Software Defined Networking
Research Field Description
The main principle behind SDN is to provide an open interface to the forwarding hardware in networks. The goal is to be able to directly influence the forwarding process of a network element using a freely programmable control software, thus no longer relying on proprietary management and control systems. This has the prospect of leading to a faster pace of innovation in the network as well as more competition on the market reducing the costs for network operators. To achieve this, a pure SDN switch does not have any conventional control-plane functionality but fully relies on the external controller entity to make forwarding decisions. Current SDN realizations often rely on the OpenFlow protocol standardized by the Open Networking Foundation for the communication between data- and control-plane.
SDN: Interfaces, Attributes, and Use Cases
The term Software Defined Networking (SDN) is prevalent in today's discussion about future communication networks. Like with any new term or paradigm, however, no consistent definition regarding this technology has formed. Establishing SDN as a widely adopted technology beyond laboratories and insular deployments requires a compass to navigate the multitude of ideas and concepts that make up SDN today. Accordingly, we are currently working at a definition of SDN and a classification of interfaces, use cases and key attributes. We aim at mapping the interfaces and attributes to SDN use cases and at highlighting the relevance of the interfaces and attributes for each scenario. In the end, we want to provide a compass which guides a potential adopter of SDN, whether SDN is in fact the right technology for his arbitrary use case.
SDN-based Application-Aware Networking on the Example of YouTube Video Streaming
Application-Aware Networking is a promising approach to provide good application quality to users in scenarios with limited network resources, like today’s access networks. With SDN, a particularly interesting method to enable flowbased traffic management in networks has become available. In this work we take a look at how a specific application, i.e., YouTube Streaming, can benefit from such an SDN-based Application-Aware Network. We implement and investigate an approach based on Deep Packet Inspection (DPI) and one based on direct information input from the application in an OpenFlow testbed in order to show, how these different types of application information can be exploited to enhance the Quality of Experience (QoE). Furthermore, we determine the overhead caused by each of the presented approaches.
On the Accuracy of Leveraging SDN for Passive Network Measurements
Network Measurement has emerged as one promising field of application for Software Defined Networking. The reason for this is that the logically centralized control plane of an SDN network inherently has to aggregate network state information in order to function. This circumstance can be leveraged for network measurements at the SDN controller without the need for additional equipment or active – and possibly disruptive – measurements in the network itself. However, the accuracy and potential resource overhead of this approach has not been discussed. In this work we compare an SDN-based solution to actual traffic measurements in order to determine its accuracy and resource demand by performing tests in an OpenFlow testbed.
Pareto-Optimal Resilient Controller Placement in SDN-based Core Networks
With the introduction of SDN, the concept of an external and optionally centralized network control plane, i.e. controller, is drawing the attention of researchers and industry. A particularly important task in the SDN context is the placement of such external resources in the network. In this work we discuss important aspects of the controller placement problem with a focus on SDN-based core networks, including different types of resilience and failure tolerance. When several performance and resilience metrics are considered, there is usually no single best controller placement solution, but a trade-off between these metrics. We introduce our framework for resilient Pareto-based Optimal COntrollerplacement (POCO) that provides the operator of a network with all Pareto-optimal placements. The ideas and mechanisms are illustrated using the Internet2 OS3E topology and further evaluated on more than 140 topologies of the Topology Zoo.