Automated Bidding in Computational MarketsAn Application in Market-Based Allocation of ComputingServices

Autonomous agents are widely applied to automate interactions in robotics,e.g. for selling and purchasing goods on eBay, and in financial markets, e.g. in theform of quote machines and algorithmic traders. Current research investigates efficienteconomic mechanisms that fully automate the provisioning and usage processes of Gridbasedservices. On the one hand, consumers want to allocate resources on demandfor their various applications, e.g. data sharing, stream processing, email, businessapplications and simulations. On the other hand, providers of Grid services want togovern business policies to meet their utilization and profit goals. The above-mentionedprocesses are not manually manageable, however, because decisions need to be takenwithin milliseconds. Therefore, such processes have to be automated to minimize humaninteractions. Hence, market mechanisms and strategic behavior play important roleswhen it comes to achieving automated and efficient allocation of Grid services.The paper begins by presenting a framework for automated bidding, providing amethodology for the design and implementation of configurable bidding strategies. Second,it presents a novel bidding strategy based on a reinforcement learning technique.This strategy is designed to automate the bid generation processes of consumers andproviders in various market mechanisms. Third, the behavior and convergence of thestrategy is evaluated in a centralized Continuous Double Auction and a decentralizedon-line machine scheduling mechanism against selected benchmark bidding strategies.Fourth, we define a bidding language for communicating consumer and provider preferencesto the market as well as report back the match of the market-based allocationprocess.