Non-Contact Vital Sign Detection With High Noise and Clutter Immunity Based on Coherent Low-IF CW Radar

Abstract

Non-contact vital sign detection using Continuous-Wave (CW) radar is subject to noises and clutters. The heterodyne architecture of the radar transceiver resolves the flicker noise. However, it still suffers from other noise components. Moreover, the presence of clutter also significantly introduces distortions in the sensing results. In this paper, an extended Noise-Immune Motion Sensing (ENIMS) technique is proposed to tackle the noise and clutters simultaneously in the low intermediate-frequency (IF) CW radar. It works by synthesizing I/Q signals at the IF peak of the spectra of the sequentially divided signal segments. Each segment generates one pair of I/Q data points and thus improves the signal-to-noise ratio (SNR). During this process, clutters are also converted into DC components of the I/Q signals. The circle-fitting-based DC compensation technique can thus be used to resolve the clutter issues. High-accurate displacement motion is then reconstructed with the DC-compensated I/Q signals. The theory and noise performance analysis are presented. Simulation and experiments show that, with the proposed technique, the SNR is improved by around 34 dB. Mechanical vibration as small as 90 μm and the subject person's breath and heartbeat at 3.2 m away from the 5. 8 GHz radar were detected under cluttered office environments with a small transmitting power of only 10 μW, whereas the conventional methods fail in the same cases.