Motion-Robust Noncontact Heartbeat Sensing Using Radar Acoustics for Healthcare IoT

The emergence of non-contact biosensing technologies is transforming the Healthcare Internet-of-Things (IoT) by enabling continuous and unobtrusive monitoring of cardiovascular function, critical for early disease detection and personalized care. Millimeter-wave radar, known for its fine spatial resolution and motion sensitivity, offers a compelling platform for remote cardiac monitoring. However, practical deployment remains challenging due to susceptibility to motion artifacts, respiratory interference, and physiological variability. In this work, we present a novel radar-acoustic cardiac monitoring framework that addresses these challenges by exploiting the spectral separation between high-frequency heart sounds (20-80 Hz) and the lower-frequency thoracic movements (0.7-1.5 Hz). This natural spectral isolation facilitates the suppression of motion artifacts and respiratory interference, enhancing the robustness of cardiac signal extraction. Building on the presented insights, we develop an active motion cancellation technique to effectively estimate and compensate for band-limited multiplicative motion interference even under complex body motions. This method enables seamless signal extraction while preserving the integrity of heart sound signals. Simulations and experimental validations demonstrate significant improvements in heart rate estimation and pulse waveform recovery, even under realistic motion scenarios. This privacy-preserving, non-contact, and clothing-penetrating sensing framework presents a promising avenue for seamless integration of cardiac monitoring into IoT ecosystems, enabling real-time health analytics in free-living environments.