Wavelet- and Cosine-Transform-Based Super-Resolution Algorithm (WCT-SRA) for Radar-Based Multi-Person Vital Sign Monitoring

This study focuses on enhancing the frequency resolution to improve the multi-person vital sign detection using a short time window. The proposed wavelet- and cosine-transform- based super-resolution algorithm (WCT-SRA) significantly improves the frequency resolution by estimating the amplitude, frequency, and initial phase of a signal using a cosine-transform-based method. Moreover, wavelet decomposition makes the decomposition process for each subject’s component more efficient. Experimental results with two actuators demonstrate that the WCT-SRA can distinguish closely spaced vibration frequencies of two metal plates and enable frequency resolutions of 0.1 Hz, 0.05 Hz, and 0.01 Hz with time windows of only 5 s, 10s, and 20s, respectively. Moreover, by adjusting the initial distances of subjects from the radar, the vital signs of four individuals were simultaneously determined during a 20-s time window.