Millimeter-Wave Radar Vital Signs Measurement With Random Body Movement Using Missing Data Model

Radar has gained increasing interest in medical and health monitoring due to its noncontact nature and privacy protection. However, the measurement of respiratory and heartbeat is influenced not only by clutter and noise, but also by human random body movement (RBM). RBM often confuses the phase information of respiratory and heartbeat echoes, making frequency estimation difficult, especially for heartbeat rate. This article addresses the problem of accurately measuring vital signs in the existence of RBM, and proposes a measurement method using the missing data model. The approach first derives and determines the pulse repetition frequency (PRF) suitable for radar transmit signals in the presence of RBM to ensure that the phase of adjacent pulses does not jump. Then, the missing data model of echo phase, reflecting variations in respiration and heartbeat, is established using a RBM recognition technique. Finally, respiration and heartbeat information is measured from the missing data. Simulation and measurement verify that the proposed method can accurately estimate respiratory and heartbeat rates even with a large RBM. Multiple experiments in different scenarios show that the root mean square error of the proposed method are 0.013 and 0.039 for respiratory and heart frequencies, respectively. That is, the respiratory and heartbeat frequencies had errors of 0.78 beats per minute (bpm) and 2.34 bpm, respectively.