Evaluating the economic efficiency of utility computing auction using a simulation model

Utility computing was a pilot program started by IBM two years ago. It uses computer network to link service providers and customers allowing customers to conduct their computer applications online. The technology aims to provide customer a cost efficient way of conducting their computer application without the need to purchase computer software. Customers conduct their computer application using vendors´ platform paying for whatever they consume. With its implementation. IBM discovered that they have to deal with the problem of insufficient buffer space when customer traffic is high and under utilized buffer space when customer traffic is low. The irregularity of customer traffic makes it impossible for vendors to price their resources cheaply and therefore less desirable to the customers. The proposed solution is to sell utility computing through an auctioneer where vendors come together to bid for the customers. The auctioneer allocates customers to the vendor with the lowest bid. Using an auctioneer to conduct transactions, it was hypothesized that resources will be efficiently utilized and prices will be significantly lower. To test the hypothesis, two simulation models were created to depict two systems, one in the presence of an auctioneer another without. Using specified indicators to measure the quality of the system, the project discovers that an auctioneer improves efficiency through better resource management. The results obtained showed that consumer and producer surplus increase insignificantly in an auction. Further analysis showed utility computing auction is also capable of serving a larger population.