Library
Repository Home account_circle Login

Kilroy Was Here

The First Step Towards Explainability of Neural Networks in Profiled Side-Channel Analysis

Conference Paper (2021)
Author(s)

Daan Valk (TU Delft - Electrical Engineering, Mathematics and Computer Science)

Stjepan Picek (TU Delft - Electrical Engineering, Mathematics and Computer Science)

Shivam Bhasin (Nanyang Technological University)

Research Group
Cyber Security
DOI related publication
https://doi.org/10.1007/978-3-030-68773-1_9 Final published version
More Info
expand_more
Publication Year
2021
Language
English
Research Group
Cyber Security
Volume number
12244
Pages (from-to)
175-199
Publisher
Springer
ISBN (print)
9783030687724
Event
11th International Workshop on Constructive Side-Channel Analysis and Secure Design, COSADE 2020 (2020-04-01 - 2020-04-03), Lugano, Switzerland
Downloads counter
139

Abstract

While several works have explored the application of deep learning for efficient profiled side-channel analysis, explainability, or, in other words, what neural networks learn remains a rather untouched topic. As a first step, this paper explores the Singular Vector Canonical Correlation Analysis (SVCCA) tool to interpret what neural networks learn while training on different side-channel datasets, by concentrating on deep layers of the network. Information from SVCCA can help, to an extent, with several practical problems in a profiled side-channel analysis like portability issue and criteria to choose a number of layers/neurons to fight portability, provide insight on the correct size of training dataset and detect deceptive conditions like over-specialization of networks.