
This dataset is part of the SETS-THIRAL dataset repository, which comprises multiple datasets suitable for performing AI-assisted Side-Channel Analysis (SCA). It consists of Power consumption traces obtained from an SRL Protected PRESENT-80 cryptographic implementation executed on ATxmega microcontroller platform.
The traces were obtained by measuring the real-time power consumption of an PRESENT-80 encryption process. These measurements are highly suitable for Side-Channel Analysis (SCA), enabling cryptographic key recovery through both statistical techniques and deep learning–based approaches. Power consumption data were collected using a ChipWhisperer CW1200 platform connected to an ATxmega microcontroller. Trigger-based synchronization is used during acquisition process to align the captured traces with encryption operations. After collection, the traces were formatted and annotated with appropriate labels to support side-channel analysis experiments and AI model development. The dataset is provided in HDF5 (.h5) format and is divided into Profiling_traces and Attack_traces groups. The Profiling_traces group consists of 150,000 power traces, each containing 95,000 sample points along with a metadata structured array containing the corresponding plaintext, key, and ciphertext values and a label array. The Attack_traces group contains 20,000 power traces together with the corresponding metadata structured array. Labels are generated from the first-round S-box output of the PRESENT encryption algorithm. For each trace, plaintext byte 0 (PT[0]) is XORed with the effective key byte. The most significant nibble (MSB) of the resulting 8-bit value is then extracted and used as the input to the PRESENT S-box. The resulting 4-bit S-box output is assigned as the label for the corresponding trace. Since the PRESENT S-box produces a 4-bit output, the labels consist of 16 possible classes ranging from 0 to 15. The Shift Register LUT (SRL) countermeasure is a hiding-based software technique designed for 8-bit microcontroller implementations. It is implemented using an array initialized with random values, on which circular shift operations are performed to introduce additional computational activity that is independent of the sensitive data being processed. In the SETS-THIRAL dataset, the SRL is inserted at randomly selected positions within the first round of the cryptographic computation. For each encryption, one of several predefined execution sequences is chosen randomly, causing the SRL to execute at different stages of the first round. This random insertion results in trace misalignment and temporal desynchronization across encryptions, thereby enhancing the implementation's resistance to side-channel attacks.
This Dataset Is Generated For Research And Educational Purposes In Side-channel Analysis (Sca). It Provides Labelled Power Traces Captured From Cryptographic Computations And Can Be Used To Develop, Evaluate And Benchmark Classical And Ai-assisted Attack Methodologies. The Dataset Also Facilitates The Study Of Leakage Characteristics, Feature Extraction Techniques, Model Interpretability And The Evaluation Of Cryptographic Countermeasures.
Attribution-Non-Commercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)
To preview this file, you need to be a registered user. Please complete the registration process to gain access and continue viewing the content.
© 2026 - Copyright AIKosh. All rights reserved.