README.md

metadata

license: cc-by-sa-4.0
DOI: https://doi.org/10.5281/zenodo.21225461
tags:
  - 5g
  - wireless-networks
  - vehicular-networks
  - simulation
  - sumo
  - omnetpp
  - simu5g

NeversNet5G

NeversNet5G is a city-scale 5G NR vehicular network dataset generated through SUMO and Simu5G/OMNeT++ co-simulation over the real urban road network of Nevers, France. The release contains processed per-UE event-level CSV files, not the raw OMNeT++ .vec/.sca outputs.

Release Contents

• data/: 716 per-UE metric CSV files, grouped by scenario part.
• configs/: OMNeT++/Simu5G configuration files used for the scenario. The released omnetpp.ini is trimmed to the NeversNet5G scenario and transition configs; generic Simu5G examples from the development workspace are omitted.
• metadata/: selected gNodeB/OpenCellID records and UE mapping files.
• scripts/: available scripts for figure generation and dataset inspection.
• manifest.csv: file-level manifest for all released per-UE CSV files.
• DATA_DESCRIPTION.md: dataset schema, units, and usage notes.
• CITATION.cff: citation metadata.
• LICENSE: dataset license and attribution notes.

Dataset Summary

Property	Value
Released data files	716 per-UE CSV files
Total rows	242,304,261
Approximate CSV size	26.4 GiB
Simulation duration	approximately 1,470 s
Restored UE trajectories	390
Maximum concurrent UEs	171
gNodeBs	19
Canonical schema	up to 23 columns
Radio/protocol metric columns	15

The raw OMNeT++/Simu5G vector and scalar outputs exceed 200 GB and are not included in this release. This archive provides the processed per-UE event-level CSV files used in the NeversNet5G dataset paper, together with configuration and metadata files documenting their generation.

Data Organization

data/
  part1/      119 files
  part1_5/     52 files
  part2/      171 files
  part2_5/     50 files
  part3/      171 files
  part3_5/     80 files
  part4/       73 files

Files are named:

<part>_ue_<id>_metrics.csv

Example:

part2_ue_047_metrics.csv

Important Note On Missing Values

Missing values in the CSV files are structural. They are not random missing measurements and are not failed simulator output. OMNeT++/Simu5G logging is event-driven: each row corresponds to a simulator event, and only metrics emitted by that event are populated. Users who need dense time-aligned matrices should apply a documented downstream alignment, interpolation, or forward-fill step appropriate to their task.

Recommended Citation

Use the citation metadata in CITATION.cff. Replace the DOI field with the final Zenodo DOI after publication.

License And Attribution

The dataset is released under Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA 4.0). The gNodeB metadata is derived from OpenCelliD records and must retain OpenCelliD attribution.