A Real-Time Phase Demodulation Method based Outlier-Robust Kalman Filter for FMCW Radar Vital Sign Detection

This paper investigates the application of S-Band FMCW radar in human vital sign estimation. To achieve more accurate estimation of respiratory cycle signals, a phase demodulation algorithm based on outlier-robust Kalman filter is proposed. Experimental comparisons are conducted with the arctangent demodulation (ARC) and digital low-pass finite impulse response (FIR) filter method. The results demonstrate that the reconstructed respiratory waveforms using the outlier-robust Kalman Kalman filter-based approach achieve a Pearson correlation coefficient of 0.78 compared with the ground truth. The proposed algorithm achieves both higher efficiency and more accuracy respiratory waveform reconstruction compared to the other two methods.