Alignment of Power Trace Based on Wavelet Analysis and Cross Correlation

Abstract

In the study of power analysis attack, the unaligned signals will increase the difficulty of analysis, causing the attack is difficult to succeed. This paper presents a method of power curves alignment based on wavelet analysis and cross correlation. The method uses the wavelet transform to denoise the original power trace, and then align the trace based on the cross correlation alignment algorithm. And the effect of the alignment was verified by Euclidean distance and Correlation Power Analysis (CPA). The experimental results show that the method can achieve good data alignment effect and improve the attack success rate.