This $1.2 million National Science Foundation project grant will support research into physically unclonable wireless systems utilizing chaotic antenna arrays at the University of South Florida from February 2023 to January 2027.
The goal is to develop various configurations of chaotic antenna arrays that introduce unclonable randomizations into their geometry. These arrays are intended to exhibit unique radio frequency fingerprints that can authenticate wireless devices joining a network and enhance physical layer security by reducing an eavesdropper's ability to interpret intercepted signals. Machine learning methods like deep neural networks will be used to perform accurate authentication based on the enhanced radio frequency fingerprints. Information theoretic bounds will also be analyzed regarding authentication and attack success probabilities. Potential adversarial machine learning attacks and defense strategies will be investigated. The project aims to advance computer and information science and engineering objectives within the NSF's Computer and Information Science and Engineering program (CFDA #47.070).
Generated 1/6/24, 10:14 PM