A fault tolerance metaheuristic-based scheme for controller placement problem in wireless software-defined networks

Abstract

Software-defined networking (SDN) paradigm has become the master solver for traditional network restrictions. The fundamental concept of SDN is moving the control plane into one or more servers called controllers and confining the data plane to several forwarding network elements allowing dynamic and flexible network management. A critical challenging issue in SDN-based network architecture is the controller placement problem (CPP) that is an NP-hard problem. Trends of researches try solving the CPP based on evolutionary algorithms focusing on either propagation delay or balancing the load among distributed controllers. In this paper, we have proposed a fault tolerance metaheuristic-based scheme for CPP in wireless software-defined networks, named FTMBS. The aim of our proposed FTMBS scheme is as follows: maximizing the network connectivity, maximizing the load balance among controllers, minimizing the worst-case latency between controllers and associated nodes and between controllers themselves, and maximizing the network lifetime. When dealing with a multi-objective-based scheme, a trade-off exists mainly when these multi-objective metrics compete with each other, and it will be up to the decision maker to decide on this trade-off. We have considered a network of 500 randomly distributed sensor nodes and demonstrated various simulations for different network performance metrics and concluded that three controllers are enough for such networks. For performance evaluation, we have verified our solutions to be close to Pareto optimal ones provided by non-dominated sorting genetic algorithm-II (NSGA-II).