Towards Dynamic Two-tier Load Balancing for Software Defined WiFi Networks

Software Defined Networking provides a promising evolution path by decoupling data plane and control plane to achieve more controllable and manageable networks. In this paper, we propose a dynamic two-tier load balancing scheme to achieve load balancing in software defined WiFi networks (SD-WiFi). User's traffic arrive at WiFi access points (APs), and these APs classify flows into real-time and non-real time classes. Three nominal features including flow size, service type and delay constraint of these flows are used to apply the Analytical Hierarchical Process method (AHP), for assigning flow processing to different controllers in large-scale SD-WiFi networks. These controllers are organized as two tiers: global and local. The global controller manages the load information of local controllers in form of clusters. This clustering is achieved based on their operation status which is periodically updated. We conducted simulation experiments to evaluate the performance of the proposed scheme using the OMNeT++ simulator. The proposed scheme outperforms the switch-based load balancing scheme by 39% higher in throughput and 24% lower in delay.