Study on Stealthy C2 Spoofing Methods for Anti-Jamming UAVs Equipped with Array Antennas

This dataset presents the clustering results obtained from the TG-HDBSCAN-PDW algorithm, designed for direction\u2013power joint analysis in UAV communication and deception scenarios. The figure visualizes the spatial distribution of clustered data points in the azimuth\u2013elevation domain, where each color represents a distinct stable cluster identified through time-enhanced density estimation. Compared with conventional HDBSCAN, the proposed TG-HDBSCAN-PDW integrates temporal continuity and power-domain weighting, effectively reducing noisy samples and improving cluster compactness and separability. The displayed results demonstrate clear boundary transitions between adjacent clusters, with a low proportion of unclassified points, indicating robust performance under complex interference and dynamic target conditions. Quantitatively, the clustering achieves a Davies\u2013Bouldin index of 0.135 and a Silhouette coefficient of 0.901, confirming high intra-cluster consistency and inter-cluster contrast. These results validate the method\u2019s ability to capture coherent spatial\u2013temporal features of received signals, providing a reliable foundation for subsequent UAV cooperative deception, direction tracking, and power control. Overall, the dataset serves as an illustrative example of stable cluster extraction under multi-source and fluctuating environments, supporting research in intelligent signal processing and physical-layer security.