Exploring the role of paths for dynamic switch assignment in software-defined networks

The distributed control plane of Software-Defined Networks improves the scalability of the control plane and enables to dynamically assign switches to controllers. However, existing works calculate the assignment mainly based on the number of flow path setup requests generated from the switches and neglect the different resource demand of the requests on the controllers. To address this problem, we analyze the controller resource demands of the requests of different paths and formulate a flow path-aware switch assignment problem to minimize the controller resource demand. To efficiently solve the problem, we propose a heuristic algorithm which is inspired by the rearrangement inequality theory to accommodate to the frequent flow fluctuation. Simulation results show that our solution can save the controller resources significantly compared with the existing solutions.