Radar-based Drone detection using ML techniques

This work proposes a novel way of using Machine Learning (ML) algorithms on publicly available radar dataset* for Unmanned Aerial Vehicles (UAV) detection that contains their Radar Cross Section (RCS) measurement and few other features, in detecting the frequency between 26-40GHz, using Decision Tree algorithm. The dataset measurements were conducted for the quasi-monostatic situation, where the transmitter and receiver are situated in the same position, in an anechoic space. The data demonstrates the manner in which radar waves are dispersed by various UAVs within the designated frequency range. ML algorithms have been used to predict the frequency (GHz) using phi (deg) for x-axis rotation, theta (deg) for y-axis rotation, RCS (dB) as the input parameters. All the results have been implemented in MATLAB.