Load Balancing Techniques in the SDN Data Center: A Review

Abstract

The ever-increasing number of internet users over the last decade has posed a serious challenge towards the available IP network structure. The limitations of the traditional networks at the data center to cope up with the high data requirements have posed a severe challenge to the resource management aspect. Software-Defined Network (SDN) has been emerged as a revolutionary technology which has changed the paradigm by improving the network management efficiency.It has emerged as a new networking framework where the architectural modifications have been done by separating the data plane and control plane. This logical decoupling between these two planes has provided an optimal network traffic management and flexible resource management. However, the uncertain and non-uniform data requirement at the data center results into a severe load imbalance among various controllers. Various load balancing techniques have been proposed by many researchers over the last two decades. These techniques varied from the conventional statistical framework to advance intelligent framework for various network environments.  However, the application of a specific technique to a defined network environment depends on different parameters and conditions. This paper presents a detailed survey of the various load balancing techniques for different network environment to encompass the complete outline for data center in SDN. It addresses the advantages, limitations and drawbacks of these load balancing techniques on the basis of different performance metrics.