Load Optimization Accessions, Ramification the QoS in Software Defined Networking

It is difficult to manage traditional networks due to network growth, an increase in user numbers, and the emergence of new technologies like big data and cloud computing. Consequently, it is required to alter the current network design. SDN is a well-liked architecture because it offers customizable control and centralized management in data centers. To achieve scalability and dependability, it was necessary to propose the geographical dispersal of a logically centralized control plane due to the huge scale of networks. Software-defined networking (SDN) load-balancing optimization has been studied for a long time. Many approaches to the load-balancing conundrum have been put forth by researchers, but very few have taken the impact of transmission delay between controllers and switches under heavy network load into account. In this article we elaborate the various techniques of load balancing in SDN those working on different areas like multi-controller, switch migration, multi-agent, strategy, time sharing, prediction, reinforcement learning etc. We discuss the methods used along with the advantages and disadvantages of various costs, service quality, and network performance metrics.