IMMERSE DATASET

This repository contains raw sub-6 GHz and mmWave measurement data gathered for and evaluated in our paper Machine Learning-aided Sensing in Private mmWave Networks for Industrial Applications.

If you use this dataset, please cite our original work in Citation.

Measurement Setup

Our paper presents details about the indoor measurement setup used for gathering sub-6 GHz and mmWave traces for LOS and passages of pedestrians and AGVs (automated guided vehicles) at different tracks.

Dataset Structure

The dataset provides CSV-formated measurement files.

    csv
     |- mode (los, user passages @ track)
      |- measurment ID
       |- modem (UE_A, UE_B, UE_C)
        |- channel metrics (prx, drx for sub-6 GHz (4G), mmWave (5G))    

For more information, see our paper.

Acknowledgements

This work was funded by the German Federal Ministry of Education and Research (BMBF) in the course of the 6GEM Research Hub under the grant number 16KISK038.

Quick start

We provide scripts for Python and Matlab, demonstrating how to parse and plot the provided measurement data.

Python

Run the following commands to get started.

1. Clone this repository:

   git clone https://github.com/tudo-cni/immerse_dataset

2. Change into repo directory:

   cd immerse_dataset
   

   Optional: Create virtual environment

    python venv venv

   Activate virtual environment (more information). On Unix and MacOS run:

    source venv/bin/activate

   On Windows run:

    venv\Scripts\activate

   Install dependencies:

    pip install matplotlib

3. Running main.py -h shows the help message with optional filtering parameters. There is no filtering if no arguments are passed.

   python main.py -h
   
   usage: main.py [-h] [--modems MODEMS] [--metrics METRICS] [--modes MODES] [--tracks TRACKS]
   
   Parse and plot trace data
   
   optional arguments:
      -h, --help         show this help message and exit
      --modems MODEMS    {"UE_A", "UE_B", "UE_C"}
      --metrics METRICS  {"5G_drx_rsrp", "5G_prx_rsrp", "4G_prx_rsrp"}
      --modes MODES      {"agv", "pedestrian", "los"}
      --tracks TRACKS    {"track1", "track2"}

4. Example: Filter 5G_drx_rsrp trace data of UE_A and UE_B for agv passages on track1:

   python main.py --modems UE_A,UE_B --metrics 5G_drx_rsrp --modems agv --tracks track1

Matlab

If you use Matlab, simply navigate into the project dir and run the main.m script.

Citation

If you use this dataset or results in your paper, please cite our work (author's version) as:

@InProceedings{haferkamp2024b,
	Author = {M. Haferkamp, S. H{\"a}ger, S. B{\"o}cker, and C. Wietfeld},
	Title = {Machine Learning-aided Sensing in Private {mmWave} Networks for Industrial Application},
	Booktitle = {IEEE Globecom Workshops (GC Wkshps)},
	Address = {Cape Town, South Africa},
	Month = dec,
	Year = {2024},
	Project = {6GEM},
}