The need to more flexibility in the management and control of today’s networks led to the development of new networking paradigms like Software-Defined Networking (SDN) and Network Function Virtualization (NFV). The SDN paradigm introduces a more flexible control of networks by separating the network control and data plane, and logically centralizing the control plane on the so-called SDN controller. The interaction between control and data plane is typically realized by a communication using a standardized protocol like OpenFlow.
The application of this SDN paradigm in operational networks assumes that the reliability and availability is ensured. For today’s network operators, this is one of the most important criteria when introducing SDN. To this end, therefore, sophisticated measures and innovative mechanisms are needed to ensure the required reliability and availability.
A recent idea in this context is based the following saying: “Fail fast is more important than fail-safe”. This means, it is more important to detect errors quickly than to build a system that is 100% accurate and faultless. The latter is de-facto not possible in an operational system, which constantly changes and receives updates. For such an approach enhanced analytics of the network including its devices is a mandatory requirement as it has to be known exactly which component has failed.
The objective of the project is to study and analyze of fast fault detection and localization techniques for SDN.