Solving Multi-Controller Placement Problem in Software Defined Networks with A Genetic Algorithm

Software Defined Networking (SDN) is a new generation network paradigm. The main purpose of SDN is to simplify network management and to achieve network communication in quickly manner by separating control and data planes using a single central controller. But this only one central controller in the SDN and the scalability of the device decreases as the network size increases for flow processing. This problem can be solved by using multiple controllers. However multiple controller placement is a very difficult problem. The number and the location of the multiple controllers should be determined so that the latency between the controllers is minimal. In this paper, a genetic algorithm based on Dijkstra algorithm is presented for CPP in SDNs. The proposed genetic algorithm is applied to the Ulaknet and Colt data sets which are within the Topology zoo database. In these experiments where the end-to-end latency is minimized, the number of controllers is determined to be three. The obtained results are shown on the map of Turkey and Europe.