基于深度学习的侧信道攻击:进展与问题

Progress and Problems of Side Channel Attack Based on Deep Learning

深度学习的兴起,促进了侧信道攻击技术的发展。相比于传统的侧信道攻击方法,基于深度学习的侧信道攻击具有攻击准确率高、效果好、能克服典型防护技术等优势,对密码设备的安全性构成了严重的威胁,但也存在着一些亟待解决的问题。文章从基于深度学习侧信道攻击技术的基本原理入手,介绍了攻击的具体流程,接着分析了基于深度学习侧信道攻击中面临的问题并给出了相应的解决方案,最后对全文进行了总结。

The rise of Deep Learning has promoted the development of side channel attack.Compared with the traditional method,the side channel attack based on Deep Learning has the advantages of high attack accuracy,good effect,and breaking the typical protection technology.It poses a serious threat to the security of cryptographic devices.However,there are some urgent problems to be solved.This paper starts from the basic principle of side channel attack technology based on Deep Learning,introduces the specific process,then analyzes the problems faced in side channel attack based on Deep Learning and gives corresponding solutions,and finally summarizes the full paper.