An Auction Based Profit-Aware Resource Allocation Mechanism for Cloud Computing

Abstract

This paper investigates the resource allocation mechanism in infrastructure as a service (Iaas) cloud computing, in which the resources are offered in the form of virtual machine (VM) instances to users based on the pay-as-you-go model. We aim to maximize the profit of cloud service provider (CSP) meanwhile ensure that the resources are utilized efficiently. We exploit the auction-based model and propose a polynomial-time approximation scheme to solve the VM allocation problem. In particular, users in the cloud are allowed to submit bids for the heterogeneous resources in the combination of VM instances, in order to determine the invoice the users have to pay by using the allocated resources, we design the system into an equilibrium so that the users do not have incentives to submit their hybrid VM requests and bids untruthfully. Extensive simulations show that the proposed mechanism can achieve near optimal allocation and resilient to the manipulation of untruthful competitions.