Class template attack method based on deep learning convolutional neural network

Abstract

The invention discloses a class template attack method based on a deep learning convolutional neural network. The method aims at characteristics of energy traces and types of Hamming weights, a deep learning convolutional neural network with a unique five-layer structure is built and is used for predicting Hamming weight of the attack energy trace; the dependence of traditional template attack ontemplate construction is well solved, the deep learning convolutional neural network is used as a class template for predicting the energy trace Hamming weight, the problem that template constructionis inaccurate is solved, the attack success rate is increased, and meanwhile the energy trace Hamming weight prediction method has high generalization capacity.

Description

technical field [0001] The invention belongs to the technical field of cryptographic algorithm analysis and detection, and more specifically, relates to a template-like attack method based on a deep learning convolutional neural network. Background technique [0002] Template attack is a very powerful side channel attack method. In the prior art, template attack is divided into two stages: template construction and template matching. [0003] In the Chinese invention patent application published on January 22, 2019, the publication number is CN109257160A, and the name is "a side-channel template attack method based on decision tree", a template attack method is announced: first collect energy traces, Establish the energy consumption matrix U, build a decision tree model, select the features that appear from the root node to the leaf node, and then extract the energy consumption matrix U according to the selected features, establish the energy consumption matrix T, and then ...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to overcome the inaccuracy of template construction in the prior art, and to propose a class template based on deep learning convolutional neural network based on the premise that the signal noise of the analyzed cryptographic equipment obeys the probability model of multivariate Gaussian distribution. Attack method to increase attack success rate

Images