一种基于支持向量机的侧信道碰撞攻击方法  

A side-channel collision attack method based on support vector machine

作　　者：张美玲[1] 袁旭鑫 郑东[1] ZHANG Meiling;YUAN Xuxin;ZHENG Dong(National Engineering Laboratory for Wireless Security,Xi’an University of Posts and Telecommunications,Xi’an 710121,China)

机构地区：[1]西安邮电大学无线网络安全技术国家工程实验室,陕西西安710121

出　　处：《西安邮电大学学报》2024年第6期73-78,共6页Journal of Xi’an University of Posts and Telecommunications

基　　金：国家自然科学基金项目(62072371);陕西省重点研发计划项目(2021ZDLGY06-04)。

摘　　要：提出一种基于支持向量机(Support Vector Machine,SVM)的侧信道碰撞攻击方法,用于对带重用掩码的高级加密标准(Advanced Encryption Standard,AES)算法进行碰撞检测和密钥恢复。在Chipwhisperer平台上采集带重用掩码防护的AES的能量迹,利用能量迹方差粗略得到各个S盒相关的最多兴趣点。将某两个S盒的兴趣点取差值后进行归一化处理得到对应的训练数据集,并利用基于三次多项式核函数的支持向量机进行训练得到分类器。在测试阶段,利用分类器给出是否碰撞的分值,判断碰撞和非碰撞两种情形。实验结果表明,在不同噪声环境下所提方法均能以更少的能量迹实现90%以上的攻击成功率,从而有效地检测碰撞。A side-channel collision attack method based on support vector machines(SVM)is proposed for collision detection and key recovery of advanced encryption standard(AES)algorithms with reused masks.The energy trace of AES with reused masks protection is collected on the Chipwhisperer platform,and the maximum interest points related to each S-box are roughly obtained by using the variance of the energy trace.After the difference between the interest points of two S-boxes is normalized,the corresponding training data set is obtained,and the classifier is trained by the Support Vector Machine based on the cubic polynomial kernel function.In the test stage,the classifier is used to score of whether there is a collision,and to judge the two cases of collision and non-collision.Experiment results show that the proposed method can achieve more than 90%attack success rate with less energy trace in different noise environments,so as to detect collisions effectively.

关 键 词：侧信道攻击 碰撞检测 支持向量机 重用掩码 高级加密标准算法 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]