Combined Use of Equivalent and Non-Equivalent Transformations of FPGA Program Code to Embedding Additional Security Data

Abstract

The paper is devoted to a stenographic approach to storing data in the space of the FPGA chips program code. In this approach, data is covertly embedded in the program code and forms a single whole with it. As a result, the presence of embedded data is hidden from the outside observer. The data hidden in this way can only be retrieved with a stego key. Traditionally for stego embedding of additional data into FPGA program code equivalent transformations of this program code are used. To do this, the program codes of the FPGA LUT units are transformed so that the FPGA functionality does not change. In this case, additional data is embedded in the target bits of the program code. This paper proposes a method that combines traditional equivalent embedding with non-equivalent embedding. The method uses an additional stego resource - inessential LUT units. This is a subset of LUT units that perform arithmetic calculations for approximate data. Inessential LUT units participate only in the calculation of the bits to be discarded as a result of rounding. Therefore, distortion for the program code of these units does not affect the correct operation of the device. The proposed method allows to increase the volume of data steganographically embedded in the FPGA program code. This increase is achieved through the use of an additional nonequivalent stego resource. An experiment was carried out in the paper, which confirmed the effectiveness of the proposed method. The experiment showed an increase in the volume of embedded data compared to using only an equivalent stego resource.