An Enhanced Framework for Traffic Load Balancing and QoS Provisioning in SDN

With the recent emergence of advanced network control and flexibility demanded by 5G networks, software defined networks (SDN) are seen as a viable solution. They provide the separation of control and data planes. It has a huge potential for a wide variety of end user applications. Load balancing and QoS provisioning are some of the applications that reap the benefits of the centralized network control that SDN provides. In our work the framework is divided into modules for congestion detection, load balancing and QoS provisioning. The major contributions to the existing literature are the integration of several network quality determining parameters like bandwidth, delay and reliability, efficient node and link disjoint routing and handling a mixture of QoS demands while not ignoring the best effort user traffic. All the results are plotted and suitable comparisons are done. Our framework performs better in elephant flow conditions and handles heavy load well. It also satisfies QoS requirements without starving the non-QoS user traffic.