Performance Analysis of Load Balancing Mechanisms in SDN Networks

Abstract

The emergence of Internet of Things (IoT) and 5G networks makes a great pressure on the underlying communication networks in terms of demand, performance requirements and dynamic management. The traditional networks routing protocols forward traffic according to the shortest path to minimize cost. This may lead to abnormal distribution of traffic and cause the overloading of the links. Recently Software Defined Network (SDN) has become increasingly popular and a praised candidate to overcome traditional networks limitations. SDN have a centralized forwarding decision and a global view of the network. These features allow the establishment of strategies that manage traffic flows according to their nature and their capacity requirements. In this perspective, we highlight the use of SDN load balancing mechanisms in increasing networkÃŢs efficiency and in enhancing its capacity to meet the QoS requirements. To this end, we compare three different types of controllers: Hub-like controller, Spanning tree protocol (STP) based controller, and a load balancing based controller, in order to establish the impact of load balancing on network performance indicators. The results show that the networks links congestion is reduced improving thereby the network's overall performance.