UWB-RADAR-SENSED HUMAN RESPIRATORY SIGNAL MODELING BASED ON THE MORPHOLOGICAL METHOD

This paper proposes a morphological ultra-wideband (UWB)-radar-based respiratory signal model. According to the detection theory, it is crucial to set up an appropriate model to fulfil the detection purpose. Previous models pay less attention to the time dimension of the respiratory signal, but the frequency domain cannot precisely describe it because of its non-linearity and non-stationarity. This model uses a morphological operator to dilate or erode the base wavelet, and the length and value of the digit in the structure element serve as the parameters in this morphological model. The result of the experiment carried out on 10 human targets with impulse radio ultra-wideband (IR-UWB) radar proves the efficiency of this model. As the UWB radar sensed human respiratory signal is nonlinear and non-stationary, the parameters in the model can be regarded as a measure of non-linearity and non-stationarity. An experiment is carried out with the simulated respiratory signal generated with the proposed model. The result shows that the detection algorithm based on Ensemble Empirical Mode Decomposition (EEMD) method has a better performance than that based on Adaptive Line Enhancer (ALE) and with the value of the digit in the structure element increases, the performance of the ALE method declines, while the EEMD method stays in a good performance, which indicates that the EEMD method has a good potential to deal with the nonlinear and non-stationary respiratory signal.