*3.1. Radar Sensor and Measurement Environment*

To verify the proposed algorithm, we established a test-bed in the DGIST lab, as shown in Figure 9. The test-bed is composed of a Doppler radar FEM (Front-end Module) with antennas, a DAQ (Data Acquisition) module, a power supply, and a PC.

The radar antennas are mounted to face the vehicle seats, and the transmitted and received ports are connected to the FEM using SMA (Sub Miniature type-A) cables. The received baseband signals are logged by the DAQ module, the data are sent to the PC through a USB (Universal Serial Bus) interface. To control DAQ module and obtain the data in real time, we developed DAQ software using NI's LabVIEW tool on the PC.

For the purposes of this paper, the FEM and antennas were manufactured by Yeungnam University. In the FEM, a VCO (Voltage Controlled Oscillator) is added, which is different from the board in an earlier circuit version [12]. A photo is shown in Figure 10.

**Figure 9.** Photo of the test-bed built in the DGIST lab.

**Figure 10.** Photo of the 2.45 GHz CW radar prototype.

The detailed specifications are presented in Table 1. The center frequency is 2.45 GHz and the FOV (Field of View) of the antenna is 80 degrees. In NI's DAQ module used here, we set the sampling rate to 1 kHz and the input dynamic range to −5~5 V through the LabVIEW tool.


**Table 1.** Parameters of the radar system used in this paper.

– In addition, the LPF and FFT parameters for the signal processing described above are also shown in Table 1. For the 2.4 GHz Doppler radar system, because the Doppler frequencies of passenger motion and breathing do not exceed 10 Hz and 1 Hz, we select the cut-off frequencies shown in Table 1. The FFT points are also selected such that they support the resolution of the Doppler frequency.

 
