-based two-phase co-allocation is proposed, which optimizes resources deployment and price scheme through a two-phase co-allocation mechanism, and applies queuing system to model the working of resources for providing quantitative guarantee for application’s deadline requirement. Extensive simulations are conducted to evaluate the effectiveness and performance of the model by comparing with other three co-allocation policies in terms of deadline violation rate, resource benefits and utilization. Experimental results show that the two-phase model can significantly improve the QoS satisfaction for those grid applications with constraints to budget and deadline.