ASCAD (ANSSI SCA Database) is a set of databases that aims at providing a benchmarking reference for the SCA community: the purpose is to have something similar to the MNIST database that the Machine Learning community has been using for quite a while now to evaluate classification algorithms performance.

ASCAD (ANSSI SCA Databases) Databases and Neural Networks models, associated to the article "Study of Deep Learning Techniques for Side-Channel Analysis and Introduction to ASCAD Database" available on https://eprint.iacr.org/2018/053.pdf.

ASCAD database version 2. This database contained the power consumption of a STM32 Cortex M4 microcrontroller (STM32F303RCT7) during 800.000 random AES encryptions. The AES encryptions are protected with shuffling and affine masking, and the implementation is available on https://github.com/ANSSI-FR/SecAESSTM32. The raw dataset is split into 8 files of 100.000 encryptions, and the extracted dataset contained the 800.000 preprocessed traces with additional metadata.

Extracted dataset from the original ASCADv2 database (https://www.data.gouv.fr/fr/datasets/ascadv2/).

The dataset is separated in three splits:

- training with 200k traces

- validation with 50k traces

- attack with 50k traces

For each trace, the extracted samples are :

- samples from the masked inputs of the sboxes

- samples from the input mask r_in

- samples from the multiplicative mask r_m

- labels for most intermediates

Comparison of training parameters of ASCAD dataset.

Dataset

This dataset was created by Tom Slooff

Contents

Deep learning, as a high-performance data analysis method, has demonstrated superior efficiency and accuracy in side-channel attacks compared to traditional methods. However, many existing models enhance accuracy by stacking network layers, leading to increased algorithmic and computational complexity, overfitting, low training efficiency, and limited feature extraction capabilities. Moreover, deep learning methods rely on data correlation, and the presence of noise tends to reduce this correlation, increasing the difficulty of attacks. To address these challenges, this paper proposes the application of an InceptionNet-based network structure for side-channel attacks. This network utilizes fewer training parameters. achieves faster convergence and demonstrates improved attack efficiency through parallel processing of input data. Additionally, a LU-Net-based network structure is proposed for denoising side-channel datasets. This network captures the characteristics of input signals through an encoder, reconstructs denoised signals using a decoder, and utilizes LSTM layers and skip connections to preserve the temporal coherence and spatial details of the signals, thereby achi-eving the purpose of denoising. Experimental evaluations were conducted on the ASCAD dataset and the DPA Contest v4 dataset for comparative studies. The results indicate that the deep learning attack model proposed in this paper effectively enhances side-channel attack performance. On the ASCAD dataset, the recovery of keys requires only 30 traces, and on the DPA Contest v4 dataset, only 1 trace is needed for key recovery. Furthermore, the proposed deep learning denoising model significantly reduces the impact of noise on side-channel attack performance, thereby improving efficiency.

This data set contains disambiguated publication data from zbMATH (www.zbmath.org) for use in author name disambiguation (AND). It covers 28321 publications with 33810 authorship records, authored by 2946 distinct authors. Authorship records have been manually annotated with author identifiers.

This download includes additional data sets for advanced, selective disambiguation.

For details, see "Mark-Christoph Müller, Florian Reitz, and Nicolas Roy (2017): Data Sets for Author Name Disambiguation: An Empirical Analysis and a New Resource", Scientometrics, doi:10.1007/s11192-017-2363-5.

ASCAD database (raw traces and extracted traces) for ATMega 8515 power consumption of boolean masked AES using a variable key.

Comparison of training parameters of AES_RD dataset.

Reducuced version of the ASCADr dataset from "Agence Nationale de sécurité des systèmes d'information" (ANSSI)

https://www.data.gouv.fr/fr/datasets/ascad/

Training set : 50k raw traces from the random key split.

Test set : 10k raw traces from the random key split

Attack set : 10k raw traces from the fixed key split

No preprocessing of any sort has been done, this is just to change the .h5 size and organisation.

ASCAD database (ruwe sporen en geëxtraheerde sporen) voor ATMega 8515 stroomverbruik van boolean gemaskerde AES met behulp van een variabele sleutel.

ASCAD (ANSSI SCA Databases) Databaser och Neural Networks modeller, associerade till artikeln ”Study of Deep Learning Techniques for Side-Channel Analysis and Introduction to ASCAD Database” finns på https://eprint.iacr.org/2018/053.pdf.

ASCAD-database (råspor og ekstraherede spor) for ATMega 8515 strømforbrug af boolesk maskerede AES ved hjælp af en variabel nøgle.

Detailed information about the Organisation SCAD.

Comparison of training parameters of DPA contest v4 dataset.

🇫🇷 프랑스

Forecast: Whole Fresh Bigeye Scad Production in Capture Fisheries in France 2024 - 2028 Discover more data with ReportLinker!

Forecast: Total Mackerel Scad Production in Capture Fisheries in France 2024 - 2028 Discover more data with ReportLinker!

Comparison of denoising performance between LU-Net and DAE models.