EFFICIENT AUTHENTICATION STEGANOGRAPHIC SYSTEMS BASED ON A CLIENT-SERVER MODEL WITH RANDOM-LIKE CODES

In this study, to increase steganographic security, a matrix embedding code was developed as a commonly used public key system by exploiting matrix decomposition. However, in systems such as the public secret cryptography system, users cannot use the public key because of the high complexity required in the networking. To address this problem, a novel method of authentication in steganographic systems, which involves the generation of random-like codes, is presented in this paper. By using a Gaussian elimi-nation technique, random-like codes can be represented as a product of a left submatrix, right submatrix, and systematic parity matrix. The proposed method involves using the aforementioned submatrices and matrix to generate public and private key. In the client, the nonshared selection channel is used to authenticate the user to reduce the risk of cheating. Experimental results confirmed that a reliable authentication performance was achieved when Hamming codes with a systematic parity matrix were used for authenti-cating steganographic systems.