High-Speed Detection of Moving Objects Using FMCW Radar Sensor

Today, high-speed navigation has become crucial in air traffic and self-driving cars, where frequency-modulated continuous wave (FMCW) radars are widely used as low-cost range sensors. Such sensors allow calculating the range-Doppler map (RDM) to increase the detection range for small drones. This map requires computing a 2D fast Fourier transform (FFT) of the time-chirp matrix to find spectral peaks above the noise floor. Unfortunately, with every newly received chirp, the 2D-FFT must be recomputed, which is computationally expensive; moreover, it may cause performance degradation when detecting fast-moving objects. In this letter, we propose a low-cost algorithm that avoids large-size FFT calculations and instead calculates block-wise 2D-FFTs with further tracking of local maxima, thus lowering complexity and improving sensitivity of radar sensors in high-speed scenarios.