Auction based Resource Allocation Strategy for Infrastructure as a Service

Abstract

The cloud providers provide various computational resources like CPU power, memory, and bandwidth to the users through Virtual Machines (VM). Various VMs differ in their resource capacity. Each cloud provider has a set of VMs which are available to users at different prices. Providers have a limited set of resources. Therefore, it is necessary to efficiently allocate available resources to the rational users and minimize the cost of resources as to gain maximum profit from the available resources. Allocation of Resource in cloud computing is one of the most challenging issues in resource management for both cloud and grid computing. In this thesis, a Vickrey-Clarke-Groves (VCG) auction based approach is used to deal with the problem of resource allocation at Infrastructure as a Service (IaaS) layer of cloud among rational users and to incentivize selfish providers. The proposed mechanism uses a reverse auction, which efficiently allocates resources, calculates payment, and also achieves social benefit among the users. The mechanism tried to solve resource allocation problem with payment calculation. The proposed mechanism fulfills properties of incentive compatibility, individual rationality and maximizes social welfare. The main target of the service providers is to maximize the utilization of their resources, maximizing the profit through minimizing the cost of the resources, achieve a social benefit, and maximizing the efficiency of the whole process. The proposed approach has been simulated using CloudSim and NetBeans IDE. Experimental analysis has proved the effectiveness of the system. This model is compared with fixed price model and has given a better performance in terms of pricing and utility.