Differential Power Analysis Attacks Based on the Multiple Correlation Coefficient
Xiaoke Tang, Jie Gan, Jiachao Chen, Junrong Liu · 2016
In this paper we research a new power analysis attack method: the differential power analysis attacks based on the multiple correlation coefficient (MCPA).This attack method is based on statistical theory that multiple linear regression.Conditions of the MCPA method of attack are weaker than ones of CPA attacks.In this paper, we indicate that MCPA attacks have same effect with CPA attacks in this weak condition.We perform the simulated attacks and the practical attacks for the two kinds of attack methods to verify this conclusion.The results show that the conclusion is correct for the first-order DPA.We also describe the CPA attacks using simple linear regression model for comparison of two attack methods.The experiments results show that this chosen traces attack is stronger than the previous methods.