A (k, n) Audio Secret Sharing with Share Authentication

Abstract

Audio is one of the multimedia's main component and is the most common way of communication, and many conversations may possess crucial information like credit card numbers, voter card details, etc. Hence the security and confidentiality of this communication must be assured. The secret is usually transmitted in the form of a cipher, which is encrypted with a key, and the key is required for revealing the secret. In this case, the key becomes a single point of failure. For enhancing the security, the secret is sent in multiple units called shares. Here, an approach is proposed for the same. The secret is divided into n units, and a minimum threshold k is set. So secret can only be revealed when k shares are available. The authenticity of the shares can also be verified to detect if the shares have been tampered or not. This scheme can be used for communication channels and secure transmission of audio. No cover is required to transmit the share. Even out of n, if some shares are lost, we may be able to reconstruct the secret without the lost share. Experimentally, Correlation Coefficient and Mean Opinion Score (MOS) for the input and revealed secret is determined. Correlation Coefficient remains intact in this scheme and equal to 1. MOS comes to be between 4.8 and 4.9 on the scale of 5.