@article{4802,
  author = {Yao-Liang Chung},
  title = {Comprehensive Multi-Domain Analysis of Radar-Derived Vital Signs for Contactless Resting Respiration and Apnea Detection},
  journal = {Signals and Telecommunication Journal},
  year = {2026},
  volume = {15},
  number = {2},
  doi = {https://doi.org/10.6025/stj/2026/15/2/64-87},
  url = {https://www.dline.info/stj/fulltext/v15n2/stjv15n2_2.pdf},
  abstract = {Contactless radar sensing has emerged as a transformative modality for non invasive, continuous vital signs
monitoring. However, a rigorous multi-domain characterization of raw radar displacement signals under
both normal and disordered breathing conditions remains a fundamental requirement for developing reliable
clinical tools. This study presents a comprehensive multi domain analysis of radar derived vital signs, focusing
on the contactless monitoring of resting respiration and the detection of apneic events. Utilizing the
VITALSENSE 120 dataset, which comprises synchronized continuous wave radar recordings (sampled at
333.33 Hz) and clinical grade reference physiological signals, we systematically evaluated the radar
displacement signal across multiple analytical domains. Results indicate that resting respiration exhibits
stable periodic oscillations with a dominant respiratory frequency concentrated in the 0.1–0.5 Hz band,
alongside strong temporal synchronization with reference signals. Conversely, apneic events are distinctly
characterized by significant spectral power reduction in the respiratory band, marked envelope reduction
indicating diminished thoracic displacement, and pronounced autocorrelation reduction reflecting the loss
of respiratory rhythm periodicity. Furthermore, cross-domain analysis reveals a sharp drop in the correlation
coefficient between the radar displacement and reference respiratory signals, while residual cardiac induced
micromovements remain detectable. This multi-domain evidence base confirms the physiological
interpretability of radar observations and quantifies the discriminative signatures necessary for separating
spontaneous breathing from apnea, laying a solid foundation for robust, explainable AI/ML pipelines in
remote healthcare.},
}