Side Channel Analysis of Hardware Trojan Based on Smooth Filtering Algorithm

Abstract

Aimed at the problem of the noise in the side channel analysis (SCA) of hardware Trojan, a novel idea to develop an optimal algorithm model based on smooth filtering is discussed in this paper. To reduce the influence of the abnormal points on the curve, the idea uses the least square method to fit the original data curve. The approach is validated using power simulation with a 128-bit Advanced Encryption Standard (AES) cipher circuit and hardware Trojan circuit whose Trojan's area is 0.16%. The experiment shows that the impact of the noise on power is reduced effectively by using the proposed smooth filtering algorithm, especially for the small-area Trojan circuit.