A Dataset of 120 GHz Millimeter-Wave Radar Vital Signals With Synchronized Reference Recordings
Ruochen Wu β,
Β Β
Laura Miro
,
Β Β
Albert Aguasca
![]()
Cosme Garcia
Β Β
Antoni Broquetas
Β Β
Montse Najar
![]()
IEEE DataPort database: A New Dataset for Millimeter-Wave Radar Vital Sensing With Reference Signals
Catalan Open Research Area. Repositori de Dades de Recerca (Alternate access): 10.34810/data2962
Important
The study was approved by the ethics committee of the Universitat Politècnica de Catalunya · BarcelonaTech (Identification code: 2024-028). Informed consent for participation and data sharing was obtained and signed in accordance with the ethical principles governing the study. All data have been fully anonymized, ensuring that no information can be traced back to individual participants.
The file structure is
VITALSENSE_120_DATASET
βββ VS01
β βββ VS01_Resting.mat
β βββ VS01_Resting_Mindray.mat
β βββ VS01_Apnea.mat
β βββ VS01_Apnea_Mindray.mat
...
βββ VS24
β βββ VS24_Resting.mat
β βββ VS24_Resting_Mindray.mat
β βββ VS24_Apnea.mat
β βββ VS24_Apnea_Mindray.mat
βββ Subject Information.xlsx
βββ Subject Info
βββ Radar Parameters
βββ Reference Parameters
The dataset contains measurements from 24 subjects, with a total of 4 minutes per subject across the Resting and Apnea scenarios. The data and parameters are stored in lightweight .mat files that can be read directly in MATLAB, and the entire dataset is about 31 MB.
In this repository you can find the reference code used for the technical validation of breathing and cardiac signals sparation (VS_separation) and synchronization study (TechValidation). In addition, the script PlotVS for data visualization is also included in the repository which can be used by configuring the subject and/or scenarios which shall be viewed.
In particular, this repository also contains the code for the validation of the radar micro-motion sensing part in the paper (Test_speaker), which contains the entire process from the raw beat tone signal to the target displacement signal extraction. The raw data (Loudspeaker) acquired during the test measurement is also included.
The mmWave radar measured data were recorded by means of a high speed digitizer, which is managed by a C++ based executable. The validation code was written and tested using MATLAB R2024b for Microsoft Windows.
@article{wu2026dataset,
title={A dataset of 120 GHz millimeter-wave radar vital signals with synchronized reference recordings},
author={Wu, Ruochen and Miro, Laura and Aguasca, Albert and Broquetas, Antoni and Garcia, Cosme and Najar, Montse},
journal={Scientific Data},
year={2026},
publisher={Nature Publishing Group UK London}
}@data{wq68-sv85-25,
doi={10.21227/wq68-sv85},
url={https://doi.org/10.21227/wq68-sv85},
author={Ruochen Wu and Laura Miro and Antoni Broquetas and Albert Aguasca and Cosme Garcia and Montse Najar},
publisher={IEEE Dataport},
title={A New Dataset for Millimeter-Wave Radar Vital Sensing With Reference Signals},
year={2025}
}