A NOVEL METHOD FOR GENERATING DIGITAL IMAGE OF ENCRYPTED SPEECH SIGNAL BASED ON FWHT AND WPLCM

This paper suggests a novel algorithm for encrypting speech signals in common image formats and retrieve them from these image files. The speech signal is encrypted in three levels. In the first level,the sample positions are permuted based on keys generated using Game of Life matrix and Piecewise linear Chaotic Map (PWLCM) in order to reduce the correlation between adjust samples. In the second level, the resulting samples are then converted to Fast Walsh Hadamard Transform (FWHT) and their transactions are encrypted by using circular transformations in the row and column depending on the generated key. At the third level, the values of encrypted samples are converted to color pixels, which are then arranged in a puzzled manner and put in a 2-D matrix to achieve the secured data transfer across networks, with the image file contains the encrypted speech signal. Several objective measures have been used to evaluate the performance of the suggested method. The experimental results and numerical analyses show that the algorithm gives a high degree of security and robust against brute force attackers, statistical attack, strong diffusion and ambiguity so that the encrypted message has been saved in a different format from the original signal, and finally give the good quality of the reconstructed speech signal from image files.