Strategies for Network Traffic Optimization: An In-Depth Analysis of Multipath Routing Algorithms in SDN

Demand for online services and cloud services has been increasing day to day. The usage of these services increases network traffic, which is difficult to handle by traditional networks. In the traditional networking paradigm, it is very difficult to integrate various networking devices of different versions. The programming capability in traditional networks is very less which requires manual installation even for small changes in the network. The traditional networking paradigm is a distributed architecture, in which individual networking devices will take control of the network devices which are under them. Overcoming the above limitations, a new networking paradigm called Software Defined Networking (SDN) came into existence. SDN possesses a centralized architecture in which the controller will take all the routing decisions and those instructions will be forwarded to the forwarding elements called switches. Through programming, networking devices can be configured which reduces manual installation. In this article, a detailed study on multipath routing in SDN is done. It provides a detailed study regarding different ways of implementing multipath routing. Multipath routing based on optimization techniques and machine learning techniques are discussed. This chapter consists description of how load balancing can be improved and how Quality of Service (QoS) can be improved using multipath routing. Multipath routing using shortest path algorithms like the Bellman-Ford algorithm, and Depth First Search (DFS) are discussed and how deep neural networks and reinforcement learning mechanisms will be used to implement multipath routing and to improve network performance is discussed. This study helps other researchers who try to do research based on multipath routing over SDN to get an overview of multipath routing.