基于时间序列分析的侧信道攻击

密码设备在进行加解密时会产生大量的物理信息,如电磁、能耗、热量等。攻击者利用这些物理信息对密码设备进行侧信道攻击,窃取密码设备的密钥。就传统的椭圆曲线加密算法简单能量分析攻击而言,攻击者通过人工识别倍点计算能耗数据中的相似模块,从而恢复出密钥。这种攻击方式导致攻击效率低,准确率不高。本文基于时间序列的模体发现提出了SPAMD算法,该算法实现了倍点计算的能耗数据中相似模块的自动识别,完成了攻击方式从人工识别到自动识别的转换。实验结果表明SPAMD攻击在效率和准确率上有了显著的提升。

Side Channel Attack Based on Time Series Analysis

The cryptographic device generates a large amount of physical information such as electromagnetic,energy consumption,heat,etc.during encryption and decryption.The attackers use these physical information to perform a side channel attack on the cryptographic device and steal the key of the cryptographic device.In the case of a traditional simple power analysis attack of elliptic curve cryptography algorithm,the attacker recovers the key by manually identifying similar modules in the power consumption data of the double point calculation.But this way results in low attack efficiency and low attack accuracy.This paper proposes the SPAMD algorithm based on the time series of motif discovery.The algorithm realizes the automatic recognition of similar modules in the power consumption data of the double point calculation,and completes the conversion of the attack mode from manual recognition to automatic recognition.Experiment results show that SPAMD attacks have a significant improvement in efficiency and accuracy.