A Novel Digital Audio Encryption and Forensics Watermarking Scheme

Abstract

There are a large amount of audio signals stored in third-party storage centers. To improve the privacy and security of audio signals, a digital audio encryption and forensics watermarking scheme is proposed. We defined the signal energy ratio feature of audio signals and designed the embedding method by quantifying the feature, aiming to improve the security of watermarking system. We firstly encrypted the original audio by using scrambling and multiplication, to get the encrypted data. Secondly, we divided the encrypted data into frames and compressed each frame by sampling to get the compressed data. After that, we embedded the compressed data and frame number into the encrypted data to get the watermarked signal being uploaded to third-party storage centers. Authorized users download encrypted data and verify the authenticity of the data. For the intact data, they decrypt it directly to get the audio signal. If the downloaded data has been attacked, they locate the attacked frame and extract the compressed data to reconstruct the attacked signal approximately. Furthermore, we decrypted the reconstructed signal to obtain the expression meaning of the original audio. The experimental results demonstrate the effectiveness of the proposed scheme.

Keywords

• Digital watermark
• encrypted audio
• content authentication
• tamper recovery