针对AES加密算法的安全检测

Security Detection of AES Encryption Algorithm

侧信道攻击(side channel attack, SCA)是一种新兴的密码分析方法,主要通过加密软件或硬件运行时产生的各种泄漏信息获取密文信息,其中相关功耗分析(CPA)是较为强大的一种攻击方法,可以用来实现AES加密算法的安全检测。CPA需要假设猜测密钥,然后根据示波器采集到的能量迹,使用具有数据相关性的汉明模型,并计算实际能量值与假设能量消耗之间的皮尔逊相关系数,利用皮尔逊相关系数来判断猜测密钥的正确与否。CPA可以把皮尔逊相关系数的计算结果限制在[-1,1]之间且CPA过程自带标准化,不需要额外对数据进行标准化。此外,还利用python的pandas库和style方法对获得的结果进行刻画,更好地实现了相关系数和正确的密钥之间的关系。本案例使用NewAE Technology Inc的芯片物理攻击平台ChipWhisperer实施CPA攻击来检测加密算法的安全性,成功破解了预设的AES-128的16字节的密钥。实验表明,CPA在针对未加防护的AES加密算法时有显著的效果,可以检测AES加密算法的安全性。

Side channel attack(SCA) is a new cryptanalysis method, which mainly obtains ciphertext information through various leakage information generated by encryption software or hardware during operation. Among them, correlation power analysis(CPA) is a relatively powerful attack method, which can be used to realize the security detection of AES encryption algorithm. CPA needs to guess the key hypothesis, and then according to the energy trace collected by the oscilloscope, the Hamming model with data correlation is used, and the Pearson correlation coefficient between the actual energy value and the assumed energy consumption is calculated, and the Pearson correlation coefficient is used to judge the correctness of the guess key. CPA can limit the calculation result of Pearson correlation coefficient to between [-1,1] and the CPA process has its own standardization, so there is no need for additional standardization of data. In addition, we also use Python’s Pandas library and style method to characterize the results obtained, better realizing the relationship between the correlation coefficient and the correct key. This case uses NewAE Technology Inc’s chip physical attack platform ChipWhisperer to implement CPA attack to detect the security of encryption algorithm, and successfully cracked the preset 16-byte key of AES-128. The experiment shows that CPA has a significant effect on the unprotected AES encryption algorithm, and can detect the security of the AES encryption algorithm.