Multiperson Localization and Vital Signs Estimation Using mmWave MIMO Radar

Abstract

This study explores the application of multiple-input multiple-output (MIMO) millimeter-wave frequency-modulated continuous wave (FMCW) radar for localizing multiple persons and estimating their vital signs. To distinguish the human targets using FMCW radar, high-resolution localization is employed first by the multiple signal classification (MUSIC) algorithm. Orthogonal projection separation (OPS) is further applied in the azimuth angle compensation to mitigate other targets’ interference. A harmonic MUSIC (HMUSIC) algorithm with multiple vital bins is designed to estimate non-pure sinusoidal signal frequency. Ray tracing simulation for FMCW radar signals to reflect the vital sign signal across various subjects is presented to verify the correctness of the proposed scheme under the additive white Gaussian noise channel. Besides, experiments in single target and two targets with different distances and the same distance are conducted. The proposed HMUSIC with multiple vital bins presents higher accuracy in vital sign estimation than the previous works. In two-target cases, their respiratory rates can be separately estimated, where an average of above 90% of respiratory rate estimation error is less than 3 breaths per minute. 83% of heartbeat rate (HR) estimation error can be maintained as less than 5 beats per minute at different distances and 71% of that at the same distances.