Car Occupancy Detection Using Ultra-Wideband Radar

In this paper, we propose car occupancy detection using ultra-wideband (UWB) radar to detect the breathing motion of a person. The target signal and multipath propagation inside the car are modeled as a pinhole channel. We show that the received signal can be described by the Kronecker product of the channel and breathing motion vectors. Thus, the covariance can be computed from the delay profile of the channel and the power spectrum of the chest motion. An estimator-correlator is applied for the detection. Measurements have been performed to confirm the structural assumptions about the received signal whereas Monte-Carlo simulations are used to evaluate the detection performance. The simulations demonstrate an improved detection rate in low signal-to-noise ratio (SNR) conditions compared to a windowed energy detector or FFT-detector.