ARHCA: Attack Resistant High-Capacity Audio Steganography
International journal of intelligent engineering and systems · 2025
Secret information is hidden within innocent data to achieve covert communication.In this paper, we chose audio as a cover medium to hide secret audio.The cover audio undergoes a short-time Fourier transform (STFT) to obtain coefficients.The secret audio undergoes discrete wavelet transformation (DWT) to compress to half its size.To the compressed secret audio STFT is applied to obtain coefficients.99% of Cover audio coefficients are fused with 1% of secret audio coefficients to form stego coefficients.After fusion, the stego audio is converted to audio format and transmitted across the network.At the receiver the 1% of secret is obtained by subtracting stego with original preshared cover, the process is called audio fission.MSE, PCC, SCC, PESQ and STOI measure the impact of embedding secret into the cover.The secret embedded is compared with the secret received to assess the impact of encoding and decoding.The stego is forced to undergo attacks to understand, it tolerates attacks and keeps the audibility of the secret to a greater extent.ARHCA achieves an embedding capacity of up to 200% and demonstrates a 95.02% reduction in mean squared error, along with improvements of 39.27% in perceptual audio quality and 11.79% in intelligibility under various attack scenarios, compared to previous work.