Agent-based Grid Resource Negotiation Algorithm

This paper introduces the current development of the resource discovery in Grid and investigates adding the models in economic Grid to the resource discovery in agent system.Using feedback mechanism and mechanism with penalty and bonus limit the resource bargain.The contributions of this work include adding the specific module to the control structure in Grid to achieve negotiation.Among them,the control based on the feedback mechanism,and the mechanism with penalty and bonus make benefit to both the resource providers and consumers.When providers asking the price of resources too high or consumers bidding too low,the proposed module can adjust the resource price based on negotiation strategies.For solving the problem that some agents who are malicious and do not act as they claimed,in this paper,it is used the mechanism with penalty and bonus to limit the credibility of Agent.Namely,the bonus depends on the advance,and the penalty depends on the delay.Obviously,the mechanism with penalty and bonus can regulate the resource market and make its healthy development.Therefore,users can get better services.At last,this paper uses popular Grid simulation tool,GridSim,make the experiments of the algorithm proposed in this paper.Simulations illustrate that the feedback mechanism and mechanism with penalty and bonus can improve the efficience of Agent system in Grid.Through the negotiate solution,both the resource discovery success rate and system profit can be improved.