Pricing Of Cloud Resources Based On Option Pricing Model And Auction Mechanism

The fast development of Cloud computing has brought the third revolution of information technology. The way people get access to information, utilize software resources, purchase hardware assets will be radically changed. Cloud computing provides users with large-scale computing power, storage space, development platforms and software services through the Internet. The resource utilization process will become relatively flexible. Users only need to pay a certain amount of fee the provider required and then get the permission to use the resources anywhere connected to the Internet. It is both simple and efficient. In addition, the price users pay for the resources they need would be much cheaper than the price they pay for buying the resources themselves. Users can also choose their own bundle of resources according to their distinctive needs and can take advantage of high performance processing power the provider offered to shorten their processing time.For different kinds of users, resource providers full range of IT services:Infrastructure as a Service:IaaS, Platform as a Service:PaaS, and Software as a service:SaaS. The goal of designing a pricing scheme is to achieve effective and rational allocation of resources between service providers and users. Resource providers will therefore adopt different pricing schemes for various users. There are three schemes most commonly used:Reserved Scheme, On-Demand Scheme and Spot Scheme.In this paper, we regard the Cloud Computing server resources as a whole and consider the devaluation characteristics of server resources caused by the Moore’s Law to establish an option on the resources. We treat the pricing of the Reserved Scheme and On-Demand Scheme of cloud computing as an option pricing problem. The pricing of the Reserved Scheme can be solved as a European option problem and the pricing of the on-Demand Scheme can be solved as an American option problem. We adopt the method of Trinomial tree to resolve the European option problem and the American option problem. The simulation results shows the model we propose can calculate the pricing of those two schemes according to the investment and contract time. In addition, we analyze the influence of various parameters on the prices and summarize the key factors needed to be considered in the pricing process.For the Spot Scheme, we design the auction model based on the VCG scheme. All the resources are divided by category in the first place, each category has a limited number of instances, users can select any combination of categories with any amount of instances less than the total amount. The model has been proved to be a dominated strategy and has the property of incentive compatible which can ensure the competitiveness and fairness of the auction process. In the specific resource allocation process, we introduce weight for different resource category according to their actual value. Then we propose an efficient allocation mechanism based on dynamic programming. Experimental result shows that the algorithm can maximize the value of allocated resources which would increase the overall revenue for resource providers.