Leveraging proactive flow to improve scalability in software defined networking / Oseghale Osemudiamen Victor

The concept of Software Defined Networking (SDN) is a concept which necessitates decoupling the network control plane from the data plane and makes each plane evolves independently. This controller architecture supporting versatile communication services is a complex paradigm. The need to coordinate the interaction of the control plane components such as QOS modules, routing information base, security policy, resource scheduling, etc. is a complex operation and based on the controller. The coordination of the actions of these components sometimes have conflicting goals and require careful handling. Some of these components exist in the network as distributed protocols and the inconsistency in coordinating these protocols could lead to critical problems in the network. Apart from the coordination of these distributed protocols, there is an issue that can arise due to flow request from data plane devices to the controller, which could eventually lead to a bottleneck as the network flow request scale at large network. The use of a centralized controller to address some of these challenges has been proposed by researchers. By centralizing the controller, the network simplifies the state distribution and hence have a better control over the network consistency. However, it is difficult to scale at larger network size and flow request. It is also important to maintain low latency in request handling and at the same time having high scalability throughput. For this thesis, an implementation of proactive flow programming on the data plane and improved I/O systems is proposed as leverage in the design of the SDN controllers. Through experiments, this work shows that implementing these concepts improve the controller scalability by enhancing the flow handling rate as the network size increases