A Dynamic Pricing and Bidding Strategy for Autonomous Agents in Grids

In this paper, we propose a dynamic pricing strategy which is used for a market-based resource allocation mechanism in a local Grid. We implement an agent based Grid economy in which the decision-making process regarding task and resource allocation is distributed across all users and resource owners. A Continuous Double Auction is used as the platform for matchmaking where consumers and producers meet. In this paper, we analyze the parameter regime of this pricing mechanism considering different network conditions. Our experiments described in the paper show that using the pricing parameters, the consumers and producers agents can decide the price to influence the way they contribute resources to the Grid or complete the jobs for which they need resources. These agents are individually capable of changing the degree of their task usage and resource contribution to the Grid.