{"title":"A Method for Extracting Plausible Images From EM Leakage Measured at Low Sampling Rates","authors":"Taiki Kitazawa;Hiroyuki Kubo;Yuichi Hayashi","doi":"10.1109/LEMCPA.2023.3301431","DOIUrl":null,"url":null,"abstract":"Threats of compromising emanations, which allow malicious eavesdroppers to screen information through electromagnetic (EM) waves that have leaked from video display units (VDUs), are expanding their target devices with the development of low-cost and portable measurement setups. On the other hand, such low-cost instruments have a lower sampling rate than the pixel clock in VDUs; pixel information is lost in the image reconstructed from the measured EM waves, and the quality of the image is also degraded. In this letter, we propose a novel method to reconstruct high-quality images from lower sampling rate observations by a nonnegative linear optimization. Our method can recover clear images even when measuring EM leakage at low sampling rates and fluctuating sampling clocks. To obtain a stable and plausible solution to the optimization problem, we focus on fluctuations in sample points caused by clock jitter in devices and instruments and use pixel information obtained by observing multiple frames. To verify the proposed method, we generated simulated data from leaked EM waves, including screen information, and evaluated the visibility of the reconstructed images. Consequently, the images reconstructed using the proposed method were of the same quality as those reconstructed by applying conventional methods to data measured at a high sampling rate sufficient for information recovery. Moreover, we confirmed that even if the image quality is degraded by a decrease in the signal-to-noise ratio of the observed signal and an increase in the clock jitter, the proposed method can still be used to obtain high-quality reconstructed images by increasing the number of observed frames.","PeriodicalId":100625,"journal":{"name":"IEEE Letters on Electromagnetic Compatibility Practice and Applications","volume":"5 3","pages":"87-91"},"PeriodicalIF":0.9000,"publicationDate":"2023-08-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/iel7/8566057/10260712/10201909.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Letters on Electromagnetic Compatibility Practice and Applications","FirstCategoryId":"1085","ListUrlMain":"https://ieeexplore.ieee.org/document/10201909/","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0
Abstract
Threats of compromising emanations, which allow malicious eavesdroppers to screen information through electromagnetic (EM) waves that have leaked from video display units (VDUs), are expanding their target devices with the development of low-cost and portable measurement setups. On the other hand, such low-cost instruments have a lower sampling rate than the pixel clock in VDUs; pixel information is lost in the image reconstructed from the measured EM waves, and the quality of the image is also degraded. In this letter, we propose a novel method to reconstruct high-quality images from lower sampling rate observations by a nonnegative linear optimization. Our method can recover clear images even when measuring EM leakage at low sampling rates and fluctuating sampling clocks. To obtain a stable and plausible solution to the optimization problem, we focus on fluctuations in sample points caused by clock jitter in devices and instruments and use pixel information obtained by observing multiple frames. To verify the proposed method, we generated simulated data from leaked EM waves, including screen information, and evaluated the visibility of the reconstructed images. Consequently, the images reconstructed using the proposed method were of the same quality as those reconstructed by applying conventional methods to data measured at a high sampling rate sufficient for information recovery. Moreover, we confirmed that even if the image quality is degraded by a decrease in the signal-to-noise ratio of the observed signal and an increase in the clock jitter, the proposed method can still be used to obtain high-quality reconstructed images by increasing the number of observed frames.