Beyond TVLA: Anderson-Darling Leakage Assessment for Neural Network Side-Channel Leakage Detection

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Mikulec, Ján, Breier, Jakub, Hou, Xiaolu
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908900861673472
author Mikulec, Ján
Breier, Jakub
Hou, Xiaolu
author_facet Mikulec, Ján
Breier, Jakub
Hou, Xiaolu
contents Test Vector Leakage Assessment (TVLA) based on Welch's $t$-test has become a standard tool for detecting side-channel leakage. However, its mean-based nature can limit sensitivity when leakage manifests primarily through higher-order distributional differences. As our experiments show, this property becomes especially crucial when it comes to evaluating neural network implementations. In this work, we propose Anderson--Darling Leakage Assessment (ADLA), a leakage detection framework that applies the two-sample Anderson--Darling test for leakage detection. Unlike TVLA, ADLA tests equality of the full cumulative distribution functions and does not rely on a purely mean-shift model. We evaluate ADLA on a multilayer perceptron (MLP) trained on MNIST and implemented on a ChipWhisperer-Husky evaluation platform. We consider protected implementations employing shuffling and random jitter countermeasures. Our results show that ADLA can provide improved leakage-detection sensitivity in protected implementations for a low number of traces compared to TVLA.
format Preprint
id arxiv_https___arxiv_org_abs_2603_18647
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Beyond TVLA: Anderson-Darling Leakage Assessment for Neural Network Side-Channel Leakage Detection
Mikulec, Ján
Breier, Jakub
Hou, Xiaolu
Cryptography and Security
Artificial Intelligence
Test Vector Leakage Assessment (TVLA) based on Welch's $t$-test has become a standard tool for detecting side-channel leakage. However, its mean-based nature can limit sensitivity when leakage manifests primarily through higher-order distributional differences. As our experiments show, this property becomes especially crucial when it comes to evaluating neural network implementations. In this work, we propose Anderson--Darling Leakage Assessment (ADLA), a leakage detection framework that applies the two-sample Anderson--Darling test for leakage detection. Unlike TVLA, ADLA tests equality of the full cumulative distribution functions and does not rely on a purely mean-shift model. We evaluate ADLA on a multilayer perceptron (MLP) trained on MNIST and implemented on a ChipWhisperer-Husky evaluation platform. We consider protected implementations employing shuffling and random jitter countermeasures. Our results show that ADLA can provide improved leakage-detection sensitivity in protected implementations for a low number of traces compared to TVLA.
title Beyond TVLA: Anderson-Darling Leakage Assessment for Neural Network Side-Channel Leakage Detection
topic Cryptography and Security
Artificial Intelligence
url https://arxiv.org/abs/2603.18647