Analysis of methods for detecting unmanned aerial vehicles using wireless sensor networks

The article presents a comprehensive analysis of modern methods for detecting unmanned aerial vehicles (UAVs) within wireless sensor networks, taking into account the evolution of aerial threats and the emergence of new types of small-sized drones. The subject of the study is UAV detection methods based on radar, radio-frequency, acoustic, and electro-optical/infrared subsystems, as well as the principles of their integration into a multisensor information and control architecture. Particular emphasis is placed on the dominance of small FPV drones on the battlefield and the emergence of platforms with fiber-optic control links, which are effectively immune to electronic warfare systems and “invisible” to RF detectors, thereby creating a critical gap in traditional Counter-Unmanned Aircraft Systems (C-UAS) approaches. The capabilities and limitations of each sensor modality are systematized. It is shown that the effectiveness of radar systems is determined by the small radar cross-section (RCS) of the target and line-of-sight conditions; as the RCS decreases, the detection range drops sharply, while RCS fluctuations caused by rotor rotation complicate target tracking. At the same time, the analysis of micro-Doppler signatures is considered a promising tool for distinguishing UAVs from birds and for classifying the type of platform based on spectral features. It is established that RF methods provide a high probability of detection only for drones with an active radio channel and are fundamentally unsuitable for autonomous and fiber-optic platforms. For acoustic systems, the key advantages of passive operation and non-line-of-sight (NLOS) capability are identified, while their limitations in terms of range, dependence on wind conditions, and elevated false alarm rates in urban noise are emphasized. For the electro-optical/infrared channel, a critical dependence on weather conditions, illumination, and thermal contrast is noted, as well as the need for external cueing due to the limited field of view at high optical magnification. The necessity of transitioning from single-channel solutions to a multisensor architecture with Multi-Sensor Data Fusion (MSDF) and adaptive weighting of information flows depending on environmental parameters and channel reliability is substantiated. A conceptual model of an information and control surveillance system based on wireless sensor networks is proposed, incorporating indicators of measurement availability for different sensors, which is especially important in the absence of an RF component for detecting fiber-optic UAVs. The role of artificial intelligence algorithms in reducing false alarms, classifying micro-Doppler, visual, and acoustic features, and improving the reliability of realtime decision-making is emphasized. It is concluded that only the integration of several sensor modalities into a distributed networked MSDF system can provide the required probability of detecting small UAVs under stochastic environmental disturbances and evolving drone control methods.