%0 Journal Article
%T 基于MLP的AKCN_MLWE算法侧信道分析<br>MLP-Based AKCN_MLWE Algorithm Side Channel Analysis
%A 尹源源
%A 吴震
%J Advances in Applied Mathematics
%P 1429-1437
%@ 2324-8009
%D 2023
%I Hans Publishing
%R 10.12677/AAM.2023.124146
%X 在量子计算机背景下，Peter Shor提出的多项式时间算法使现有的公钥密码体制面临严重威胁，因此需要研究后量子密码算法。后量子密码算法可以抵抗量子计算机的威胁，但在实际应用中容易受到侧信道攻击。本文分析了AKCN-MLWE算法在STM32F1开发板上的实现，针对该算法解密过程中消息解码时的侧信道脆弱点，提出一种结合机器学习的侧信道分析方案。实验表明，使用PCA降维方式比SOSD提取兴趣点方式攻击效果更好。<br />
In the context of quantum computers, the polynomial time algorithm proposed by Peter Shor poses a serious threat to the existing public-key cryptography, so post-quantum cryptography algorithms need to be studied. Post-quantum cryptography algorithms can resist the threat of quantum com-puters, but are vulnerable to side-channel attacks in practical applications. This paper analyzes the implementation of AKCN-MLWE algorithm on STM32F1 development board, and proposes a side-channel analysis scheme combined with machine learning for the side-channel vulnerability point during message decoding during the decryption process of the algorithm. Experiments show that PCA dimensionality reduction is better than SOSD extraction of points of interest.
%K 侧信道分析，模板攻击，后量子密码，多层感知器，AKCN-MLWE<br>Side Channel Analysis
%K Template Attacks
%K Post-Quantum Cryptography
%K Multilayer Perceptron
%K AKCN-MLWE
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=63744