Vital Sign Estimation using FMCW Radar in Various Scenarios

Millimeter-wave (mm-wave) radars enable non-contact detection of human vital signs by capturing chest movements caused due to respiration and heart activity [1].This study analyses a frequency-modulated continuous wave (FMCW) radar based heart rate estimation algorithm for determining its robustness across diverse scenarios.The accuracy of the estimated heart rate is also checked for scenarios where measurements have been performed in rooms and cars of various dimensions.The algorithm performance is validated through both real and simulated radar data where the accuracy and precision of the estimated heart rates have been quantified and will be presented in this work.The measurements of heart and breathing rates have been performed using FMCW radar operating in the 58-63.5 GHz frequency range [2].The radar signal processing algorithm used here operates on in-phase (I) signals from the radar and generates quadrature-phase (Q) signals using Hilbert transform.The quadrature-phase signal is examined to determine the impact of noise on the Hilbert transform.Subsequently, a range computation algorithm (range FFT) is applied to the I-Q signal corresponding to each of the chirps.This is followed by a static clutter removal algorithm to mitigate impact of stationary objects on the radar echo signals.In addition to this, range of the targets are dynamically calculated across multiple chirps using an adaptive range bin selection method.