Accurate Through-Wall Breathing Rate Measurements using a C-Band FMCW Radar

This dataset presents a collection of through-wall breathing rate (BR) measurements obtained using a custom-built C-Band Frequency Modulated Continuous Wave (FMCW) radar operating within the 4–6 GHz frequency range. The dataset was created as part of a research study on non-invasive biometric identification through physical barriers and is particularly valuable for researchers working on radar-based vital sign monitoring, signal processing, and artificial intelligence (AI) models applied to biomedical applications.

The dataset consists of breathing rate measurements from 10 individuals, including 5 males and 5 females. The radar signals were recorded through a 30 cm thick brick and cement wall at distances of 1m, 3m, and 4m. The data collection process was carefully designed to ensure high accuracy, with ground truth validation provided by the Biopac system (MP36R). The radar system used a center frequency of 5 GHz, a bandwidth of 2 GHz, a transmitted power of 20 dBm, and an antenna gain of 12 dBi. The chirp duration was set to 1 ms, and the sampling rate was 1 Msps, ensuring high-resolution signal acquisition.

The dataset includes raw radar signals (intermediate frequency signals), extracted phase signals for breathing rate estimation, corresponding ground truth breathing rates from the Biopac system, and both time-domain and frequency-domain processing results. These data points make the dataset valuable for applications such as radar-based biometric identification, through-wall sensing and security applications, healthcare monitoring with contactless breathing rate detection, AI-based signal processing for vital signs extraction, and the optimization of radar signal processing techniques.

In terms of performance, the FMCW radar system demonstrated high accuracy when compared to the Biopac reference measurements, achieving a Mean Absolute Error (MAE) of 0.07 BPM and a Root Mean Square Error (RMSE) of 0.09 BPM. Agreement analysis showed a strong correlation (R² ≈ 0.999) between the radar-based measurements and the ground truth data, confirming the reliability of the system for through-wall breathing rate detection.

This dataset was collected in compliance with ethical guidelines approved by the Ethics and Deontology Committee of the University of Aveiro (No. 10-CED/2023). All participants provided informed consent before data acquisition. The dataset serves as a valuable resource for researchers developing advanced non-contact monitoring systems and exploring new methodologies in radar-based physiological sensing.