Comparative Analysis of Steganographic Methods for Covert Embedding Information into Speech Signals
M. Poleshchuk Оlga, Yaqin Zhao, Longwen Wu, Liang Zhao, Chong Wu, Dongming Luo · 2025
Modern data transmission systems are not only standard information systems but also involve encryption. It is important to note that the latter can be implemented both for the purpose of protecting confidentiality as well as for hiding information and even for programming and psychological influence. As a result, today steganography is studied both in specialized educational institutions and by interested parties. Moreover, this topic is widely popularized in programming since innovative technologies open new possibilities for data hiding. It is especially relevant for working with text and speech signals, which are transmitted by audio and video files and allow attracting the attention of ordinary people/consumers. Results indicate that subband methods such as filter banks and wavelet transform demonstrate good resilience to various attacks. The comparative analysis reveals that while the Critical Band Embedding method scores highest in reliability, the filter bank and wavelet transformation methods also offer substantial robustness and are simpler to apply. This highlights the balance between security and ease of implementation in different steganographic methods. While several challenges remain in the fields of steganography and steganalysis, these are expected to be addressed through emerging technologies.