ELF Passive Radio Sensing and AI-Perception of Micro-UAS

Abstract: The rapid proliferation of micro-unmanned aerial systems (UAS) has raised significant security and airspace management concerns, necessitating the development of robust detection methods. This paper presents a passive sensing approach for drone detection based on extremely low frequency (ELF) electromagnetic emissions. Unlike traditional active radar or acoustic methods, the proposed system does not require signal transmission, making it undetectable and interference-free. A highly sensitive ferrite rod and circular loop antenna were designed to capture ELF emissions, followed by an optimized analog front-end for noise reduction. The acquired signals were processed using advanced Digital Signal Processing (DSP) techniques, including the Spectral Correlation Function (SCF), to extract distinctive spectral features of drone emissions. A deep learning-based classifier was implemented to distinguish drone-generated ELF signatures from environmental noise. Experimental results demonstrated a detection accuracy of 96% and a drone classification accuracy of 87%, highlighting the reliability of the proposed method. The findings suggest that ELF-based passive drone detection offers a promising solution for real-time surveillance applications in security and defense without requiring active signal transmission.