Network Data Sharing and Licensing

An Enabler for Autonomous Networking

Technical Backround

Autonomous networking is one of the key enablers for digital transformation and moving towards low-margin operation regime of telecom ecosystems. However, the scarcity of real field-collected network data is one the main showstoppers in their development. At Fraunhofer HHI, we gather and make such datasets accessible, to allow researchers and solution providers to develop and/or validate their solutions for the realization of smart and autonomous networks.

Features

  • Real field-collected data
  • Clean and standardized datasets
  • Task and use-case specific datasets
  • 4G/5G RAN datasets

Applications

  • Training and testing of machine learning algorithms
  • Validation of DSP algorithms through sampled waveforms
  • Development of network automation solutions

 

Fraunhofer Heinrich Hertz Institute hosts a 3-node microROADM metro network testbed that provides a great opportunity to perform real-field experiments including 5G-ready RAN infrastructure and edge compute capability for realizing low-latency end-to-end use-cases.


This testbed will host several demonstrations, including the final demonstration of Metro-Haul project focusing on improving the public safety. The purpose of this demonstration is to show that optical metro network needs to provide a high level of flexibility and transport capacity as well as ultra-low latency to enable real-time surveillance and analytics. We will show that network slicing and the use of edge compute capability are essential to improve public safety through video surveillance services in the future 5G networks. Moreover, the testbed is connected to the German metro ring network, which provides connectivity to our partners’ premises to perform field trials and optical transmission tests.

Metro network characteristics

  • 3 microROADM nodes
  • Edge compute capability per node
  • 400G coherent transponders and beyond
  • Filtered and filterless add/drop options

Edge cloud environment

  • Multi-rack edge cloud connected to RAN
  • Flexible container-based cloud environment
  • AI-enabled by NVIDIA DGX-1 deep learning platform

RAN characteristics

  • 2.6 GHz TDD LTE macro-cell
  • 2.6 GHz TDD indoor LTE small-cell
  • 3.7 GHz 5G macro-cell upgrade

Project and partners