IDENTIFICATION AND POSITIONING OF UAV IN SHELTERED AREAS OF BUILDINGS BASED ON FMCW RADAR

The past few decades have witnessed the monumental success of Unmanned Aerial Vehicles (UAVs) in both military and civilian applications. However, this success has also led to the increased need for UAV detection and recognition, as improper management of UAVs can pose serious risks. Detecting UAVs is particularly challenging due to their small Radar Cross Section (RCS) and consequently low Signal-to-Noise Ratio (SNR) in received signals. This difficulty is further compounded in urban environments where UAVs are often located in Non-Line-of-Sight (NLOS) areas, shielded by massive buildings. In this paper, we propose an approach to identify and locate UAVs in building shadow areas using a Frequency Modulated Continuous Wave (FMCW) radar system. Firstly, we extract the micro-Doppler spectrum of the drone rotor from the one-dimensional FFT spectrum of the echo signal. Then, we employ the peak extraction method to estimate the parameters of the drone rotor speed and blade length, facilitating drone identification. Subsequently, we extract range bins data capable of recognizing drones, which serves as input for the localization algorithm. By developing a multi-channel and multipath imaging fusion algorithm, we accumulate strong values corresponding to the drone location, thereby enabling its precise localization. Extensive simulations and experiments commendably validate the effectiveness and accuracy of the proposed method.