Enhancing UAV Security: A Comparative Analysis of Machine Learning Algorithms for Intrusion Detection

The escalating reliance on unmanned flying vehicles in many domains highlights the need for robust security measures against cyber attack, physical attack, and the combination of cyber and physical attack. This paper explores the performance of several machine learning techniques (Random Forests, K-Nearest Neighbors, Support Vector Machines, and K-means Clustering) in identifying such threats using a comprehensive dataset that combines cyber and physical properties. We evaluate the performance of these algorithms in three different contexts: we evaluate the performance of these algorithms in three different contexts: Cyber-Only, Physical-Only, and cyber-physical fusion. Our results show that the cyber-physical fusion approach can significantly improve the detection rate compared to network or physical analysis alone. This study highlights the advantages of integrating different data sources to enhance intrusion detection capability and suggests that future research focuses on developing sophisticated models, exploring hybrid methods, and empirically validating them in real-world environments to enhance the performance of intrusion detection systems.