Zhang Yang


Year of Graduation:

Project Title:
Computational Resource Management using Auctions




Area(s) of Expertise:
Mechanism design; Network modeling and simulation

Research Keyword(s):
Auction; Mechanism design; Resource management

About My Research:

The auction theory developed in Economics has recently been applied to solve various problems in engineering. An auction process works as follows. First, buyers submit bids for purchasing commodities, and sellers submit asks to sell commodities. The auction commodity may be an actual item or a service. Each bid can contain a bid price and quantity of commodity to be purchased. Similarly, each ask can contain an ask price and quantity of commodity to be sold. The bids and asks are collected by an auctioneer who clears the market by matching bids and asks. Thereafter, every matched buyer-seller pair must deal at a clearing price. The buyer and seller assess the commodity involved, and have a private reserved valuation on it. Basically, buyers will not bid higher prices than their valuations, and sellers will not ask lower prices than their valuations.
In the context of cloud computing, the buyers are computational resource users who can use cloud services by making payments, and the resource sellers (i.e., Google and Amazon) provide the services to buyers and receive monetary gains from buyers. The commodities here are the cloud services (i.e., computational resources). In addition, to gain revenue, the auctioneers can charge commission fee from arranging auction for both seller and buyer side. A marketplace named computational resource exchange can be further set up.