Although the management of resources is essentialudfor realizing a computational grid, providing an efficient resourceudallocation mechanism is a complex undertaking. Since Gridudproviders and consumers may be independent bodies, negotiationudamong them is necessary. The contribution of this paperudis showing that market-driven agents (MDAs) are appropriateudtools for Grid resource negotiation.MDAs are e-negotiation agentsuddesigned with the flexibility of: 1) making adjustable amounts ofudconcession taking into account market rivalry, outside options,udand time preferences and 2) relaxing bargaining terms in theudface of intense pressure. A heterogeneous testbed consisting ofudseveral types of e-negotiation agents to simulate a Grid computingudenvironment was developed. It compares the performanceudof MDAs against other e-negotiation agents (e.g., Kasbah) in audGrid-commerce environment. Empirical results show that MDAsudgenerally achieve: 1) higher budget efficiencies in many marketudsituations than other e-negotiation agents in the testbed andud2) higher success rates in acquiring Grid resources under highudGrid loadings.
展开▼
