Masking properties of APD communication systems

In this paper, we explore the ability of wavelet multilevel analysis to unmask the information hidden in a chaotic mask. This digital filtering technique has been recently reported as particularly well suited for the detection of coherent structures. In a recent paper the capability of wavelets to unmask a message, in cases where a chaotic signal is just added to the information, has also been demonstrated. The case of an active passive decomposition scheme, where message is mixed with the chaotic signal, is studied here. A representative case employing Daubechies wavelets and a typical Rossler-oscillator-based communication system is reported. Using a time scaling parameter modifies the spectrum of the mask. The results show that wavelets are effective only in particular cases with poor masking. The fast Fourier transform analysis demonstrates that the spectrum of the chaotic mask shows no holes and then other digital filtering techniques such as Wiener filters or comb filters must also been disregarded.