Radar System for Detecting Respiration Vital Sign of Live Victim Behind the Wall
Using radar systems for non-contact respiration detection has emerged as a significant method of detecting the vital signs in living things. It holds immense potential for applications in medical, security, and disaster management fields.
Frequency Modulated Continuous Wave (FMCW) radar system detects a living being by capturing the phase change of the beat signal due to the movement of the chest or abdominal wall. A simple FMCW radar system may not accurately detect the respiratory signs of humans behind or under an object like a building or debris. The debris creates multiple reflections, increasing the clutter, phase shift effects, and beat frequencies that can interfere with the radar's readings.
A new detection method is proposed using two-step Fast Fourier Transform (FFT) computation to address the problem. The first step is to identify the response from the surrounding structures. Beat frequencies created by obstructions are eliminated, and a region of interest (ROI) is established. ROI is a range of the FFT Index after the obstacle response considered in a situation.
A weighting process is employed further to improve the signal-to-clutter ratio (SCR). The strongest reflection signal in ROI is taken as a reference or weight factor. All phase detector outputs are cross correlated with reference, and fringe signals are discarded. The values of SCR before and after weighting were -4.64 dB and 5.41 dB, respectively, indicating a vast improvement.
A phase detection method extracts the Doppler response corresponding to human respiratory rate information in the second step. The FFT process with the largest peak spectrum is selected as the target vital sign data.
The proposed method was evaluated in a laboratory setup consisting of an FMCW radar and a mini PC for computation. The radar has a 24 GHz frequency with a bandwidth of 50 MHz and a chirp period of 0.4μs. The number of FFT sequences is 512, and the sampling frequency is 100 MHz. Each FFT point represents a gap of 12.5 cm. The ROI is defined as the range of the FFT index after sampling the obstacle response in the detection area. A live victim is placed in a reclined position under concrete debris.
The phase detector output shows time-varying small displacements of the chest or abdominal wall to represent breathing. In the frequency domain, a peak position describes the respiratory rate of the victim. Several target lying positions were also analyzed. It is found that the back/supine position shows a more significant amplitude phase detection value compared to the recumbent/side position.
Radar system detection results can be represented in b-scans and c-scans. A b-scan gives an estimated respiratory rate, while a c-scan provides information about the position of a living person in the scanning area.
The results show an error rate of approximately 3.375 cm in location detection, which is reasonably accurate considering the size of the human target and scanning area.
The proposed two-step FFT and ROI determination could improve SCR and assist in detecting vital signs to save lives from under the rubble.