Navigating Threats
Open Access DepositedIntelligent Models for Advanced Intrusion Detection in UAVs
Unmanned Aerial Vehicles (UAVs) play a critical role in modern surveillance, logistics, and defense operations, yet their integration of cyber and physical systems exposes them to a range of sophisticated cyber-physical threats. As these threats continue to evolve, traditional intrusion detection systems are not always able to identify complex, multi-layered attacks. This research investigates the performance of machine learning and deep learning models used to detect cyber intrusions in UAV environments. To do so, this research was able to leverage a dataset that linked the interconnected cyber and physical features of a UAV. The mapping of these features will become the focal point of testing various models to better detect cyber and physical intrusions in UAVs. To this end, the interconnected cyber and physical features selected for this research were used to train and test a variety of classification algorithms or models. Specifically, these models which consist of, but not limited to, Decision Trees, Random Forest, Naive Bayes, Support Vector Machines, and Multilayer Perceptron, were trained and evaluated on the leveraged cyber-physical dataset. Emphasis was placed on accuracy, precision, and recall, which determine each model’s ability to detect distinct attack types such as Denial Service (DoS), False Data Injection (FDI), Replay, and Evil Twin attacks. Further, this study will help advance better detection of intrusions in UAVs by developing a set of hybrid models (ensemble learning) that combine the strengths of multiple algorithms to improve detection performance. This research will focus on combining the strengths of multiple algorithms to detect intrusions specifically for interconnected cyber and physical data. The findings in this research underscore the importance of using interconnected cyber and physical data to ensure intrusions in UAVs are not missed. All in all, this research contributes to a novel UAV intrusion detection framework and provides a foundation for real-world deployment in defense, aviation, and autonomous systems.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.