基于GHM多小波算法的功耗分析攻击

功耗分析的密钥获取是基于采集的功耗信号，功耗信号的信噪比是影响分析密钥成功率的重要因素，所以噪声能否被有效去除是提高功耗分析成功率的关键，针对该问题引入了基于GHM多小波的预处理方法。该方法首先对功耗曲线进行GHM多小波阈值去噪处理，其目的是最大限度地去除功耗曲线中不相关的噪声，提高功耗曲线中真实信号的信噪比，从而提高攻击效率。在MEGA16微控制器上，采集固定密钥随机明文的ASE算法的功耗曲线，对比原始功耗曲线与去噪后的功耗曲线执行相关功耗分析。实验结果表明，使用去噪后的功耗曲线执行相关功耗分析所需的功耗曲线减少了89.5%，相关系数平均提高了107.9%，验证了新方法的有效性。

Power Analysis Attack Based on GHM Multiwavelet Algorithm

In power analysis, key acquisition for power analysis is based on the collected power signal, and one of the most important factors impacting the success rate of key analysis is the signal to noise ratio of real power consumption. So the noise can be effectively remove is the key to improve the success rate of power analysis, to solve this problem based on the preprocessing method of GHM multiwavelet. This method is to denoise power traces by GHM multiwavelet thresholding, with an aim to remove irrelevant noise from the power traces as far as possible, and raise the signal to noise ratio of real signal in the power traces. Power traces of AES algorithm were collected in MEGA16 micro controller hardware platform for the same key with different plaintexts. Perform correlation power analysis with original power traces and the denoised power traces. Experimental results show that the power traces required for correlation power analysis performed with the denoised power traces was reduced by89.5%, and the correlation coefficient was raised by 107.9% on average; this verifies the effectiveness of the new method.