Towards Efficient Training in Deep Learning Side-Channel Attacks

Datasets used to develop my Master's Thesis Towards Efficient Training in Deep Learning Side-Channel Attacks at Politecnico di Milano. The datasets contain power consumption measurements taken from multiple STM32F4 devices (3) considering multiple keys (11) while executing AES-128. unprotected-AES.zip contains the traces related to the execution of a software unprotected implementation of AES-128. masked-AES.zip contains the traces related to the execution of a software masked implementation of AES-128.本数据集用于支撑我在米兰理工大学（Politecnico di Milano）完成的题为《面向深度学习侧信道攻击（Deep Learning Side-Channel Attacks）的高效训练》的硕士学位论文。该数据集包含从3台STM32F4开发板采集的功耗测量数据，实验过程中使用了11组不同密钥，并运行了128位高级加密标准（AES-128）算法。其中，unprotected-AES.zip 存储了软件无保护实现的AES-128对应的功耗轨迹；masked-AES.zip 则存储了软件掩码实现的AES-128对应的功耗轨迹。