Artificial Intelligence and Hardware Security
Artificial Intelligence (AI) and hardware security are two important topics nowadays. Recently, researchers have been focusing on applying AI technologies into security applications. For example, deep neural network (DNN) are not just widely deployed in various security-crucial scenarios, including image recognition, natural language processing and autonomous vehicles but also in hardware security domains such as network intrusion detection, hardware Trojan identification, etc. AI techniques help strengthen existing hardware security solutions. However, a deep look into both areas would reveal that the relations between AI and hardware security are more complicated.
In this talk, I will introduce two recently work from my team to show that, if not properly addressed, AI and hardware security may compromise each other. For example, due to economic and privacy concerns, the hardware implementations of structures and designs inside DNN accelerators are usually inaccessible to the public. However, these accelerators still tend to leak crucial information through Electromagnetic (EM) side channels in addition to timing and power information. Along this direction, we will prove that EM-based side channel can reveal large-scale DNN designs. On the contrary, AI and DNN techniques can enhance the hardware level attacks such as the previously mentioned EM-based side channel attacks. We will then show that supported by DNN techniques, side channel attacks become more powerful and may require much less EM traces. Further, AI-based side channel attacks may even invalidate existing hardware protection methods so that new protection solutions are needed.
Date and Time
Location
Hosts
Registration
-
Add Event to Calendar
Loading virtual attendance info...
- 1000 River Road
- Teaneck , New Jersey
- United States 07666
- Building: Muscarelle Center, M105,
- Room Number: M105
- Contact Event Host
- Co-sponsored by North Jersey Section, Signal Processing Chapter
- Starts 13 October 2020 02:35 PM UTC
- Ends 11 November 2020 07:00 PM UTC
- No Admission Charge
Speakers
Dr. Yier Jin of University of Florida
Artificial Intelligence and Hardware Security
Biography:
Dr. Yier Jin is the Endowed IoT Term Professor in the Warren B. Nelms Institute for the Connected World and also an Associate Professor in the Department of Electrical and Computer Engineering (ECE) in the University of Florida (UF). Prior to joining UF, he was an assistant professor in the ECE Department at the University of Central Florida (UCF). He received his PhD degree in Electrical Engineering in 2012 from Yale University after he got the B.S. and M.S. degrees in Electrical Engineering from Zhejiang University, China, in 2005 and 2007, respectively. His research focuses on the areas of embedded systems and Internet of Thing (IoT) design and security, trusted hardware intellectual property (IP) cores and hardware-software co-design for modern computing systems. He is also interested in artificial intelligence (AI) security and its applications in hardware domain. Dr. Jin received Young Investigator Grant from Southeastern Center for Electrical Engineering Education (SCEEE) in 2015, early CAREER award from Department of Energy (DoE) in 2016, Outstanding New Faculty Award (ONFA) from ACM's Special Interest Group on Design Automation (SIGDA) in 2017, Young Investigator Award (YIP) from Office of Naval Research (ONR) in 2019, and Ernest S. Kun Early Career Award from IEEE Council on Electronic Design Automation (CEDA) in 2020. He also received the Best Paper Award of the 52nd Design Automation Conference (DAC) in 2015, the 21st Asia and South Pacific Design Automation Conference (ASP-DAC) in 2016, the 10th IEEE Symposium on Hardware-Oriented Security and Trust (HOST) in 2017, the 2018 ACM Transactions on Design Automation of Electronic Systems (TODAES), the 28th edition of the ACM Great Lakes Symposium on VLSI (GLSVLSI) in 2018, and the Design, Automation and Test in Europe Conference and Exhibition (DATE) in 2019. He is the IEEE CEDA Distinguished Lecturer. He is also a senior member of IEEE.
Agenda
Artificial Intelligence (AI) and hardware security are two important topics nowadays. Recently, researchers have been focusing on applying AI technologies into security applications. For example, deep neural network (DNN) are not just widely deployed in various security-crucial scenarios, including image recognition, natural language processing and autonomous vehicles but also in hardware security domains such as network intrusion detection, hardware Trojan identification, etc. AI techniques help strengthen existing hardware security solutions. However, a deep look into both areas would reveal that the relations between AI and hardware security are more complicated.
In this talk, I will introduce two recently work from my team to show that, if not properly addressed, AI and hardware security may compromise each other. For example, due to economic and privacy concerns, the hardware implementations of structures and designs inside DNN accelerators are usually inaccessible to the public. However, these accelerators still tend to leak crucial information through Electromagnetic (EM) side channels in addition to timing and power information. Along this direction, we will prove that EM-based side channel can reveal large-scale DNN designs. On the contrary, AI and DNN techniques can enhance the hardware level attacks such as the previously mentioned EM-based side channel attacks. We will then show that supported by DNN techniques, side channel attacks become more powerful and may require much less EM traces. Further, AI-based side channel attacks may even invalidate existing hardware protection methods so that new protection solutions are needed. The Zoom link is as follows:
https://fdu.zoom.us/j/97888412349
Meeting ID: 978 8841 2349