Continuous Wave Radar based Vital Sign Estimation: Modeling and Experiments

This paper studies the use of continuous wave (CW) radar for non-invasive estimation of vital signs, in particular respiratory and heart rate. We derive a realistic reference model for a stationary human subject. We propose methods for estimating the vital signs, including nonlinear and linear signal demodulation, as well as time-frequency methods utilizing the micro-Doppler structure of the signal. We show that real-world measurements with 24 GHz CW-radar taken from different positions close to human are in line with the simulation results. The results indicate that reliable estimation of the heart rate can be obtained with the nonlinear demodulation if the radar is suitably positioned w.r.t. the subject.