Paper 2023/1110

Breaking Free: Leakage Model-free Deep Learning-based Side-channel Analysis

Abstract

Profiling side-channel analysis has gained widespread acceptance in both academic and industrial realms due to its robust capacity to unveil protected secrets, even in the presence of countermeasures. To harness this capability, an adversary must access a clone of the target device to acquire profiling measurements, labeling them with leakage models. The challenge of finding an effective leakage model, especially for a protected dataset with a low signal-to-noise ratio or weak correlation between actual leakages and labels, often necessitates an intuitive engineering approach, as otherwise, the attack will not perform well. In this paper, we introduce a deep learning approach that does not assume any specific leakage model, referred to as the multibit model. Instead of trying to learn a representation of the target intermediate data (label), we utilize the concept of the stochastic model to decompose the label into bits. Then, the deep learning model is used to classify each bit independently. This versatile multibit model can align with existing leakage models like the Hamming weight and Most Significant Bit leakage models while also possessing the flexibility to adapt to complex leakage scenarios. To further improve the attack efficiency, we extend the multibit model to simultaneously attack all 16 subkey bytes, which requires negligible computational effort. Based on our preliminary analysis, two of the four considered datasets could only be broken using a Hamming Weight leakage model. Using the same model, the proposed methods can efficiently crack all key bytes across four considered datasets. Our work, thus, signifies a significant step forward in deep learning-based side-channel attacks, showcasing a high degree of flexibility and efficiency without any presumption of the leakage model.