APPROACH TO DETECTION AND CLASSIFICATION OF RADIO-CONTROLLED MODELS BY THEIR RADIO SIGNAL

Abstract

The ongoing development of affordable embedded sensors, microcontrollers, and wireless communication technologies has led to a rapid rise in the popularity and use of unmanned aerial vehicles (UAVs), commonly known as drones, across various fields – including agriculture, environmental monitoring, reconnaissance, and defense. However, alongside these benefits, the potential for malicious or unauthorized drone usage poses significant risks to public safety, critical infrastructure, private property, and industrial facilities. Therefore, there is an increasing need for systems capable of detecting and classifying drones based on the analysis of their radio frequency (RF) communication signals. This paper explores an approach for monitoring and identifying drones by analyzing the RF signals generated by their communication modules. Specifically, the proposed system leverages the capabilities of a software-defined radio (SDR), which enables real-time signal reception, processing, and analysis without the need for specialized hardware. The implementation utilizes the GNU Radio framework, offering flexibility in designing custom signal processing workflows. The focus is placed on classifying drone communication signals based on their modulation type – particularly, orthogonal frequency-division multiplexing (OFDM). Signal classification is achieved by analyzing its energy profile and structural features, which vary depending on the communication protocol used. Machine learning algorithms are employed to train a model that can recognize and classify such signals in real-world environments. This enables the identification of the device type, behavior analysis, and, in some cases, estimation of the operator’s location – all without direct interaction with the drone. Overall, the study demonstrates the effectiveness of combining SDR technology with artificial intelligence methods for drone detection and classification, potentially forming the basis for early warning systems and countermeasures against unauthorized drone activity.