Optimizing flow setup time in software defined network

To address the issues of scalability in Software Defined Network (SDN), researchers have proposed a distributed control plane, where controllers are physically distributed but logically centralized. Network traffic is dynamic with respect to time and space, however, the key limitation of state-of-art work is that they make the unrealistic assumption that the traffic matrix is static. In this work, we assume that the network traffic is dynamic and to optimize the load on the SDN controller, we propose adaptive mapping between switches and the controller. The objective is to minimize the flow setup time as well as reduce the cost of reassignment.