Extracting CFAR and RD map from drivingRad​arDataGene​rator

Hi,

I wanted to understand how drivingRadarDataGenerator api generates detections? Does it actually simulate Radar waveforms such as FMCW or pulsed radar , as i dont see these specific parameters are passed to this function. In parallel, i want to get RD-map and CFAR detections , which would be helpful for my algorithm development.

Please help !

Answer 1

[Divit]

Hi Vinay,

The “drivingRadarDataGenerator” in MATLAB’s Automated Driving Toolbox does not directly simulate radar waveforms such as FMCW (Frequency Modulated Continuous Wave) or pulsed radar. Instead, it generates detections based on specified parameters like object positions, velocities, and radar sensor characteristics.

To generate RD-map (Range-Doppler map) and CFAR (Constant False Alarm Rate) detections, you would need to implement these processing steps separately in your algorithm using the radar detections generated by the function.

• For RD-map, you will need to convert the radar detections into a format suitable for RD-map processing and then perform a 2D FFT (Fast Fourier Transform) on the radar detections.
• For CFAR detections, you will need to implement a CFAR algorithm on the RD-map data to detect and track objects.

To know more you can refer to the following documentation links:

• drivingRadarDataGenerator: https://www.mathworks.com/help/driving/ref/drivingradardatagenerator-system-object.html
• Range-Doppler Response: https://www.mathworks.com/help/phased/ug/range-doppler-response.html
• CFAR Detection: https://www.mathworks.com/help/radar/ug/cfar-detection.html

I hope this helps!