"Trajectory-Driven Intelligent System for UAV Location Integrity Checks" by Mincheol Shin

Trajectory-Driven Intelligent System for UAV Location Integrity Checks

Author

Mincheol Shin

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Computer Science and Information Systems

Date of Award

Spring 2025

Abstract

Unmanned aerial vehicles (UAVs) are now widely used across different industries, and concerns about maintaining location integrity have become more significant to ensure secure deployment and management. Many studies have addressed this issue by applying hardware sensors, cryptographic techniques, and machine learning (ML) methods. However, most of these approaches place their focus on GPS signal-related data, such as jamming and noise. In this study, we propose an alternative method that examines movement patterns based on sequential flight records. Instead of depending on isolated signal-specific data points, our approach evaluates location accuracy by tracking updates across the entire flight trajectory. To accomplish this, we define a set of attributes that represent aerial vehicle movements effectively over time. Additionally, we introduce a deep sequence analysis method implemented using either a recurrent neural network (RNN) or a Transformer architecture with a backend classifier. The results from extensive experiments confirm the reliability of our trajectory-based approach, demonstrating classification performance of up to 98.9% with false positive rates kept below 1%, even without referencing GPS-specific information.

Advisor

Jinoh Kim

Subject Categories

Computer Sciences | Physical Sciences and Mathematics

Share

COinS