The Spatio-Temporal Data-Centric Detection in Geographic-Homogeneous Unmanned Cluster

The credibility of wireless communications in unmanned aerial vehicle (UAV) networks present significant security challenges. Internal malicious nodes may inject false messages, causing receiving UAVs to take erroneous actions potentially leading to accidents. This necessitates passive security mechanisms to detect false messages from internal malicious nodes. Existing research primarily focuses on detecting anomalies through inter-agent traffic identification and cloud-based rationality checks, rarely considering the high dynamicity and randomness of UAV network interactions. This paper proposes a novel data detection method based on spatio-temporal information correlation. We first divide UAVs into clusters based on geographic coordinate correlation, then apply a spatio-temporal data-centric detection approach to identify abnormal data within these clusters. Our method leverages the inherent spatial and temporal relationships in UAV communications to enhance detection accuracy. Analysis results demonstrate that this method successfully detects anomalous data within UAV networks, addressing the unique challenges posed by the dynamic nature of UAV communications. This approach not only improves security in current UAV networks but also provides a foundation for developing more sophisticated and adaptive security measures in future unmanned aerial systems.