Optimization of Error Pattern Embedding Steganography within Error-Correcting Code Frameworks

The importance of secure wireless communication is increasing as adversaries’ eavesdropping capabilities become more advanced. In this paper, we propose a novel steganography method that utilizes error pattern embedding to minimize the likelihood of detection by eavesdroppers. Unlike existing error pattern embedding steganography, we introduce a secret codebook generation algorithm designed to maximize the secret codebook size. Our algorithm is applicable to any coding scheme that possesses a predefined maximum number of correctable errors. In addition, we propose a novel steganalysis scheme for error pattern embedding steganography. Our method is based on comparing two distinct empirical relative entropies: one derived from the empirical probability mass function (pmf) of observed transmitted signals and the other from the empirical pmf of randomly generated signals following a Bernoulli $\left( {\frac{1}{2}} \right)$ distribution. Simulation results indicate that our algorithm enhances security by effectively reducing the detection probability by the eavesdropper while simultaneously increasing the capacity for secret information.