Enhanced Audio Steganography: Audio-in-Audio Steganography with Multi-Bin Phase Coding and Adaptive Post-Processing

This paper presents an enhanced audio steganography technique for hiding audio within audio files using phase coding. We introduce a Multi-Bin Phase Coding (MBPC) approach that extends traditional phase coding by utilizing multiple frequency bins within each audio segment, significantly increasing embedding capacity. A header structure is embedded to preserve the secret audio parameters, enabling accurate reconstruction during extraction. Additionally, we implement various post-processing techniques to enhance the quality of the extracted audio, including Wiener filtering, spectral gating, and median filtering. Experimental results demonstrate the effectiveness of our approach, achieving a Signal-to-Noise Ratio (SNR) of approximately 2.80 dB between host and stego audio when using optimized frequency bin selection. Our findings indicate that while the Multi-Bin approach substantially increases capacity compared to traditional phase coding, the optimized bin selection strategy significantly improves the quality-capacity trade-off by minimizing audible distortion. The proposed technique enables robust and efficient audio-in-audio steganography with practical applications in covert communication, copyright protection, and secure data transmission.