A Comprehensive Study on Cybersecurity Vulnerabilities in Connected Diabetes Devices

This research investigates cybersecurity vulnerabilities in connecteddiabetes devices, focusing on Bluetooth Low Energy (BLE) communication protocols within Continuous Glucose Monitoring (CGM) systems and the broader Internet of Medical Things (IoMT) ecosystem. The study addresses the gap between recognizing BLE weaknesses and implementing practical, real-time protection mechanisms for resource-constrained medical devices. Using a dual-dataset approach, combining public BLE traffic data and simulated CGM attack scenarios, the research develops and evaluates a supervised machine learning framework for multi-class threat classification. Key attack vectors such as denial-of-service, GATT abuse and link layer abuse are analyzed through statistical tests and model-based experiments. The methodology includes feature engineering, rigorous data preprocessing, and comparative analysis of classifiers, with a focus on accuracy and computational efficiency. Results demonstrate that BLE attack traffic exhibits distinct feature signatures, and that a tuned Decision Tree model achieves high classification accuracy (macro-F1 > 0.95) with minimal inference latency, making it suitable for deployment in low-power IoMT devices. The findings support the development of scalable, explainable, and resource-efficient intrusion detection systems, contributing to improved patient safety, data integrity, and regulatory compliance in connected healthcare environments.

Author

• Aitlafquih, Abderrahim

Language

• en

Date created

• 2025

Type of Work

• Dissertation

License

• All rights reserved

Rights statement

• In Copyright

GW Unit

• Cybersecurity Analytics

Degree

• D.Engr.

Advisor

• Togwe, Thembani

Committee Member(s)

• Fossaceca, John
• Clements, Kenneth
• Togwe, Thembani

Persistent URL

•  https://scholarspace.library.gwu.edu/etd/d217qq512