Millimeter-wave Radar Vital Signs Detection Based on Modified Independent Component Analysis

This study proposes a novel method for human vital signs detection using a frequency-modulated continuous-wave (FMCW) radar. First, the phase signal of the human thoracic displacement is extracted using radar. Subsequently, the phase signal is decomposed into intrinsic mode functions (IMF) using complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and the frequency of each IMF is calculated. Finally, the IMF with frequencies falling at 0.1-0.5 Hz and 0.8-2.0 Hz are selected, respectively. The selected two sets of signals are processed using independent component analysis (ICA) and the second-order derivative independent component analysis (SOD-ICA) algorithm. The respiration and heartbeat rates can be obtained from the final results.