Utility-based price proportion in cloud resource allocation

As cloud computing is a new emerging distributed computing paradigm driven by economies of scale it is urgently to find better solutions for cloud resource allocation problem for economy reasons. Although there are lots of research efforts for cloud resource allocation, most of them introduce auction model to analyze the competition among cloud consumers in the condition of resource provisioned by means of indivisible VMs. For divisible resources, previous works are focus on minimizing the cost without considering features of diverse cloud consumers. In this study, utility-based price proportion approach for divisible resource is proposed. In the presented approach, utility is used to specify the features of diverse cloud consumers. With utility introduced, each cloud consumer is focus on the profit but not the cost, where the profit is the utility consumer gained from using resources minus the cost. Furthermore, experimental results show that the approach gains 21 and 15% more profit than common equal resource and cost minimization approaches separately with the same cost. © 2013 Asian Network for Scientific Information.