Securing audio data using K-shuffle technique

Abstract

This paper presents a k-shuffle based audio scrambling technique which yields cipher audio with varying audibility that is useful in perceptual video encryption algorithms. In this technique the encryption and decryption algorithms group the input audio signals into m-piles of n-values each. The encryption procedure retains the first signal value (of the first pile) and the last signal value (of the last pile) in their respective positions while re-arranging the rest by taking in turns, the 1st values of each of the m-piles, then the 2nd values, up to the nth values. At the receivers end, the decryption algorithm re-orders the scrambled audio signals back in to their respective original positions using the reverse of the k-shuffle technique. It is shown that, the audibility of the cipher audio is controlled by varying the number of piles. The performance of the proposed technique is demonstrated by simulations. The results of the simulations show that the proposed technique offers some level of resistance to chosen/known plaintext attacks.