arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.16309cs.CRcs.AI
收藏

比特的重量:嵌入式深度学习模型的EMFI敏感性分析

The Weight of a Bit: EMFI Sensitivity Analysis of Embedded Deep Learning Models

TTControl GmbH(TTControl公司);Faculty of Informatics and Information Technologies, Slovak University of Technology(斯洛伐克理工大学信息学与信息技术学院);State Key Laboratory of Blockchain and Data Security, Zhejiang University(浙江大学区块链与数据安全国家重点实验室)

Jakub Breier, Štefan Kučerák, Xiaolu Hou

AI总结:

本文研究四种数表示对EMFI攻击成功的影响,发现整数表示比浮点表示更抗攻击,尤其8位整数在较大网络中保持较高精度。

摘要:

针对嵌入式神经网络模型的电磁故障注入攻击已被证明是严重威胁。现有研究多从不同角度探讨模型的鲁棒性,但尚无综合研究评估模型参数的数表示对EMFI攻击的影响。本文研究四种数表示(32位和16位浮点,8位和4位整数)对EMFI攻击成功的影响。我们选择两种常见浮点表示(32位和16位)和两种整数表示(8位和4位),在嵌入式内存芯片上部署四个常见图像分类器(ResNet-18、ResNet-34、ResNet-50和VGG-11),并利用低成本EMFI平台触发故障。除了准确性评估外,我们通过分析位错误率、损坏字节的空间分布以及不同格式中0xFE/0xFF字节值的普遍性,识别导致鲁棒性差异的机制。实验结果表明,浮点表示在单次故障注入后几乎完全降级,而整数表示整体更具抗性。特别是,8位整数在相对较大的网络(VGG-11)中保留约70%的Top-1精度和约90%的Top-5精度。

英文摘要:

Fault injection attacks on embedded neural network models have been shown as a potent threat. Numerous works studied resilience of models from various points of view. As of now, there is no comprehensive study that would evaluate the influence of number representations used for model parameters against electromagnetic fault injection (EMFI) attacks. In this paper, we investigate how four different number representations influence the success of an EMFI attack on embedded neural network models. We chose two common floating-point representations (32-bit, and 16-bit), and two integer representations (8-bit, and 4-bit). We deployed four common image classifiers, ResNet-18, ResNet-34, ResNet-50, and VGG-11, on an embedded memory chip, and utilized a low-cost EMFI platform to trigger faults. Beyond accuracy evaluation, we characterize the injected fault pattern by analyzing the bit error rate, the spatial distribution of corrupted bytes, and the prevalence of 0xFE/0xFF byte values across formats, identifying the mechanisms responsible for the observed differences in resilience. Our results show that while floating-point representations exhibit almost a complete degradation in accuracy (Top-1 and Top-5) after a single fault injection, integer representations offer better resistance overall. In particular, the 8-bit representation on a relatively large network (VGG-11) retains Top-1 accuracy of around 70% and Top-5 at around 90%.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑