DigSiNet Network Digital Twin prototype presented at CNSM 2024 conference

Parts of the work were fortunately supported by an internal research project at Fulda University of Applied Sciences.

The prototype makes it possible to start several twins of a real network. In this way, experiments such as performance or error analyses, as well as "what-if scenarios" can be carried out using historical network management and monitoring data.

Various dockable interfaces, builders, controllers and apps can be used to operate the siblings with customised topologies of the real network. Containerlab, Arista, cEOS and gNMI are primarily used as reference implementations.

The extension enables the assessment of various network management and monitoring data and assesses their aptitude for practical network digital twin setups. Particularly challenging in this area is the granularity in terms of volume and speed of the data.

Other authors of the paper "Challenges of Event-based Streaming and Queuing as Data Exchange for Network Digital Twins" (see https://lnkd.in/eBygvraJ) are the students Sven Schickentanz, David Hermann and Thomas Mott.

They have actively contributed to extending the prototype with message queues (RabbitMQ) and streaming platforms (Apache Kafka). These form an essential component of scalable digital twins of network environments and are recommended by the IRTF in its NDT architecture design(https://lnkd.in/e2i3vaaM).

Prof Dr Rieger would like to thank the entire team behind DigSiNet, which is publicly available as open source https://lnkd.in/emGuZ5fB.

Furthermore, he would like to thank all participants and especially the organisers of the excellent CNSM 2024 conference and the NDT4IN workshop for the valuable feedback on the project, e.g. in the reviews and discussions!