Perceived Information Revisited II: Information-Theoretical Analysis of Deep-Learning Based Side-Channel Attacks

ePrint Report: Perceived Information Revisited II: Information-Theoretical Analysis of Deep-Learning Based Side-Channel Attacks

Akira Ito, Rei Ueno, Naofumi Homma

In conventional deep-learning-based side-channel attacks (DL-SCAs), an attacker trains a model by updating parameters to minimize the negative log-likelihood (NLL) loss function. Although a decrease in NLL improves DL-SCA performance, the reasons for this improvement remain unclear because of the lack of a formal analysis. To address this open problem, this paper explores the relationship between NLL and the attack success …

akira analysis attacker attacks channel eprint report function information log loss performance report sca side-channel side-channel attacks trains