77GHz Millimeter-Wave Radar Vital Signs Detection Based on GA-VMD Algorithm

Abstract:
In recent years, millimeter-wave radar has been widely used in various fields, and it has also begun to flourish in the noncontact monitoring of physiological signals. In this paper, we realize the acquisition of static human’s vital sign signals with 77GHz millimeter-wave radar and introduce GA-VMD to separate the respiratory and heartbeat signals. First, the frequency modulated continuous waves (FMCW) is transmited from millimeter-wave radar, then the intermediate frequency (IF) signals containing vital sign signals are separated from the echo signals. For the obtained phase signal, the Variational Modal Decomposition (VMD) algorithm is applied to obtain the potential modal components contained in the breathing and heartbeat signals after phase unwrapping. Aiming to better determine the uncertain parameters in the VMD algorithm, we introduce the Genetic Algorithm (GA) to optimize some of them first. Meanwhile, the sample entropy is used as the fitness function. At last, the breathing signal and heartbeat signal of the static human are reconstructed by adaptively selecting some modal components. Compared with the bandpass filtering method and the VMD algorithm without GA, the mean standard deviation (MSD) of GA-VMD algorithm in the determination of the heart rate is increased by 6.57% and 4.93% respectively, and the signal-to-noise ratio (SNR) is increased by 5.72dB and 2.48dB respectively. In conclusion, the SNR of the obtaind vital sign signals have been successfully improved because of the introduction of GA.