-
摘要:
电子信息设备工作时无意发射的电磁波中包含有用信息，会导致电磁信息泄漏，从而威胁设备的信息安全。现有的电磁信息泄漏检测方法，在复杂现场环境下，难以从具有不确定性的电磁泄漏信号中提取有用信息。面向电磁信息安全问题，开展了电磁信息泄漏检测研究，提出了一种基于深度学习的检测方法。设计了一个适用于电磁泄漏信号的一维卷积神经网络，并结合改进的梯度加权类激活映射方法，在未知电磁信息泄漏特征的前提下，通过深度学习实现电磁信息泄漏特征的智能标定和自动提取，从而解决了现场环境下电磁信息泄漏检测难以提取有用信息的问题。分别通过实测和仿真对比实验，验证了所提方法的有效性。
Abstract:Electronic information equipment will emit electromagnetic wave unintentionally, which contains useful information. It will lead to the electromagnetic information leakage, thus threatening the information security. The traditional electromagnetic information leakage detection methods are difficult to extract useful information from uncertain electromagnetic leakage signals in complex environments. Aimed at the problem of electromagnetic information security, the electromagnetic information leakage detection is studied. A detection method based on deep learning is proposed. The method designs a one-dimensional convolutional neural network suitable for electromagnetic signals, and combines an improved gradient-weighted class activation mapping algorithm. It can locate and extract the electromagnetic leakage information characteristics intelligently under the condition of unknown the characteristics through deep learning so as to solve the problem of extracting electromagnetic leakage information in complex environments. The effectiveness of the proposed method is verified by experiments and simulation.
-
图 1 基于深度学习的电磁信息泄漏检测原理
Figure 1. Electromagnetic information leakage detection principle based on deep learning
图 2 EM-CNN处理电磁泄漏信号示意图
Figure 2. Schematic diagram of electromagnetic leakage signal processing by EM-CNN
图 3 计算机显示器电磁信息泄漏标定的实测验证
Figure 3. Experimental verification for electromagnetic information leakage detection on computer displayer
图 6 不同CNN的检测准确率比较
Figure 6. Comparison of detection accuracy rates using different CNNs
表 1 EM-CNN结构参数
Table 1. Parameters of EM-CNN structure
神经网络层次 卷积核/滤波器尺寸 计算步长 通道数 卷积层1 4 2 32 池化层1 2 2 32 卷积层2 12 2 32 池化层2 2 2 32 卷积层3 12 2 32 池化层3 2 2 32 卷积层4 12 2 64 池化层4 2 2 64 
下载: 导出CSV
-
[1] 刘泰康, 李咏梅, 等. 电磁信息泄漏及防护技术[M]. 北京: 国防工业出版社, 2015: 11-125.LIU T K, LI Y M, et al. Electromagnetic information leakage and countermeasure technique[M]. Beijing: National Defense Industry Press, 2015: 11-125(in Chinese). [2] VAN ECK W. Electromagnetic radiation from video display units: An eavesdropping risk [J]. Computers and Security, 1985, 4(4): 269-286. doi: 10.1016/0167-4048(85)90046-X [3] KUHN M G. Eavesdropping attacks on computer displays[C]//Proceedings of Information Security Summit, 2006: 1-10. [4] KUHN M G. Optical time-domain eavesdropping risks of CRT displays[C]//Proceedings of IEEE Symposium on Security & Privacy. Piscataway: IEEE Press, 2002: 3-18. [5] KUHN M G. Security limits for compromising emanations[C]//International Workshop on Cryptographic Hardware and Embedded Systems. Berlin: Springer, 2005, 3659: 265-279. [6] KUHN M G. Compromising emanations of LCD TV sets[J]. IEEE Transactions on Electromagnetic Compatibility, 2013, 55(3): 564-570. doi: 10.1109/TEMC.2013.2252353 [7] KUHN M G. Electromagnetic eavesdropping risks of flat-panel displays: Lecture notes in computer science[C]//Proceedings of the 4th International Conference on Privacy Enhancing Technologies. Berlin: Springer, 2005, 3424: 88-107. [8] SEKIGUCHI H. Information leakage of input operation on touch screen monitors caused by electromagnetic noise[C]//Proceedings of IEEE International Symposium on Electromagnetic Compatibility. Piscataway: IEEE Press, 2010: 127-131. [9] SHI J, YONGACOGLU A, SUN D, et al. A novel wavelet based independent component analysis method for pre-processing computer video leakage signal[C]//Proceedings of IEEE Symposium on Computers and Communications. Piscataway: IEEE Press, 2016: 334-339. [10] SUN D, SHI J, WEI D, et al. A low-cost and efficient method of determining the best frequency band for video leaking signal reconstruction[C]//Proceedings of International Conference on Systems and Informatics. Piscataway: IEEE Press, 2015: 624-628. [11] TOSAKA T, YAMANAKA Y, FUKUNAGA K. Method for determining whether or not information is contained in electromagnetic disturbance radiated from a PC display[J]. IEEE Transactions on Electromagnetic Compatibility, 2011, 53(2): 318-324. doi: 10.1109/TEMC.2010.2103562 [12] MAO J, LIU P, LIU J, et al. Method for detecting electromagnetic information leakage from computer monitor[J]. Control and Intelligent Systems, 2017, 45(1): 37-42. http://smartsearch.nstl.gov.cn/paper_detail.html?id=1ac05d2eff92b309a45c8445b1efc977 [13] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[C]//Annual Conference on Neural Information Processing Systems, 2012: 1097-1105. [14] CHO S I, KANG S. Gradient prior-aided CNN denoiser with separable convolution-based optimization of feature dimension[J]. IEEE Transactions on Multimedia, 2019, 21(2): 484-493. doi: 10.1109/TMM.2018.2859791 [15] DENG Z, SUN H, ZHOU S, et al. Multi-scale object detection in remote sensing imagery with convolutional neural networks[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2018, 145: 3-22. doi: 10.1016/j.isprsjprs.2018.04.003 [16] CHEN S, XU L, MA L, et al. Convolutional neural network for classification of solar radio spectrum[C]//Proceedings of IEEE International Conference on Multimedia and Expo Workshops. Piscataway: IEEE Press, 2017: 198-201. [17] ZHANG X, LIN T, XU J, et al. DeepSpectra: An end-to-end deep learning approach for quantitative spectral analysis[J]. Analytica Chimica Acta, 2019, 1058: 48-57. doi: 10.1016/j.aca.2019.01.002 [18] SELVARAJU R R, COGSWELL M, DAS A, et al. Grad-CAM: Visual explanations from deep networks via gradient-based localization[C]//Proceedings of the IEEE International Conference on Computer Vision. Piscataway: IEEE Press, 2017: 618-626. [19] CAO J, PANG Y, LI X, et al. Randomly translational activation inspired by the input distributions of ReLU[J]. Neuro Computing, 2018, 275: 859-868. http://www.onacademic.com/detail/journal_1000040110605810_a66d.html [20] SRIVASTAVA N, HINTON G, KRIZHEVSKY A, et al. Dropout: A simple way to prevent neural networks from overfitting[J]. Journal of Machine Learning Research, 2014, 15(1): 1929-1958. http://jmlr.org/papers/volume15/srivastava14a.old/srivastava14a.pdf [21] VESA. VESA and industry standards and guidelines for computer display monitor timing (DMT)[S]. San Jose: VESA, 2013. -
下载:
点击查看大图
计量
- 文章访问数: 827
- HTML全文浏览量: 142
- PDF下载量: 163
- 被引次数: 0
