Bid-Proportional Auction for Resource Allocation in Capacity-Constrained Clouds

Currently most commercialized cloud providers mainly offer cloud computing resources by the fixed price approach. However, this may not be efficient for cloud resource usage especially when the total computing capacity is limited. Since the auction mechanism is already shown to be an approach to many resource-limited problems, how to apply such a mechanism to the cloud computing environment for improving revenues and resource utilization of providers is worth investigations. In this paper, we have used a bid-proportional auction model which is capable of adaptively adjusting resource price. We also study the decision of the optimal bid among all users in this environment. Finally, we conduct two simulations to validate the operation of the proposed model in dynamic and stochastic demand environments. Furthermore, several remarkable observations of the relationships between users and providers regarding price dynamic are elaborated.