Collaborative Research: SaTC: CORE: Small: Towards Robust, Scalable, and Resilient Radio FingerprintingProject Information
Project OverviewThis project develops new methods to enhance the robustness, scalability, and resilience of radio fingerprinting in wireless networks. Radio fingerprinting authenticates wireless devices over radio frequency (RF) signals at the physical layer based on hardware variations from manufacturing. This project adopts a cross-layer approach, which synergizes signal processing at the physical layer and deep learning at the data layer. This research project includes three thrusts: (1) developing a new robust radio fingerprinting method by designing complex-valued triplet neural networks. This method can achieve high accuracy when a classifier is trained with RF signals from one day but is tested with RF signals from a different day; (2) building a new receiver-agnostic radio fingerprinting method by building Physical-Layer-Assisted Generative Adversarial Networks. This method can train classifiers that can be utilized by different receivers; (3) developing a new cross-domain radio fingerprinting method by building neural networks across the time domain, frequency domain, and time-frequency domain. This method will be resilient against adversarial RF signals perturbed by white-box attackers. This project will integrate the research activities into curriculum development, render research opportunities to female and underrepresented students, and advance research experience for students in STEM (Science, Technology, Engineering and Math). Publications
|