Radar performance

The videos below showcase some of the radar capabilities. The module is connected to a PC for visualization purposes only.

Plot of the FFT magnitude:

The radar is strapped to my chest while walking around with a camera to capture the scene. The FFT magnitude of the received signal is plotted. A basic peak search algorithm is implemented where obstacles are detected based on their magnitude.

Range direction graph:

In the following example, the direction information is estimated thanks to the phase difference between the received signals from both antennas. When setting up the radar horizontally, the device indicates if the detected target is located up or down (elevation angle).

Kalman filter tracking:

A Kalman filter is implemented to track multiple moving targets. The instantaneous detection information pass first by a measurement-to-target-association algorithm which associates each measurement to its corresponding logged target. Then, the Kalman filter uses the new measurements to update its state.

Doppler processing:

Thanks to multiple emitted pulses per scan, the radar computes the phase shift between the received pulses to estimate the target velocity. This technique can resolve two closely located targets moving with different speeds.