Janghyuk Kauh, Changmin Jeong, Koohyung Kwon, Journal of Internet Computing and Services, Vol. 26, No. 1, pp. 221-233, Feb. 2025
10.7472/jksii.2025.26.1.221, Full Text:
Keywords: Hardware Trojan, Anomaly Detection, Electromagnetic Analysis, Side channel analysis, FPGA, ASIC

Abstract

This paper demonstrates the possibility of a non-destructive method that inspects the assembled state of the equipment circuit rather than a destructive method such as disassembling the equipment circuits, using side-channel analysis to detect abnormal behaviors in weapon systems and IoT equipment circuits. The abnormal behaviors generated in weapon system circuits are caused by hardware Trojans (HTs), and electromagnetic signal analysis was used among the side-channel analysis methods. The initial experiments were conducted with the AES-128 GM, an encryption module provided by the HT sharing portal, and HTs. The target hardware platform was first implemented in the FPGA, and then tested on the manufactured ASIC chip. Initially, the tests were conducted using supervised learning machine learning as an abnormal behavior detection algorithm, and then unsupervised learning deep learning models were used. After proving the possibility with the FPGA and the manufactured ASIC chip, the tests were conducted on the RC car as IoT equipment, and the possibility of the circuit abnormal behavior detection technology was demonstrated by conducting experiments on the same product FPGA applied to the weapon system.

Statistics
Cite this article