Khalifa University PhD Thesis: UAV Pilot Identification with ML

This PhD thesis from Khalifa University explores the use of machine learning for Unmanned Aircraft Vehicle (UAV) pilot identification. The research focuses on capturing behavioral data from different pilots during UAV flights and using this data to train machine learning algorithms capable of identifying the pilot of each dataset. The study uses 20 datasets with various flight modes (triangle, vertical, random) and features derived from Radio Controlled (RC) signals and Internal Measurement Unit (IMU) sensors. The Random Forest algorithm proved to be the best classifier, achieving high accuracy in pilot identification. The thesis also investigates tree reduction strategies to optimize the model's power and memory consumption. The primary goal is to develop a mechanism that enables UAVs to recognize the pilot to prevent unauthorized access. The research covers UAV applications, security issues, machine learning techniques, data storage, and results. The thesis highlights the application of machine learning in enhancing drone security and authentication.