Date of Award
Master of Science (MS)
The rapid development and deployment of 5G/6G networks have brought numerous benefits such as faster speeds, enhanced capacity, improved reliability, lower latency, greater network efficiency, and enablement of new applications. Emerging applications of 5G impacting billions of devices and embedded electronics also pose cyber security vulnerabilities. This thesis focuses on the development of Global Positioning Systems (GPS) Based Anomaly Detection and corresponding algorithms for Unmanned Aerial Systems (UAS). Chapter 1 provides an overview of the thesis background and its objectives. Chapter 2 presents an overview of the 5G architectures, their advantages, and potential cyber threat types. Chapter 3 addresses the issue of GPS dropouts by taking the use case of the Dallas-Fort Worth (DFW) airport. By analyzing data from surveillance drones in the (DFW) area, its message frequency, and statistics on time differences between GPS messages were examined. Chapter 4 focuses on modeling and detecting false data injection (FDI) on GPS. Specifically, three scenarios, including Gaussian noise injection, data duplication, data manipulation are modeled. Further, multiple detection schemes that are Clustering-based and reinforcement learning techniques are deployed and detection accuracy were investigated. Chapter 5 shows the results of Chapters 3 and 4. Overall, this research provides a categorization and possible outlier detection to minimize the GPS interference for UAS enhancing the security and reliability of UAS operations.
Mohan, Jaya Preethi, "GPS Anomaly Detection And Machine Learning Models For Precise Unmanned Aerial Systems" (2023). Theses and Dissertations. 5684.