A Revenue Maximization Approach for Provisioning Services in Clouds

Abstract

With the increased reliability, security, and reduced cost of cloud services, more and more users are attracted to having their jobs and applications outsourced into IAAS data centers. For a cloud provider, deciding how to provision services to clients is far from trivial. The objective of this decision is maximizing the provider’s revenue, while fulfilling its IAAS resource constraints. The above problem is defined as IAAS cloud provider revenue maximization (ICPRM) problem in this paper. We formulate a service provision approach to help a cloud provider to determine which combination of clients to admit and in what Quality-of-Service (QoS) levels and to maximize provider’s revenue given its available resources. We show that the overall problem is a nondeterministic polynomial- (NP-) hard one and develop metaheuristic solutions based on the genetic algorithm to achieve revenue maximization. The experimental simulations and numerical results show that the proposed approach is both effective and efficient in solving ICPRM problems.