Compressing And Ciphering Digital Signal By Using Wavelet Transform

Abstract

With the rapid development of multimedia and network technologies, the security of multimedia becomes more and more important, since multimedia data are transmitted over open networks more and more frequently. Typically, reliable security in storage and transmission of digital speech data, images, and videos is needed in many real applications, such as medical imaging
systems, military image databases, as well as confidential video conferences. In recent years, some consumer electronic devices, such as mobile phones, have also started to provide the function of saving and exchanging digital speech/music data, images, and video clips under the support of multimedia messaging services over wireless networks, which is urgently demanding for
multimedia security.
A new proposed method is presented with high degree of security to calculate the transform domain of Discrete Wavelet Transform (DWT) and Inverse Discrete Wavelet Transform (IDWT) by the pyramid algorithm. The pyramid
algorithm is permitted to construct the bases of the vectors of its transform, this leads to more security in the system. Also by the pyramid algorithm, the
transformation process gives double the number of samples of the original signal, while in any transformation the process gives square the number of samples of the original signal. This work includes a mathematical presentation
the modification of DWT (with IDWT) followed by a second stage of DWT (multi resolution) to compress the signal. Thus, the system ciphers the signal by two dimensions (time and frequency) then compresses and ciphers it again. The
keys used throughout this paper are generated manually; the first algorithm is for time domain, the second algorithm is for frequency domain and the last
algorithm is for compressing signal samples. Finally, all the preceding methods give very inapprehensible compressed and
ciphered signal and intelligible decompressed and deciphered signal according to numerical and graphical results, so this work gives 50% compression ratio and approaches 10% error ratio.