Nonlinear Filter Soft Tempest Techniques Investigation

The eavesdropper could capture the video signal by antennas from a distance and recover the plaintext which is displayed on the computer. By this way eavesdroppers can get the encrypted message they want more timely,accurately,extensively and continuously. It will be extremely easy to reveal the content of an encrypted message by this way. It is proved that the leakage of electromagnetism of computer video signals is mainly transmitted to the space by high-frequency components. At present, with the increasing of the display units processing speed, the video frequency and it's harmonic have been in the frequency range of 100MHz～3GHz. According to the theory of antenna transmission, all of the display units at this frequency range can be transmitted effectively and therefore become source of leakage.Soft Tempest technique is a new computer video information security protective technique which is developed in recent years. It tries to run a programme on computers to process the video signals so as to attain the goal of information protection. As the high-frequency component of the video signals is concentrated on the edge and the contour of the picture as well as the normal direction of some veins. Using normal technique in this part may make the picture quality poorer. How to ensure the picture quality while realizing computer video information tempest protection is the core question of this paper.By analyzing the leaky mechanism and the HVS characteristics,this paper discusses this question from two angles. The first method is using direction weighted nonlinear filter to process computer video signals. This method is mainly based on the fact that character has obvious directional feature. We process the computer video signals by dividing the characters' directions and using fuzzy control to adjust the weighed size of filter windows. In this way, we can reduce the higher frequency video signals while keeping the image quality, which increases the difficulty from the side of receiver.Based on the pictures' feature, the second method is that we use fake transmittion technique to process the displayed signals by utilizing HVS (Human Visual System) characteristics.By intentionally inserting a picture,we do the dithering modulation on an original picture which has been nonlinear filtered,so that the video signals received by the eavesdropper is totally different from that displayed on the computer. In this way, we can realize the soft tempest security protection of video information.We use VC++ to process the image and to simulate the above two ways of hiding video information. The result of experimental test on reverting the video signal has proved the two methods are effective and feasible.