Traffic Steering in SDN Based Cellular Networks

Abstract

To handle the increased traffic demands on cellular networks, ultra-dense and heterogeneous deployments are required. The natural evolution of the existing cellular networks will lead towards the multi-tier structure of overlapping coverage and heterogeneous resources pool. The spatial-temporal traffic dynamics and non-uniformity of traffic cause inefficient utilization of radio resources and energy. A strong desire of finding the way to make an optimal use of diverse network resources while providing satisfactory services to the users which has shifted the interest of researchers towards traffic steering in heterogeneous cellular networks. The concept of traffic steering, which intelligently redistributes traffic across the network entities, and designs optimal management schemes for the heterogeneous networks. The steering technique requires real-time network and user state information for taking the steering decisions which can be achieved by the promising software defined networking (SDN) platform. The idea of integrating SDN concept in cellular networks has gained much interest due to its up-to-date global view, logically centralized but physically distributed architecture, and programmability. In this thesis, various aspects of traffic steering technique and literature review on the integration of SDN in cellular network is presented. The key benefits and challenges of SDN integration in cellular network are briefly described in terms of network architecture design and management. A complete traffic steering framework for SDN based cellular network is proposed by considering both user’s and network current status. Finally, a traffic steering scheme is introduced which maximize network radio resource utilizations by redistributing the user’s traffic to the optimal destination. A comprehensive system level simulation has been conducted to represent the effectiveness of the proposed scheme.