Service Overlay Network Resource Adaptations Based on an Economic Model

In the current Internet, service overlay networks (SON) can provide a means for offering end-to-end quality of service (QoS) required by real time services such as VoIP, streaming multimedia and interactive games. We consider the approach where the SON operator leases bandwidth with QoS guarantees for the overlay links from Internet autonomous systems. Available bandwidth is managed by the SON admission and routing policy to provide end-to-end QoS connections to the service user. Maximizing profit is a key objective for the SON operator. In this paper, we propose a novel resource management approach which uses an economic model, including costs and revenues, to drive resource adaptations to changing network conditions, so that network profit can continuously be optimized. The approach integrates link capacity adaptations with connection admission control and routing policies based on Markov decision process theory. Numerical analysis results on a network example shows that the approach is effective at attaining maximized profit under varying network conditions