Load Balancing on Software Defined Networks

Abstract

The development of cloud computing has made it possible to transport desired content to other locations via virtual servers. However, the fact that the traditional network devices are fixed and complex, it is difficult and costly to upgrade these devices to adapt to these developments. This inconsistency, led to the search for a new generation network instead of the traditional network structure. Software Defined Network (SDN) is a new generation network paradigm and separates data and control layers in network structure. With the separation of the control layer from the data layer, the management of the network resources can be performed by software. Thus, important tasks such as load balancing, traffic engineering and flow control can be carried out by loading on the controllers. From these components, the load balancing is a technique that divides the workload on the network into multiple sources and prevents overloading of a resource. In recent years, load balancing in SDN is one of the important research concerns of both industry and academia. In this paper, first load balancing problem in SDN is described and then the performances of two different load balancing algorithms (Round Robin and Dijkstra) are examined.