The current wireless Unmanned Aerial Vehicle (UAV) approaches based on supervised learning have a high false detection rate, making it difficult to recognise novel types of attack activities, and increasing the cost of collecting labelled network data. The method employs unsupervised learning and does not require human tagging of large amounts of wireless network data. As a result, it can quickly obtain training data sets and discover unknown assault patterns. The multi-view cosine distance method for hierarchical clustering is introduced. This makes clustering results more rational, network behavior assessment more accurate, and intrusion detection false detection rate lowered. The experiment uses the AWID (Aegean WIFI Intrusion Dataset) attack data collection. The principle component analysis approach decreases the dimensionality of the experimental data set, reducing the temporal complexity of the intrusion detection algorithm. The experimental findings reveal that the proposed multi-view hierarchical clustering wireless network intrusion detection algorithm outperforms the classic wireless network intrusion detection method in terms of detection rate, false detection rate, and unknown attack types

Multi-View; UAV; Hierarchical Clustering; Wireless Network; Intrusion Detection; Principal Component Analysis