A Randomized Approach for Data Embedding in Audio Steganography

Exchange of information over the internet or an intranet is inevitable in today’s technological atmosphere. Information hiding using steganographic techniques like LSB Steganography do provide us with a decent defense against hackers, however, recent advancements in computing, internet usage have paved the way for hackers to breach data hiding measures like steganography to a considerable extent. This paper presents a novel approach with a randomized embedding scheme for audio steganography, which absolve the lack of security, and the localized nature of the distortions produced in the stego object. The main advantage of a randomized embedding scheme is that, the data values can be stored at any random position in the audio data set. This protects the stego object against an unauthorized breach by unwanted entities. Although the random number generation increases the overall computational complexity of the entire process, the resultant improvement in security and dataset integrity is certainly worth it. Experimental results indicate that the proposed method has an improved security and dataset integrity check with genuine protection for the hidden information, when compared to a basic embedding scheme like the sequential embedding scheme as referred to in the earlier paper [16]. The scope of earlier work confirms the practical feasibility of implementation of a randomized embedding scheme, which could challenge attackers with impregnable protections for data hiding.