A truthful and budget-balanced double auction model for resource allocation in cloud computing

This study aimed to improve the truthfulness and efficiency of double auction mechanism in the cloud computing environment. In an auction, bidders have a high incentive to announce an unrealistic price, that is, each cloud user tends to announce a lower price and each cloud service provider tends to bid higher price than the actual one. This untruthful behavior will have catastrophic consequences to the market, leading to participants' fear of being misused. Therefore, it seems vital to provide a mechanism for the realization of truthfulness involving no unrealistic offer by the users. On the other hand, it has been proven that it is not possible to achieve the characteristics of truthfulness, budget balance and economic efficiency at the same time. There is a weak budget balance in a truthful mechanism with a price difference greater than zero between the total price paid by the cloud users and that received by the cloud service providers. In this paper, a mechanism is proposed to encourage participants to offer truthful prices by rewarding honest participants. The proposed called the reward mechanism uses a repository to store bidders' information, including the number of bids and the number of honest bids per auction as well as a ranked list of bidders based on their bidding history in offering honest prices. The results show that the proposed mechanism not only motivates the bidders to announce honest bids, but also solve the problem of weak budget balance by dividing the bid amount among the winning bidders.