ECE Joint Faculty Candidate Seminar: Rozhin Yasaei
Tuesday, January 16, 2024 - 2:00 pm
Rozhin Yasaei, Ph.D.
Assistant Professor
Cyber, Intelligence & Information Operations
College of Applied Science & Technology
The University of Arizona
"Applied Machine Learning for Cross-Layer System Security"
ECE 530 | Zoom link
Abstract:
The 21st century is marked by the integration of elegantly engineered components, devices, and systems that communicate across platforms, necessitating the ability to connect real-world applications with modeling techniques. This research envisions a holistic model for cross-layer security in hardware and cyber-physical systems, emphasizing the importance of multi-modal data integration and contextual understanding. The proposed strategy involves identifying system components, collecting data, understanding their relationships, and employing machine learning techniques like graph neural networks to create comprehensive models.
The current presentation focuses on cross-layer security in embedded and cyber-physical systems, particularly in the Internet of Things (IoT). A context-aware anomaly detection method is introduced, leveraging multi-modal data fusion and GNN models to enhance security. The research extends to hardware security, addressing the challenges of hardware trojan detection and IP piracy detection. Graph neural networks prove effective in these domains, offering automated solutions for detecting Trojans in pre-silicon design stages and locating trojans post-silicon using side-channel emissions. The research further extends to using GNN for IP piracy detection and design automation, showcasing the versatility and potential impact of the proposed methodologies.
Bio:
Dr. Yasaei earned her Ph.D. in 2023 from the University of California, Irvine (UCI), where her research culminated in the publication of her thesis, titled “Graph Neural Network for Integrated Circuits and Cyber-Physical Systems Security.” Her academic journey commenced with a master’s degree in computer engineering from UC Irvine in 2021, following her bachelor’s degree in electrical engineering from Sharif University of Technology in 2018. Her research pursuits revolve around the innovative application of machine learning techniques to cross-layer security, explicitly focusing on hardware, embedded systems, and cyber-physical systems. Dr. Yasaei’s work leverages cutting-edge methodologies like graph neural networks to model these systems, enriching security and design automation.