This content will become publicly available on August 11, 2025
Scalable Physics-Embedded Neural Networks for Real-Time Robotic Control in Embedded Systems
- Award ID(s):
- 2208573
- PAR ID:
- 10545457
- Publisher / Repository:
- IEEE Midwest Symposium on Circuit and System
- Date Published:
- ISBN:
- 979-8-3503-8717-9
- Page Range / eLocation ID:
- 823 to 827
- Format(s):
- Medium: X
- Location:
- Springfield, MA, USA
- Sponsoring Org:
- National Science Foundation
More Like this
-
One aspect of system security is evaluating a system’s vulnerability to Trojan attack. A hardware Trojan attack can have potentially devastating effects, especially given the increased reliance on integrated circuits within critical systems. A significant amount of research concerns attacks on digital systems, but attacks on AMS and RF systems have recently been of interest as well. A class of Trojans has been proposed that uses undesired alternate modes of operation in nonlinear systems as the Trojan payload. These Trojans are of particular interest because they do not cause deviations from the ideal system performance and cannot be detected until the Trojan is triggered. This work addresses this class of Trojans by listing different payloads, trigger mechanisms, and examples of system architectures vulnerable to attack.more » « less
