Evaluation of signal masking by a two-component steganographic system in windowed information processing

This paper examines a two-component steganographic system operating under windowed processing of audio signals. A mathematical model is proposed for selecting the first and second container components during multilayer data embedding within a fixed-size window. The influence of the system’s parameter selection on the quality of message masking and the stegosystem’s resistance to detection is analyzed. Simulation results for the process of hidden embedding in musical fragments are presented, confirming an increase in masking efficiency compared to classical LSB methods. It is shown that the use of a windowed approach and preliminary cryptographic encryption ensures a more uniform distribution of the least significant bits and reduces the probability of detecting hidden information.