cs.CRMar 31, 2026

Deep Learning-Assisted Improved Differential Fault Attacks on Lightweight Stream Ciphers

Authors: Kok Ping Lim, Dongyang Jia, Iftekhar Salam

Organizations: School of Computing and Data Science, Xiamen University Malaysia, Sepang 43900, Malaysia

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Lightweight cryptographic primitives are widely deployed in resource-constrained environments, particularly in Internet of Things (IoT) devices. Due to their public accessibility, these devices are vulnerable to physical attacks, especially fault attacks. Recently, deep learning-based cryptanalytic techniques have demonstrated promising results; however, their application to fault attacks remains limited, particularly for stream ciphers. In this work, we investigate the feasibility of deep learning assisted differential fault attacks on three lightweight stream ciphers, namely ACORNv3, MORUSv2, and ATOM, under a relaxed fault model in which a single-bit bit-flipping fault is injected at an unknown location. We develop and train multilayer perceptron (MLP) models to identify the fault locations. Experimental results show that the trained models achieve high identification accuracies of 0.999880, 0.999231, and 0.823568 for ACORNv3, MORUSv2 and ATOM, respectively, and outperform traditional signature-based methods. For the secret recovery process, we introduce a threshold-based method to optimize the number of fault injections required to recover the secret information. The results show that the initial state of ACORN can be recovered with 21 to 34 faults, while MORUS requires 213 to 248 faults, with at most 6 bits of guessing. Both attacks reduce the attack complexity compared to existing works. For ATOM, the results show that it possesses a higher security margin, as the majority of state bits in the Nonlinear Feedback Shift Register (NFSR) can only be recovered under a precise control model. To the best of our knowledge, this work provides the first experimental results of differential fault attacks on ATOM.

Figures & tables

CipherMethodKeystream lengthAccuracyRef.
ACORNv3MLP (this work)152 bits0.999880Sec. 4.1
Set-based149 bits0.999774[ 29 ]
Signature-based152 bits0.999747Sec. 4.1
MORUSv2MLP (this work)384 bits0.999231Sec. 4.2
Signature-based384 bits0.952606Sec. 4.2
ATOMMLP (this work)56 bits0.823568Sec. 4.3
Table 1: Summary of fault location identification accuracy across different ciphers
CipherFault type# Faults# Recovered bitsComplexityRef.
ACORNv3Bit-flipping9Full initial state225.402^{25.40}[ 22 ]
Bit-flipping27–42Full initial statec⋅2146.5−3.52nc\cdot 2^{146.5-3.52n}[ 29 ]
Bit-flipping21–34Full initial stateNegligibleSec. 4.1
MORUSv2Random–256 state bits–[ 24 ]
Bit-flipping213–248Full initial state262^{6} (At most)Sec. 4.2
ATOMBit-flipping46Majority of NFSR state≥260\geq 2^{60}Sec. 4.3
Table 2: Summary of differential fault attacks on ACORNv3, MORUSv2, and ATOM
Figure 1: Threshold-based secret recovery process.
AccuracyPrecisionRecallF1-score
0.9998800.9998800.9998800.999880
Table 3: Performance of trained model for identifying fault locations on ACORNv3
Figure 2: Correlation between fault injection counts and equation thresholds for ACORNv3
Figure 3: Number of linear equations produced by each fault location in ACORNv3.
AccuracyPrecisionRecallF1-score
0.9992310.9992350.9992310.999231
Table 4: Performance of trained model for identifying fault locations on MORUSv2
Figure 4: Frequency of the number of faults for each successful attack.
AccuracyPrecisionRecallF1-score
0.8235680.8211860.8235680.816163
Table 5: Performance of trained model for identifying fault locations on ATOM

Explore similar work

CardsList
Aug 24, 2026

Graph Representation Learning of Lightweight IoT Ciphers

SIMON and SIMECK belong to a family of Lightweight Cryptographic Algorithms (LCAs) based on the Feistel block cipher, designed for Internet of Things (IoT) devices. As with all Feistel ciphers, they are susceptible to differential cryptanalysis, necessitating rigorous resilience evaluations. While state-of-the-art techniques leverage heuristics and sampling to improve efficiency, little work has applied Machine Learning (ML) guided Graph Representation Learning (GRL) to efficiently identify and visualise high-probability differential clusters. We address this gap by introducing an efficient feature engineering strategy that extracts four differential attributes from a partial Difference Distribution Table (pDDT), revealing structural information concealed in raw differential data. Utilising the enriched features, we construct and compare three ML-guided directed graphs for SIMON3232 and SIMECK3232 using K-Nearest Neighbour (KNN), Decision Trees (DT), and Random Forests (RF). To the best of our knowledge, our framework produces the first graph-based visualisation of the differential clustering effect, in which high-probability single-bit differentials form geometrically close clusters in the learned embedding. All three models achieve a precision of 1.01.0 in identifying high-probability differentials, confirming zero false positives. KNN achieves the strongest cluster separation, the highest F1 score and the lowest graph construction time of approximately 2.32.3 seconds, while DT and RF produce optimal paths with near-perfect regression. The results are consistent across both LCAs, demonstrating the applicability of the framework to other AND-rotation LCA families.

Jonathan Cook, Sabih ur Rehman, M. Arif Khan

School of Computing, Mathematics and Engineering, Charles Sturt University, Australia

Graph Representation Learning
AskPDF
Aug 31, 2026

A Simple Transformer Pipeline for Full-Key Side-Channel Attacks on Uncropped Datasets

Deep learning-based side-channel analysis has historically focused on single-byte targets and manually cropped traces, which risks discarding exploitable leakage. While recent work has proposed specialized architectures and resampling techniques to address this gap, the literature lacks a simple transformer baseline for simultaneous full-key attacks on uncropped traces. We present an open-source transformer implementation for uncropped full-key attacks which uses the standard transformer encoder backbone, adapting only the input and output layers to the side-channel setting. We release our implementation, training recipes, and pretrained weights for uncropped ASCADv1f, ASCADv1r, and CHES-CTF-2018 which achieve performance competitive with previously-reported results, while using less than 10GB of VRAM and requiring at most 3.34 hours of training on a single NVIDIA A6000.

Jimmy Gammell, Kaushik Roy

Elmore Family School of Electrical and Computer Engineering Purdue University, West Lafayette, IN 47907, USA

Topic pending
AskPDF
Apr 5, 2026

Improving ML Attacks on LWE with Data Repetition and Stepwise Regression

The Learning with Errors (LWE) problem is a hard math problem in lattice-based cryptography. In the simplest case of binary secrets, it is the subset sum problem, with error. Effective ML attacks on LWE were demonstrated in the case of binary, ternary, and small secrets, succeeding on fairly sparse secrets. The ML attacks recover secrets with up to 3 active bits in the "cruel region" (Nolte et al., 2024) on samples pre-processed with BKZ. We show that using larger training sets and repeated examples enables recovery of denser secrets. Empirically, we observe a power-law relationship between model-based attempts to recover the secrets, dataset size, and repeated examples. We introduce a stepwise regression technique to recover the "cool bits" of the secret.

Alberto Alfarano, Eshika Saxena, Emily Wenger +2

Axiom Math · FAIR at Meta · Duke University

Post-Quantum Cryptography
AskPDF