Project Watch Our Your Phone Calls: Inference Attacks Over Encrypted 5G Calls Program DUIRI - Discovery Undergraduate Interdisciplinary Research Internship Term Spring 2024 Status Accepted Research Area Security; Machine Learning (ML); 5G network; Description This project aims to expose potential threats to 5G voice phone calls which have been overlooked by the community and the public. 5G voice calls are encrypted and thus anticipated to be secure (at least they should not be eavesdropped to leak confidential call information). However, our recent research shows that it is not the case, particularly when 5G voice calls are made to an answering machine. We find that these phone calls follow certain patterns, which can be inferred through advanced machine learning techniques. In this project, we aim to develop ML techniques to infer 5G call information over radio signals captured in the air. Supervisor Chunyi Peng Mentor Chunyi Peng Type of work required - Work with the research team and participate the project meeting every week;
- Run real-world experiments with US operators to collect traces with different types of 5G voice calls;
- Process data traces and develop ML techniques to improve the inference accuracy;
- Deliver a demo or a publication for this project.
Websites https://www.cs.purdue.edu/homes/chunyi/index.html Qualifications Major in CS/DS/ML preferred; GPA >= 3.5; Courses or knowledge on ML (say, classification, neural networks etc); Python Programming; Android Programming is a PLUS Credit Hours 3 Weekly Hours 20 (estimated)