Incollection,

Extraction of buried signals from an Ikeda-type microwave system via empirical mode decomposition

, and .
Abstract Book of the XXIII IUPAP International Conference on Statistical Physics, Genova, Italy, (9-13 July 2007)

Abstract

Secure communications using chaotic waveforms has became a particular application of nonlinear dynamics appeared at the beginning of 1990s. For example, electronic circuits modeled by nonlinear ordinary differential equations were used very often for the generation of chaotic dynamics, however, the encryption efficiency of these electronic setups is limited by the embedded low-dimensional complexity. Recently, Ikeda-type delay dynamics (DD) has shown to be an outstanding candidate for chaos-based encryption in modern high speed optical telecommunications. It is known that the unique feature of DD is to exhibit extremely complex chaotic behaviours, which can be quantified in terms of Lyapunov spectrum of a given chaotic regime in a reconstructed phase space of finite dimension, and finally a Lyapunov dimension is derived. Dorizzi et al. had shown that the Lyapunov dimension has linear dependence with the ratio $T/\tau$, where $\tau$ is a characteristic response time and $T$ is the delay time. Larger et al. experimentally demonstrated a high masking efficiency in Ikeda-type optoelectronics intended for practical applications, where the Lyapunov dimension can be up to 470 when $T/\tau$ equals 60. Chaos synchronization is widely used for the encoding and decoding technique, in which the output of the receiver replicates the same chaotic waveform as that in the emitter. However, the decoding quality is strongly relied on matching conditions between the emitter and receiver elements. Thus, it would be interesting to ask that Is there a possible way to directly extract message from high-complexity encoded signals?. In this study, empirical mode decomposition developed by Huang et al. is employed to decode message from an Ikeda-type microwave system. Under considerations of the high-dimensional chaos encryption as well as the high masking efficiency, it is shown that a meaningful data set embedded in chaotic fluctuations can be directly filtered out in a reasonable way. Our approach is quite different from the well-known encoding/decoding technique by use of chaos synchronization. Furthermore, our satisfactory results suggest that empirical mode decomposition can be generally applied to data mining in various systems.

Tags

Users

  • @statphys23

Comments and Reviews