A Novel Method of Fisher Discriminant Analysis for Hardware Trojan Horse Detection

Abstract

With the rise and rapid development of the network, security issues cannot be ignored. At the same time, the hardware Trojan horse to the security of the chip caused a threat. Therefore, for the detection of hardware Trojan horse in the chip, we propose a hardware Trojan horse detection method based on Fisher discriminant analysis. This method combines the discrimination ability and uses Fisher classification algorithm to extract the small difference between information, so as to realize the detection of hardware Trojan horse. However, in practice, the interference of data and the area of Trojan horse will affect the discrimination results. We conduct discriminant analysis on the number of sample chips. The analysis shows that the proposed method can effectively detect hardware Trojan horse to verify the effectiveness of the method.