Context-Aware Traffic Forwarding Service for Applications in SDN

Traffic forwarding is a fundamental function in network operations. And forwarding traffic dynamically according to the current situation is attractive to network providers. With its programmability, Software-Defined Networks (SDN) provide network providers dynamic traffic forwarding ability. But SDN only provides trivial flow table to control traffic forwarding that is not an effective way for applications above SDN architecture. Also, SDN does not percept the situation itself. A more abstract and context-aware traffic forwarding scheme for applications in SDN is needed. In this paper, we propose a context-aware traffic forwarding service for applications in SDN. We firstly propose a traffic forwarding service composition framework special for SDN that utilize the advantages of SDN. We conclude that traffic forwarding service composition problem in SDN is essentially a multi-constrained optimization path problem. And based on this conclusion, a graph-based model and service composition algorithm are applied to the service composition scheme. To take the dynamic context factors in SDN into consideration, a regression analysis method is adopted. Finally, we offer an evaluation to validate the methodology. The results of the evaluation demonstrate that our approach is context-aware.