Two-Stage Zoom FFT-Enhanced Deep Learning-Aided Weighted Scheme for Wireless Vital Sign Estimation Using mmWave FMCW Radar

The accurate estimation of vital signs using millimeter-wave (mmWave) frequency-modulated continuous wave radar promises a future of enhanced health monitoring with superior convenience and reliability. This technological advancement is closely related to precise distance detection and user-adaptable signal processing solutions. The well-known fast Fourier transform (FFT) operation offers simple implementation but has limitations in distance detection performance. In this letter, we propose the application of a two-stage zoom FFT (zFFT) to precisely detect the distance by enhancing the detection resolution. The zoom FFT-enhanced scheme yields high-resolution measurements to facilitate the subsequent adaptable deep-learning (DL) based weighted scheme to achieve better performance with low complexity. Extensive experimental results demonstrate superior performance with a low parameter overhead for the proposed scheme.