Multiple Human Heart Rate Variability Detection Using MIMO FMCW Radar With Differential Beam Techniques

A differential beam technique to measure the human heartbeat and its variability (HRV) is proposed, which is implemented with digital beamforming (DBF) using multi-input multi-output (MIMO) frequency-modulated continuous wave (FMCW) radar. Respiration and heartbeat signals are detected from a pair of differential beams directed to specific areas in the human chest. Because the third-order harmonic of the respiration signal appears near heartbeat frequencies, it becomes one of the major noise sources during HRV measurements. A macro-differential beam technique that selects the differential beam pair with the maximum signal to noise ratio (SNR) is introduced. N beams directed to a human chest using DBF provide
${_{N}}\text{C}{_{2}}$
pairs of macro-differential beams, with these pairs consisting of all combinations of the N beams. For
${_{N}}\text{C}{_{2}}$
differential beams, the amplitude ratio of the respiration and heartbeat signals is estimated. The pair of differential beams with the highest SNR of each phase difference is selected. Between the pair, the beam with the stronger heart signal is assigned as beam H, with the other being denoted as beam
${\overline {\text {H}}}$
. A micro-differential beam technique is also applied to increase the SNR. A set of differential beam pairs is generated, consisting of beam
${\overline {\text {H}}}$
and other beams generated by linear contrast minimal variation (LCMV) DBF. The summation and bandpass-filtering of the phase-difference signals of the set provide heart signals with a higher SNR. The proposed differential beam method is validated experimentally with two human targets, demonstrating an average absolute error percentage of 1.03% to detect HRV.