EMISSION SECURITY LIMITS FOR COMPROMISING EMANATION AND ITS RECONSTRUCTION

Abstract

In this dissertation, reconstruction of electromagnetic emanation security (EMSEC)-channel information for video display units and printer are reconstructed using the averaging technique and proposed adaptive deringing filter. Also, emission security limits are proposed based on the analysis of the indoor EMSEC-channel. An emitted waveform from equipment which manages the important information can be detected and restored intentionally using the sensitive antenna and high performance receiver. These documents related to the EMSEC have classified by high confidentiality so that these are prohibited to publish by military organization. For this reason, reasonable emission security limits for various electronic devices dealing with significant information are necessary. Firstly, we try to identify the exact a signal characteristics and the frequency components to measure and analyze the spectrum of electromagnetic waves which are contained information on personal computer (PC) and printer. The target devices are the desktop, laptop and laser printer which is generally used in the domestic offices in this study. The printer processed a large amount of information for a short period of time, there may be leaked the information in this process. To verify the leakage of electromagnetic spectrum that contains information, we measure and analyze the whole spectrum from 100 MHz to 1000 MHz. Secondly, we represent how to build the EMSEC-system and to restore the signal leakage of electromagnetic waves on the basis of the signal characteristics of the electromagnetic wave leakage of printer and video display unit (VDU) of PC. The parameters that can improve the performance of signal recovery of the leakage electromagnetic wave, it can be given antenna sensitivity, resolution bandwidth (RBW) of the receiver, and signal processing gain. To adjust the signal processing gain, antenna which have the high antenna gain, and the use of wider RBW on receiver are improved hardware of EMSEC system. Whereas image restoration algorithm for EMSEC system as post-processing is a portion corresponding to the software of EMSEC system. Techniques for increasing signal strength and noise reduction are particularly important when trying to measure compromising emanations because the magnitude of these signals can be extremely small. Averaging technique find to achieve maximum cross correlation between recorded electromagnetic leaked signals. That method is a practical, highly effective and widely used technique for increasing the signal-to-noise ratio (SNR) of a periodic signal, such as that generated by the image-refresh circuitry in a video display system. But, the printer and facsimile exhibit aperiodicity in their EMSEC-channel information during their operation state unlike video display systems. Since the aperiodic EMSEC-channel information of equipments such as printers and faxes is not involved in processing gain, the differences between periodic- and aperiodic compromising emanations need to be considered in order to establish emission security limits. In addition to, we propose the adaptive deringing filter to reconstruct the EMSEC- channel information from PC and printer. We can obtain that the minimum peak signal-to-noise ratio (PSNR) enhancement is 2 and maximum PSNR enhancement is 10 compared with the original reconstructed image. Next, we perform the EMSEC-channel measurements in the 100?1000 MHz frequency bands. Second, we analyze the pathloss characteristics of the indoor EMSEC-channel based on these measurements. We find the frequency correlation pathloss characteristics of compromising emanations to determine the reasonable total radio attenuation (TRA). Also, the pathloss exponent value have a range from 1.06 to 2.94 depending on frequency band and the CMs, which in turn differed with propagation environments. Through this EMSEC-channel analysis, we affirm that the TRA, which is one of the key parameters for determining the security limits for compromising emanations, follows the Rician distribution. However, previous work assumed that radio attenuations would have constant values. We found that the TRA does not show significant differences depending on the frequency bands and has the following range depending on the environment, 29?41dB at CM2, a 42?57 dB at CM3, a 47?57 dB at CM4, and 24?29 at CM5. In addition to, CM3 and CM4 have greater TRA than CM2 and CM5. Based on the experimental results of this study, we propose security limits on periodic as well as aperiodic EMSEC-channel information. The proposed security limits on compromising emanations are classified into two levels according to the TRA and the level of required confidentiality. Periodic emission security limits for class A is 24, 28, 35 dBμV/m in the 100-400 MHz, 400-900 MHz and 900-1000 MHz, respectively. And periodic emission security limits for class B is 4, 1, 3, 5 dBμV/m in the 100-200 MHz, 200-600 MHz, 600-700 MHz and 700-1000 MHz, respectively. Aperiodic emission security limits are weaker than the processing gain Gp, 23 dBi than periodic emission security limits owing to the redundancy caused by repetitive signals. So, that the periodic EMSEC-channel information is easily leaked and reconstructed, which results in a potential risk. Thus, the periodic emission security limits must be stronger than the aperiodic emission security limits. We can then compare our security limits with other security limits and existing civil and military EMC standards. Future works may include characterization and reconstruction of FAX, smartcard and other electronics. And it is need to EMSEC-channel analysis in more complex environments.