An audio encryption technique through compressive sensing and Arnold transform

In this paper, an audio encryption scheme using compressive sensing (CS) and Arnold transform-based scrambling is presented. In the proposed method, compressive sensing is done by using a key-based measurement matrix and the scrambling is carried out with the help of a key-based Arnold matrix. The use of these key-based matrices not only provides better security but also do away with their transmission and storage requirement. The measurement matrix is constructed by using the random numbers generated by a linear feedback shift register (LFSR) whose initial state is generated by a piece wise linear chaotic map (PWLCM), using three 32-bit keys whereas the Arnold matrix is constructed by the random numbers generated by using another 32-bit secret key, PWLCM and a logistic map. By combining secure compressive sensing and Arnold scrambling techniques, very high security can be ensured in addition to efficient channel usage, good resistivity to noise, best reconstruction performance, little encoder complexity and excellent scrambling of data. Experimental results prove the effectiveness of the proposed scheme.