Researchers have developed the method for environments where reflections from walls, furniture, and moving objects often overwhelm physiological signals.
Get all the details: Grab your PDF here!
Non-contact monitoring aims to track vital signs, such as respiratory rate and heart rate, without attaching physical sensors to the body.
Radar systems are well-suited to this task because they can detect tiny chest movements using electromagnetic waves, without relying on light, temperature, or cameras.
Frequency-modulated continuous wave (FMCW) radar is particularly attractive because it can extract both distance and phase information, allowing simultaneous localisation of a subject and measurement of physiological motion.
However, real-world environments pose serious difficulties. Strong reflections from nearby objects can mask the true subject, while traditional clutter-suppression methods, such as moving target indication (MTI), often remove the slow chest movements that carry vital-sign information.
Even when a subject is correctly identified, heart-rate estimation remains vulnerable to interference. Breathing produces strong periodic motion, and its harmonics frequently leak into the heart-rate frequency band, distorting or obscuring the much weaker cardiac signal.
Two Algorithms Work Together to Target Two Key Failure Points
To address these problems, the researchers introduced two algorithms designed to work together.
The first, the Matrix Coefficient Selection Method (MCSM), focuses on distance detection. Rather than relying on signal amplitude alone, MCSM operates in the range-Doppler domain, suppressing static clutter and non-periodic low-frequency interference.
By applying low-frequency zeroing and autocorrelation, the method identifies the distance bin most likely to contain periodic vital-sign motion, improving both distance accuracy and stability in multi-target scenes.
The second method, Recursive Least Squares Respiratory Harmonic Suppression (RLSRHS), tackles the frequency-domain problem.
Inspired by harmonic mitigation techniques used in power systems, the approach applies adaptive recursive least-squares filtering to automatically suppress respiratory harmonics.
Unlike fixed band-pass filters or signal reconstruction methods, RLSRHS continuously adapts to changing breathing patterns and does not require manual parameter tuning, making it well-suited to non-stationary physiological signals.
Testing the Technique in a Complicated Environment
The researchers evaluated the two methods separately and jointly using a 77 GHz millimetre-wave FMCW radar system. All experiments were approved by the Ethics Committee of Xiamen University, with informed consent obtained from participants.
Simulation studies showed that MCSM reduced the mean absolute error in distance detection by around 40 % compared with commonly used techniques, particularly in noisy and multi-person scenarios.
For heart-rate estimation, applying RLSRHS increased accuracy to 83 % under defined error-tolerance criteria, substantially outperforming conventional harmonic-suppression approaches.
The team then tested the full system in real indoor environments containing strong static reflections and dynamic interference. Radar-derived respiration and heart-rate measurements were compared with electrocardiogram monitors, Xiaomi wristbands, and respiratory sensors.
Across these experiments, the average error was approximately 4 %, closely matching simulation results. In longer recordings, simple post-processing steps, such as median filtering, were used to improve heart rate stability.
Implications for Radar-Based Health Monitoring
The study demonstrates that improving non-contact vital-sign monitoring requires jointly addressing distance localisation and frequency-domain interference.
By combining strong target selection with adaptive harmonic suppression, the proposed approach substantially improves accuracy in conditions that typically defeat radar-based systems.
While the experiments focused on indoor environments, the authors note that the methods could also be applicable to other settings where contact-free monitoring is desirable, such as smart homes or vehicle cabins.
Journal Reference
Zhang, C. et al. (2026). A radar vital signs detection method in complex environments. Scientific Reports, 16(1), 2333. DOI: 10.1038/s41598-025-32042-6
Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.