A steganographic method based on optimized audio embedding technique for secure data communication in the internet of things. (13th November 2019)

• Record Type: Journal Article
• Authors: S, Anguraj; S P, Shantharajah; J, Jeba Emilyn
• Is Part Of: Computational intelligence. Volume 36:Number 2(2020)
• Page Start: 557
• Page End: 573
• Publication Date: 2019-11-13
• Subjects: audio steganography -- embedding technique -- internet of things -- optimization -- random bit selection -- security; Artificial intelligence -- Periodicals; Computational linguistics -- Periodicals; 006.3
• DOI: 10.1111/coin.12253
• Languages: English
• ISSNs: 0824-7935

Abstract

Abstract: The rapid growth of the internet and the internet of things (IoT) refers to the next phase of information revolution whose context involves billions of smart devices and sensors interconnected to facilitate speedy information and data exchange under soft real‐time constraints. Digital information revolution has caused significant changes in the data communication. This data communication may require private, secure, and sometimes malicious communication. Competent secrecy can be accomplished by applying novel and inventive audio steganography. This article focuses on the secret message followed by shuffled embedded bit substitution in original audio stream by adopting optimized audio embedding technique (OAET) from the technological observation. To hide the information in the deeper layer of the audio stream, this method uses a new elevated bit range least significant bit (LSB) audio steganography technique that decreases distortion and improves the robustness of the embedded audio stream. The proposed technique proves that the perceptual quality of audio steganography is better than the previous standard LSB technique. Experiment results proved that the cladding of the OAET provides high‐level security to the universal cyber data. The interpretation of results shows that embedding data in audio enhances the level of security when used as IoT smart speakers, where the attackers could not distinguish between the original audio and the embedded audio streams.