Cryptology ePrint Archive: Report 2020/904

A Comparison of Weight Initializers in Deep Learning-based Side-channel Analysis

Huimin Li and Marina Krček and Guilherme Perin

Abstract: The usage of deep learning in profiled side-channel analysis requires a careful selection of neural network hyperparameters. In recent publications, different network architectures have been presented as efficient profiled methods against protected AES implementations. Indeed, completely different convolutional neural network models have presented similar performance against public side-channel traces databases. In this work, we analyze how the choice of weight initializers influences deep neural networks' performance in the profiled side-channel analysis. Our results show that different weight initializers provide radically different behavior. We observe that even high-performing initializers can reach significantly different performance when conducting multiple training phases. Finally, we found that this hyperparameter is more dependent on the choice of dataset than other, commonly examined, hyperparameters. When evaluating the connections with other hyperparameters, the biggest connection is observed with activation functions.

Category / Keywords: applications / Weight initialization, Deep learning, Side-channel Analysis

Date: received 17 Jul 2020

Contact author: h li-7 at tudelft nl,m krcek@tudelft nl,g perin@tudelft nl

Available format(s): PDF | BibTeX Citation

Version: 20200718:161350 (All versions of this report)

Short URL: ia.cr/2020/904


[ Cryptology ePrint archive ]