Joint Optimization of Heterogeneous Cellular and Backhaul Networks for Services Awareness in SDN

In next generation network, types of services will be more diverse due to the increment of users demand. Traffic Engineering (TE), which solves the problem of performance optimization in network, cannot meet the QoS needs of different services because of the limitations of routing strategies in existing networks. Software-defined network (SDN) is a promising network paradigm that separates control plane and data plane, and enhances network flexibility. SDN brings benefits to TE via collecting and analyzing data dynamically. Existing mobile network solutions mostly focus on single network, such as radio access network or core network. In this paper, edge nodes in software-defined backhaul network are selected as the entrances of network flows from base stations in heterogeneous cellular network. An Access Selection Method (ASM) is proposed to choose access nodes by measuring the similarity between the service characteristics and performance of backhaul edge nodes with weighted Euclidean distance. Network flows traversing the edge nodes split into multiple sub-service flows among backhaul based on types of services. A Service Awareness Routing (SAR) scheme is proposed to minimize overhead from perspective of service combining ASM. The joint design with SAR and ASM solves entrances selection and network flowing routing problems in consideration of network services. Extensive simulations show the effectiveness and reliability of the proposed scheme.