A comparison of weight initializers in deep learning-based side-channel analysis

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review

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 weight initializers’ choice 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.

Original languageEnglish
Title of host publicationApplied Cryptography and Network Security Workshops - ACNS 2020 Satellite Workshops, AIBlock, AIHWS, AIoTS, Cloud S and P, SCI, SecMT, and SiMLA, Proceedings
EditorsJianying Zhou, Chuadhry Mujeeb Ahmed, Mauro Conti, Eleonora Losiouk, Man Ho Au, Lejla Batina, Zhou Li, Jingqiang Lin, Bo Luo, Suryadipta Majumdar, Weizhi Meng, Martín Ochoa, Stjepan Picek, Georgios Portokalidis, Cong Wang, Kehuan Zhang
PublisherSpringer
Pages126-143
Number of pages18
ISBN (Print)9783030616373
DOIs
Publication statusPublished - 2020
Event2nd ACNS Workshop on Application Intelligence and Blockchain Security, AIBlock 2020, 1st ACNS Workshop on Artificial Intelligence in Hardware Security, AIHWS 2020, 2nd ACNS Workshop on Artificial Intelligence and Industrial IoT Security, AIoTS 2020, 2nd ACNS Workshop on Cloud Security and Privacy, Cloud S and P 2020, 1st ACNS Workshop on Secure Cryptographic Implementation, SCI 2020, 1st ACNS Workshop on Security in Mobile Technologies, SecMT 2020, and 2nd ACNS Workshop on Security in Machine Learning and its Applications, SiMLA 2020, held in parallel with the 18th International Conference on Applied Cryptography and Network Security, ACNS 2020 - Rome, Italy
Duration: 19 Oct 202022 Oct 2020

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12418 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference2nd ACNS Workshop on Application Intelligence and Blockchain Security, AIBlock 2020, 1st ACNS Workshop on Artificial Intelligence in Hardware Security, AIHWS 2020, 2nd ACNS Workshop on Artificial Intelligence and Industrial IoT Security, AIoTS 2020, 2nd ACNS Workshop on Cloud Security and Privacy, Cloud S and P 2020, 1st ACNS Workshop on Secure Cryptographic Implementation, SCI 2020, 1st ACNS Workshop on Security in Mobile Technologies, SecMT 2020, and 2nd ACNS Workshop on Security in Machine Learning and its Applications, SiMLA 2020, held in parallel with the 18th International Conference on Applied Cryptography and Network Security, ACNS 2020
CountryItaly
CityRome
Period19/10/2022/10/20

Keywords

  • Deep learning
  • Side-channel analysis
  • Weight initialization

Fingerprint Dive into the research topics of 'A comparison of weight initializers in deep learning-based side-channel analysis'. Together they form a unique fingerprint.

Cite this