Motion-Tolerant Measurement of Respiration and Heartbeat via mmWave Radar Across Diverse Healthcare Scenarios

Noninvasive vital sign (e.g., respiration and heart-beat) monitoring via millimeter wave radar (mmWave) presents a promising paradigm for health management and disease diagnostics, yet the widespread adoption of this technology is often hindered by motion artifacts, environmental noise, and limited adaptability to diverse scenarios. This paper proposes a robust respiratory and heartbeat measurement approach that integrates enhanced radar sensing with adaptive decomposition algorithms for diverse healthcare environments. Initially, we introduce a target-signal cooperative mechanism that combines target detection for spatial localization with beamforming-driven directional adjustments and channel coherent accumulation, thereby enhancing the signal to noise ratio (SNR) in the region of interest. Then, the marine predators algorithm (MPA) is implemented to optimize variational mode decomposition (VMD) parameters for precise vital sign extraction. Prior to this, a cascaded signal processing pipeline is applied, comprising static clutter suppression, differentiation and cross multiplication (DACM) phase extraction, and signal smoothing to mitigate noise effects. Experimental validation encompasses two configurations: (1) multi-angle and multi-distance static indoor scenarios, (2) four motion scenarios with controlled body movements, designed to emulate real-world conditions. The experimental results indicate superior respiratory rate (RR) estimation performance, achieving a mean absolute error (MAE) of 1.02 bpm. For heartbeat rate (HR) monitoring, MAEs of 3.23 bpm (static scenarios, 95.78% accuracy) and 4.16 bpm (lightweight motion scenarios, 94.88% accuracy) demonstrate significant feasibility and substantial reliability.