Type something to search...
5G SRS Outdoor Dataset
5G SRS Outdoor Dataset

5G SRS Outdoor Dataset

Mohsen Ahadi
Florian Kaltenberger
Omid Esrafilian
Adeel Malik

This dataset contains raw Channel Impulse Response (CIR) measurements collected from uplink Sounding Reference Signals (SRS) transmitted by a 5G User Equipment (UE) in a real-world outdoor deployment at the EURECOM campus. The dataset enables research on localization, positioning, channel charting, and AI/ML-driven wireless network analysis using multi-antenna 5G measurements.

release

Detailed Description

This repository provides datasets containing raw Channel Impulse Response (CIR) measurements collected from uplink Sounding Reference Signals (SRS) transmitted by a 5G CrossCall phone acting as the User Equipment (UE). Measurements were collected using two Radio Units (RUs) from VVDN deployed at the EURECOM campus located at coordinates [43.6144094, 7.0711747]. Each RU is equipped with four spatially distributed SISO directional antennas with a 50° 3dB beamwidth in azimuth and elevation. The dataset is designed to support research in wireless localization, positioning, channel charting, and AI/ML-based radio analytics. It provides synchronized CIR measurements together with accurate RTK-based ground-truth UE positions.

Antenna Placement

The local Cartesian coordinate reference is antenna 1 on RU1 with:

  • Local coordinates: [x, y, z] = [0, 0, 2.2]
  • GPS coordinates: [43.6139498, 7.0714626, 170]

RU1 — EURECOM South Terrace

AntennaX (m)Y (m)Z (m)
1002.2
2902.2
32702.2
43602.2

RU2 — EURECOM North Rooftop

AntennaX (m)Y (m)Z (m)
1502011
2502511
3503011
4503511

Dataset Structure

The dataset contains T samples, M antennas, and NFFT CIR indices.

{
    time: [T×1 double]                # recording timestamp in milliseconds
    lat: [T×1 double]                 # true latitude from RTK
    lon: [T×1 double]                 # true longitude from RTK
    alt: [T×1 double]                 # true altitude from RTK
    local_x: [T×1 double]             # converted local x coordinate
    local_y: [T×1 double]             # converted local y coordinate
    cir_real: [NFFT×M×T double]       # real part of CIR
    cir_imag: [NFFT×M×T double]       # imaginary part of CIR
}

Example: Load in MATLAB

time = h5read('datasetname.h5', '/time');
Share :

Related Publications

From Concept to Reality: 5G Positioning with UL-TDoA in OpenAirInterface


Published at IEEE INFOCOM 2025
View Details

EURECOM 5G SRS Dataset


Published at IEEE DataPort
View Details

Experimental Insights From OpenAirInterface 5G Positioning Testbeds: Challenges and Solutions


Published at ACM WiNTECH 2025
View Details

An O-RAN Framework for AI/ML-Based Localization with OpenAirInterface and FlexRIC


Published at ArXiv
View Details

TDoA-Based Self-Supervised Channel Charting With NLoS Mitigation


Published at IEEE Transactions on Machine Learning in Communications and Networking 2026
View Details

Technical Details


Version1.0.0
Published07 Oct, 2025
Base
Platform:
OpenAirInterface
Data Format:
HDF5
Network:
5G Standalone
Hardware:
CrossCall 5G Phone, VVDN Radio Units, Directional Antennas

Authors


Adeel Malik
Adeel Malik
Florian Kaltenberger
Florian Kaltenberger
Mohsen Ahadi
Mohsen Ahadi
Omid Esrafilian
Omid Esrafilian

Tags


  • OpenAirInterface
  • O-RAN
  • Machine Learning
  • VVDN

Affiliation


  • Firecell
  • EURECOM
  • EURECOM
  • EURECOM

Comments

Sign in to join the discussion

More from Mohsen Ahadi et al.

5G NVIDIA Aerial Testbed SRS Indoor and Outdoor Dataset

5G NVIDIA Aerial Testbed SRS Indoor and Outdoor Dataset

Max Mayershofer
Florian Kaltenberger
Omid Esrafilian
Reem Bahsoun

This dataset contains raw Channel Frequency Response (CFR) measurements collected from uplink Sounding Reference Signals (SRS), the corresponding FAPI metadata, and the slot of fronthaul (FH) IQ samples where SRS was transmitted by a 5G User Equipment (UE). Data is provided for real-world multi-antenna 5G indoor and outdoor deployments using the Aerial Testbed at the EURECOM campus in scenarios with and without a dynamic scatterer present. In the indoor scenario this is a person walking through the cell, while in the outdoor setting it is a UAV/drone flying through the cell.

OpenAirInterfaceO-RANMachine Learning+1
release