DLBA - A Dynamic Load-balancing Algorithm in Software-Defined Networking

Abstract

Networks have become increasingly important in communication as it has evolved tremendously. There have been many changes in networks over the years. Currently, networked systems are requiring multiple packets to be transferred simultaneously from different sources to different destinations on the same network. To do the same, several efficient Heuristic algorithms are available to identify the best paths. Nevertheless, parameters like delay constraints and increased throughput must be considered in real-time scenarios since Quality of Services (QoS) is the major metric in networks. This research work has considered two real-time topologies - Fat Tree topology and Abilene topology and have demonstrated dynamic load balancing using the $k$ nearest path algorithm in the software-defined networking (SDN) using OpenDaylight (ODL) controller to ensure an increase in throughput and reduction in end-to-end latency.