边信道攻击和学习向量量化(英文)
发布时间:2021-04-09 22:29
尽管加密算法已得到改进,加密系统的安全性仍然是密码系统设计者关注的重点。边信道攻击可利用加密系统的物理漏洞来获取秘密信息。目前提出的多种边信道信息分析方法中,机器学习被认为是一种有前景的方法。基于神经网络的机器学习可获得指令标志(功耗与电磁辐射),并自动识别。本文对椭圆曲线加密(Elliptic curve cryptography,ECC)的现场可编程门阵列(field-programmable gate array,FPGA)实现展开了新的实验研究,探讨了基于学习向量量化(Learning vector quantization,LVQ)神经网络的边信道信息表征的效率。LVQ作为多类分类器的主要特点是它具有学习复杂非线性输入-输出关系、使用顺序训练程序和适应数据的能力。实验结果表明基于LVQ的多类分类是边信道数据表征的强大且有前景的方法。
【文章来源】:Frontiers of Information Technology & Electronic Engineering. 2017,18(04)EISCICSCD
【文章页数】:9 页
【文章目录】:
1 Introduction
2 Neural networks as multi-class classi-?ers
2.1 Side-channel attacks based on neural net-works
3 Multi-class classi?cation based on learning vector quantization
3.1 Learning vector quantization algorithm
4 Experimental results based on learn-ing vector quantization
4.1 Experimental setup
4.2 Empirical results and discussions
5 Conclusions
【参考文献】:
期刊论文
[1]一套具备使用者不可追踪性的轻量化身分鉴别机制(英文)[J]. Kuo-Hui YEH. Frontiers of Information Technology & Electronic Engineering. 2015(04)
本文编号:3128426
【文章来源】:Frontiers of Information Technology & Electronic Engineering. 2017,18(04)EISCICSCD
【文章页数】:9 页
【文章目录】:
1 Introduction
2 Neural networks as multi-class classi-?ers
2.1 Side-channel attacks based on neural net-works
3 Multi-class classi?cation based on learning vector quantization
3.1 Learning vector quantization algorithm
4 Experimental results based on learn-ing vector quantization
4.1 Experimental setup
4.2 Empirical results and discussions
5 Conclusions
【参考文献】:
期刊论文
[1]一套具备使用者不可追踪性的轻量化身分鉴别机制(英文)[J]. Kuo-Hui YEH. Frontiers of Information Technology & Electronic Engineering. 2015(04)
本文编号:3128426
本文链接:https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/3128426.html