Audio Steganography Using Multi-Level DWT to Improve PSNR, Hiding Capacity and Imperceptibility

Abstract

A world without digital communication and transmission of data is hard to imagine. Need for secure data transmission over the Internet is growing every day and security has become very crucial for various reasons. To ensure this security, Steganography is being implemented. This paper provides a novel method for audio steganography where text is encrypted and is hidden by implementing two-level discrete wavelet transform (DWT) in audio. By implementing discrete wavelet transform, approximate and detail coefficients are obtained. Data is embedded in these coefficients of the audio. This method gives very good hiding capacity and a high value of peak signal-to-noise ratio (PSNR). The file that is used for hiding information is called cover file and the file that has the secret message embedded in it is called stego file. The stego audio file is imperceptible from the cover audio file. Mean square error (MSE) is very low and reconstruction is quite accurate.