Affiliations: [a] Computer Engineering College, Jimei University, China
| [b] Institute of Electronic Countermeasure, National University of Defense Technology
| [c] Electromagnetic Protection Evaluation Technology Research Center, Institute of CETC, China
Abstract: The unintentional electromagnetic (EM) emission of computer monitors may cause the leakage of image information displayed on the monitor. Detection of EM information leakage risk is significant for the information security of the monitor. The traditional detection method is to verify EM information leakage by reconstructing an image from EM emission. The detection method based on image reconstruction has limitations: adequate signal sampling rate, accurate synchronization signal, and dependence on operational experience. In this paper, we analyze the principle of image information leakage and propose an innovative detection method based on Convolutional Neural Network (CNN). This method can identify the image information in EM emission to verify the EM information leakage risk of the monitor. It overcomes the limitations of the traditional method with machine learning. This is a new attempt in the field of EM information leakage detection. Experimental results show that it is more adaptable and reliable in complex detection environment.
Keywords: Convolutional neural network, electromagnetic information leakage, image
identification, information security, computer monitor
