A modified genetic algorithm for controller placement problem in SDN distributed network

In this paper, a framework of Controller Placement Problem (CPP) is implemented for dynamic mapping and switch migration across controllers in SDN distributed network architecture. Then a heuristic multi-level capacitated CPP (MCCPP) is formulated to efficiently assigned each switch to a corresponding controller to improve the reliability and scalability of the SDN system. The proposed method simultaneously identifies the optimum number of controllers, minimizes the delay, and ensures load balancing among controllers. Then a Genetic Algorithm (GA) is integrated with the allocation components to improve the controller locations. The effectiveness of the GA-based MCCPP is compared against heuristics CCPP over different network topologies. The adaptive GA-based yields excellent results in latency and only requires a few iterations to attain convergence