Backhaul, QoS, and channel-aware load balancing optimization in SDN-based LTE networks

Abstract

Future cellular networks utilizes complex network technologies that makes it hard to achieve fine-grained traffic control. Moreover dynamic channel conditions alongwith finite backhaul capacity, limits the desired quality of service (QoS) objectives. This work aims to design a framework for software-defined cellular networks (SDCN) and suggests channel and QoS-aware load balancing procedures that jointly consider both access and back-haul networks. Based on the overall access and backhaul load information we first formulate optimization problems for both QoS and non-QoS users, respectively, considering their specific objectives. Then a realistic algorithm is proposed consisting of users scheduling, load estimation, handover decision, admission, and rate control procedures. To achieve optimal control of backhaul and access network segments in a timely efficient manner, the procedures are envisioned to run in a distributed fashion. Using our system-level SDN-LTE testbed developed in the network simulator (NS3), the proposed system is evaluated and compared with other state-of-the-art cell association algorithms that either consider the fairness in load distribution factor or just maximizes the end user rate while ignoring the load metric.