PhD Thesis

Identical IoT device identification via hardware performance fingerprinting and Machine Learning

Identificación de dispositivos IoT idénticos mediante fingerprinting del rendimiento del hardware y Machine Learning

Download Thesis (PDF)

Summary

In the evolving landscape of cybersecurity, the identification and protection of Internet of Things (IoT) devices have become paramount. This thesis focuses on device behavior fingerprinting, especially for resource-constrained systems such as single-board computers used in Smart Cities, Industry 4.0, and the Internet of Battlefield Things (IoBT).

Key research questions addressed include:

• Can ML/DL techniques effectively identify individual IoT devices?
• How resilient are these techniques to adversarial attacks?
• Can advanced models like transformers improve fingerprinting?

The methodology involves extensive data collection from Raspberry Pi models and the development of LwHBench, a tool for low-level benchmarking of CPU, GPU, memory, and storage. This dataset supports ML/DL training for device identification.

Another major contribution is the SpecForce framework, integrating hardware fingerprinting with behavioral analysis to detect malware and SSDF attacks in IoBT. Adversarial defense strategies and authentication frameworks based on transformers are also explored.

This work advances the field by:

• Establishing new methodologies, datasets, and benchmarks
• Demonstrating robust ML/DL-based fingerprinting methods
• Validating integration with cybersecurity frameworks in real-world scenarios

Publications

• Sánchez Sánchez, Pedro Miguel et al. (2021). A survey on device behavior fingerprinting: Data sources, techniques, application scenarios, and datasets. IEEE Communications Surveys & Tutorials, 23(2), 1048-1077. IF: 33.84 (D1). DOI

• Sánchez Sánchez, Pedro Miguel et al. (2023). A methodology to identify identical single-board computers based on hardware behavior fingerprinting. Journal of Network and Computer Applications, 103579. IF: 8.7 (D1). DOI

• Sánchez Sánchez, Pedro Miguel et al. (2023). LwHBench: A low-level hardware component benchmark and dataset for Single Board Computers. Internet of Things, 22, 100764. IF: 5.9 (Q1). DOI

• Sánchez Sánchez, Pedro Miguel et al. (2022). SpecForce: A Framework to Secure IoT Spectrum Sensors in the Internet of Battlefield Things. IEEE Communications Magazine. IF: 11.2 (D1). DOI

• Sánchez Sánchez, Pedro Miguel et al. (2024). Adversarial attacks and defenses on ML-and hardware-based IoT device fingerprinting and identification. Future Generation Computer Systems, 152, 30-42. IF: 7.5 (D1). DOI

• Sánchez Sánchez, Pedro Miguel et al. (2024). Single-board device individual authentication based on hardware performance and autoencoder transformer models. Computers & Security, 137, 103596. IF: 5.6 (Q2). DOI

Directors

• Alberto Huertas Celdrán
  Postdoctoral Researcher, University of Zürich / University of Murcia.
  Webpage

• Gregorio Martínez Pérez
  Full Professor, University of Murcia.
  Webpage

Examining Board

• Juan Tapiador (Committee President)
  Full Professor, University Charles III.
  Webpage

• Lorenzo Fernández Maimó (Committee Secretary)
  Professor, University of Murcia.
  Webpage

• Vincent Lenders (Committee Member)
  Director of the Cyber-Defence Campus.
  Webpage

Substitutes:

• Pedro Peris López (Substitute President)
  Professor, University Charles III.
  Webpage

• Félix Gómez Mármol (Substitute Secretary)
  Professor, University of Murcia.
  Webpage

• Joaquín García Alfaro (Substitute Member)
  Full Professor, Telecom SudParis.
  Webpage

Experts Supporting the PhD

• Kallol Krishna Karmakar
  Postdoctoral Researcher, University of Newcastle.
  Webpage

• Muriel Figueredo Franco
  Postdoctoral Researcher, Federal University of Rio Grande do Sul.
  Webpage