发表机构
Elmore Family School of Electrical and Computer Engineering; Purdue University（埃尔莫尔电气与计算机工程学院; 普渡大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究提出一种适配侧信道场景的简单Transformer流水线，用于未裁剪数据集的全密钥侧信道攻击，发布了相关实现与权重，性能相当且资源消耗较低。
AI 中文摘要
基于深度学习的侧信道分析以往聚焦于单字节目标和手动裁剪的迹线，这存在丢弃可利用泄漏的风险。尽管近期研究提出了专门架构和重采样技术以解决该问题，但文献中缺乏针对未裁剪迹线的同时全密钥攻击的简单Transformer基线。我们提出了一种用于未裁剪全密钥攻击的开源Transformer实现，采用标准Transformer编码器骨干，仅将输入和输出层适配到侧信道场景。我们发布了针对未裁剪ASCADv1f、ASCADv1r和CHES-CTF-2018的实现、训练方案和预训练权重，其性能与先前报告的结果相当，同时使用不到10GB的显存，在单个NVIDIA A6000上训练最多需要3.34小时。
英文摘要
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.
CommentsAccepted to the OPTIMIST Workshop '26 at CHES 2026. 6 pages, 1 figure. Code can be found at https://github.com/jimgammell/simple-transformer-pipeline-for-sca
