Reversible data hiding in audio based on discrete cosine transform and location maps

Abstract

Transmitting information via public networks is prone to illegal attacks. Thus, data hiding is a suitable aspect that is useful to assure information security. Nowadays, multimedia tempering becomes the most variety problem that occurs on public networks. Audio steganography methods aim to hide secret data in an audio file called cover. Many methods in the time domain can hide large data (payload) with relatively low robustness. On the other hand, transformed based frequency algorithms can be more robust but the recovery of the cover audio is not guaranteed. In this paper, we investigate the time and frequency domains and tend to bridge gaps in the area of data hiding with the frequency transformation methods. In addition, we propose a new reversible audio data hiding based on Discrete Cosine Transform (DCT) and location maps. This kind of map helps to fully recover both the secret message and the cover file from the stego where the frequency sample values are increased or decreased by 10 in correspondence with the intended bit to be embedded. The experimental results prove that this method is able to hide high payload, and it is also robust since it guarantees 100% the reversibility of both the secret message and the cover audio file. The quality measured by Signal-to-noise ratio (SNR) is approximately 75 dB, Normalized Correlation (NC) value is 1.000, and Bit Error Rate (BER) is around 0.006. Additionally, this method is able to maintain the quality and invisibility, which may not be achieved by some previous research.