An Optimal Admission Control Policy for Cloud Computing Services with Tandem Queues Based on Game Theory

In a cloud computing system, each customer can pay for a service based on usage, and the service provider usually wants to maximize revenue while the cloud customer expects the best quality of service. If the service provider accepts all incoming requests, it may lead to resource overloading thus affecting the performance of the whole system. In this paper, to address this kind of problem, we propose a computer service admission control model with tandem queues based on game theory and verify that the optimal admission control policy for when to admit or reject an arriving customer request is a control limit (threshold) policy, thus maximizing the service provider’s revenue to achieve best performance and maximize the revenue of the whole system.