Deep-Learning Side-Channel Attacks
Date:
About the Event
The AI for Cybersecurity Research Lunch is a weekly event at Texas A&M University where students and researchers present their latest work in cybersecurity and AI. Talks are open to all, but regular attendees are expected to present their research at least once.
đź“… Date: February 5, 2025
đź•› Time: 12:00 PM - 1:00 PM
đź“Ť Location: PETR 214, Texas A&M University
Speaker
Mabon Ninan - PhD Student @ Texas A&M University
Mabon Ninan is a first-year PhD student working under Dr. Marcus Botacin, focusing on machine learning for malware detection. His research builds on his prior work at the University of Cincinnati, where he explored deep learning-based side-channel attacks.
Abstract
A side-channel attack exploits unintentional information leaks—such as power consumption, timing variations, or electromagnetic emissions—to infer sensitive data from a system. Recent studies have demonstrated that deep learning can significantly enhance the success of these attacks compared to traditional statistical methods. However, deep-learning-based attacks are highly sensitive to software and hardware discrepancies, making their real-world applicability more complex than initially assumed.
This talk provides an in-depth analysis of how software and physical discrepancies affect side-channel attacks, discussing their implications for attack portability. We explore strategies for building more robust and transferable attacks, ensuring effectiveness across diverse environments.
Event Details
- Mailing List: aicybersecurity-research-lunch@lists.tamu.edu
- Previous Meetings: View All Past Events