Classification of Drone Type Using Deep Convolutional Neural Networks Based on Micro- Doppler Simulation

摘要

Monitoring drones has become an increasingly significant area of study for surveillance and safety purposes. The use of radar is one of the most feasible approaches to detecting a drone as RF waves propagate with a low attenuation constant. In this study, we suggest classifying drones based on the micro-Doppler signatures in the spectrogram when observed by radar. We simulate micro- Doppler signatures from various drones and investigate the feasibility of classifying the type of drone. A deep convolutional neural network is suggested as a classifier and its classification accuracy is reported.