A novel signal processing algorithm improves the signal-to-noise ratio and accuracy of noncontact vital sign measurements using millimeter-wave radar in the presence of mutual interference.
Noncontact vital sign measurements based on millimeter-wave radar can realize long-range detection of respiratory and heartbeat signals, therefore it is gradually applied in more and more scenes. With an increase in the number of millimeter-wave radar devices, the mutual interference between radar signals will occur, which will increase the noise floor. Moreover, respiratory and heartbeat signals are relatively weak and easy to be submerged by the high noise floor. To solve this problem, we present a novel method for suppressing mutual interference and extracting vital signs using improved morphological component analysis (IMCA) and an adaptive parameter optimization variational mode decomposition (APVMD) algorithm. The IMCA algorithm is used to suppress the mutual interference based on the different sparsities of components in different sparse domains, and this improvement solves the inability of the MCA algorithm to detect the interfered signal. Then, the APVMD algorithm is used to extract respiratory and heartbeat signals from the signal containing residual noise. The rate of energy loss is used as the index to optimize the parameters, and the step size of the penalty value is adaptively adjusted according to the center frequency of the mode. The simulation and field experiment results show that the proposed method can significantly improve the signal-to-noise ratios (SNRs) of respiratory and heartbeat signals in the case of mutual interference between radar signals. Moreover, a comparison with other interference suppression and extraction methods shows that the proposed method better improves the SNR and accuracy.
Zhang et al. (Mon,) studied this question.