Emerging Trends in Radar: Drone Characterization Using Deep Learning on Micro-Doppler Data

Identifying whether a drone is being used for recreational or malicious purposes, a process referred here as drone characterization, is essential for ensuring security, safety, and privacy. One of the most effective methods for this is applying deep learning to the drone's micro-Doppler spectrogram captured by radar. This article reviews current trends and future advancements in this field, focusing on less-explored deep learning approaches that have the potential for significant impact, rather than revisiting the more commonly discussed topics in the existing literature.