Nowadays more and more traffic demands appear in telecommunication networks, but due to the static nature of the lower layers it is difficult to act on these changing needs. The transmission links of transport networks need to be sized for peak traffic, but the lower layers would need more flexible operation and control. The concept of Software-Defined Networking is the separation of centered control, which guarantees the programmability of the network by real-time applications. It is worth dealing with the extension of programmability to transport networks, because joint control of the lower layers can increase resource utilization and reduce costs due to optimal path computation.
The purpose of my thesis is the presentation of the recent Transport SDN concept, as well as examination of the applicability in transport – e.g. optical or microwave – networks. Firstly I introduce the principles of SDN, and then I describe the advantages of extending the traditional solution into lower network layers. I summarize the changings that this technology brings to recent networks through the Transport SDN architecture and use-cases. There is no accepted standpoint for using transport-specific parameters so I analyze these parameters and present my suggestions. With this I demonstrate the benefits of Transport SDN by simulation in a traditional SDN environment. During the simulation I describe how to configure the network by actual bandwidth demands, delay or attenuation. Based on the evaluation of the tests I offer use-cases and opportunities to use SDN in transport networks.