Compact convolutional neural networks for AI-based drone detection system
Machine Learning
Summary
The authors studied how to detect drone signals using small neural networks that can work on simple devices. They turned radio signals into easy-to-process pictures and trained models on many labeled examples to recognize drones quickly and accurately. Their method works in real-time without complex signal processing, needing less computing power than traditional techniques. This makes it promising for use in real electronic warfare systems to spot drones early.
Authors
Gábor Farkas, Gábor Fazekas, Karakai Patrik, András Németh, Gábor Farkas
Abstract
The increasing use of first-person-view drones in modern conflicts has created a demand for compact and reliable detection systems capable of operating in complex electromagnetic environments. These drones continuously transmit video signals through onboard video transmitters, generating radio-frequency emissions that can be exploited for early detection. This study investigates the use of lightweight convolutional neural networks for automated detection of drone signals captured by a software-defined radio-based electronic warfare framework. Samples are converted into rasterized time-domain images, providing a computationally efficient input representation suitable for embedded systems. Several custom model architectures were designed and benchmarked in terms of accuracy, model size, and inference performance using a dataset containing approximately 40,000 labeled images. In addition to offline testing, the models were integrated into a GNU Radio signal processing chain for real-time evaluation. The results show that compact models can achieve high detection accuracy while maintaining low computational requirements, making them suitable for embedded radio-frequency monitoring applications. Compared with existing spectrogram-based RF detection methods, the proposed approach eliminates frequency-domain preprocessing and achieves comparable accuracy with significantly reduced computational cost.