Date Approved

2025

Degree Type

Open Access Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department or School

College of Engineering and Technology

Committee Member

Omar Darwish, PhD

Committee Member

Fathi Amsaad, PhD

Committee Member

Suleiman Ashur, PhD

Committee Member

James Banfield, PhD

Abstract

The utilization of recreational drones has experienced a substantial increase in both the United States and globally. However, it is noteworthy that most drones, classified as Internet of Things devices, are produced with a limited security lifecycle. This study's findings are of paramount importance, as traditional computing exploits can be applied to drones, designating them as high- value targets. This study examines the detectability and disruptability of covert timing channel traffic in secure drones. The investigation aims to ascertain the effects of multiple interarrival times, distances ranging from 1 to 330 feet, various detection algorithms, and stream sizes between 32-bit and 256-bits on the detection of covert timing channels in drones. A comprehensive dataset was meticulously generated from covert channels utilizing the secure Parrot Anafi Ai drone across varying distances, stream sizes, and interarrival times. Machine learning and deep learning models were employed to classify covert and non-covert timing channels. Additionally, a classification based on distance was conducted to assess covert timing channels. The final phase of the experiment involved disrupting the interarrival times associated with covert timing channels. The study revealed that interarrival time, machine learning, deep learning algorithms, and disruption methods significantly influenced the detection of covert timing channels in drones. This is evident from the metrics of classification models, including accuracy, precision, recall, F1 score, and percentage of correctly decoded bits. The long short- term memory, one-dimensional convolutional neural network, and adaptive boosting models achieved a remarkable classification accuracy of 100%. In contrast, network shaping disruption resulted in a 50% reduction in correctly decoded bits. These findings possess practical implications, suggesting that machine and deep learning models demonstrate reliability in detecting covert timing channels in drones. Moreover, slight network time shaping adversely affects the sensitivity of interarrival times. This research unveils the discrepancy between the rapid growth of recreational drones and their weakened security posture, offering a strategic roadmap for researchers, drone manufacturers, regulators, and corporations to enhance drone cybersecurity through an initial analysis of timing channels and network shaping.

Download

Share

COinS