Controller Location and Load Balancing Integrated Solution

Abstract

The usage of multi-controller SDNs is currently the most efficient approach for constructing the core of communication networks of the fifth and following generations. One of the top priorities in the study of this topic is occupied by the optimisation of the network core construction since it involves relatively high expenses when developing communication networks of the fifth and future generations. Due to the complexity of the problems being tackled, there are currently a number of load balancing algorithms and algorithms for arranging controllers in multicontroller networks that are based on meta-heuristic methods. These algorithms allow for the optimum possible utilisation of controller resources in such networks. However, a comprehensive solution to the load balancing and controller placement issues has yet to be discovered. The answer to such an issue is the focus of this article. The report suggests using network clustering in conjunction with the meta-heuristic chaotic salp swarm technique, which has  shown to be effective in prior research on the challenges of creating multi-controller networks, to accomplish this goal. The salp swarm algorithm in the paper is adjusted to take into account the integral solution to the problem of deploying controllers based on clustering of a multi-controller network and load balancing. By contrasting the simulation results with those from the well-known meta-heuristic particle swarm algorithms optimization and the grey wolf GWO, as well as the previous version of the chaotic salp swarm algorithm CSSA, the effectiveness of the proposed solution was evaluated.