Intelligent Counter-UAV Threat Detection Using Hierarchical Fuzzy Decision-Making and Sensor Fusion

Abstract

Highlights: This study presents a hierarchical fuzzy decision-making approach for detecting UAV threats using data from multiple sensors. It introduces a three-level fuzzy system that evaluates sensor performance and selects the best sensor combinations to improve threat detection in Counter-UAV operations. What are the main findings? A hierarchical fuzzy inference framework was developed to evaluate and optimize multi-sensor configurations for Counter-UAV applications. The system demonstrates that combining fuzzy logic with sensor fusion improves detection accuracy while balancing cost, range, and operational effectiveness. What is the implication of the main finding? The proposed approach enables intelligent and adaptive sensor selection in real-time threat environments, enhancing decision-making in Counter-UAV systems. The modular and interpretable framework can be applied to other domains requiring uncertainty management and multisensor optimization. This paper proposes an intelligent hierarchical fuzzy decision-making framework for threat detection and identification in Counter-Unmanned Aerial Vehicle (Counter-UAV) systems, based on the fusion of heterogeneous sensor data. To address the increasing complexity and ambiguity in modern UAV threats, this study introduces a novel three-stage fuzzy inference architecture that supports adaptive sensor evaluation and optimal pairing. The proposed methodology consists of three-layered Fuzzy Inference Systems (FIS): FIS-A quantifies sensor effectiveness based on UAV flight altitude and detection probability; FIS-B assesses operational suitability using sensor range and cost; and FIS-C synthesizes both outputs, along with sensor capability overlap, to determine the composite suitability of sensor pairs. This hierarchical structure enables detailed analysis and system-level optimization, reflecting real-world constraints and performance trade-offs. Simulation-based evaluation using diverse sensor modalities (EO/IR, Radar, Acoustic, RF), supported by empirical data and literature, demonstrates the framework’s ability to handle uncertainty, enhance detection reliability, and support cost-effective sensor deployment in Counter-UAV operations. The framework’s modularity, scalability, and interpretability represent significant advancements in intelligent Counter-UAV system design, offering a transferable methodology for dynamic threat environments.