Periodicity-Driven Millimeter Wave Radar Framework for Robust Non-Contact Heart Rate Monitoring

With the advances in millimeter wave radar, it is being explored as a contactless alternative to traditional wearable sensors for monitoring vital signs. However, heart rate (HR) monitoring with millimeter wave radar sensors face a significant challenge, including the weak heartbeat signal's susceptibility to noise, which can degrade the accuracy of HR monitoring. To address the challenge, this paper proposes an innovative periodicity-driven framework for contactless HR monitoring. The framework includes extracting heartbeat signals using a second-order differentiator, localizing cardiac signals by exploiting their periodicity, and estimating heartbeat timing via template matching. Evaluated on 10 subjects in typical office settings, our method achieves a mean HR estimation error of 0.60 bpm and a median interbeat interval (IBI) error of 2.58 ms. These results highlight the accuracy and robustness of our method and suggest its strong potential for healthcare applications.