Research On Blind Source Extraction Algorithm Based On Virtual Receiving Array In Chaotic System

With the development of information technology, society is moving into the world of globalization and the information age. Communications security is becoming the hotspot and focus issues studied by scholars. Technologies related communication systems privacy are more and more appreciated and the realization of secure communications of information data is one of the globally hot topics. The introduction of chaotic signal into secure communication is particularly common among numerous studies in the field of information security reports according to its non stationarity, sensitivity to initial value, wide spectrum and so on which makes difficulty to crack chaotic signal completely providing effective means for secure transmission of information. In this paper, taking harmonic signal transmission under chaotic background as an object, researching on harmonic signals extraction based on underdetermined system, considering the actual communication channels with additive noise, overcoming the difficulty that chaotic signal and noise signal are both wide spectrum, comprehensive utilization of time-frequency analysis and independent component analysis theory to solve blind extraction of harmonic signals in underdetermined systems, provide important theoretical basis for verifying the effectiveness of chaotic secure transmission.In this thesis, harmonic signal transmission is under the background of typical three-dimensional chaotic dynamical system. Due to chaotic signals are non-stationary, conventional Fourier transform would be helpless. In order to effectively hide harmonic signals, use wavelet transform including in time-frequency analysis to identify chaotic signal energy accumulation zone. Harmonic signals will be hidden in this accumulation zone to transfer, with an emphasis on "chaos energy-band harmonic signals extraction." The verifiability simulation in this paper is divided into two parts. Firstly, analysis time-frequency characteristics of chaotic signals according to wavelet transform to determine parameters of harmonic signals and interference signal. Form the original data through positive definite matrix channel and introduce additive white Gaussian noise into channel, analysis observational data by independent component analysis for blind extraction of the desired signals. Change the model into underdetermined mixed to close to the practical application of the communication system, namely underdetermined channel mixing matrix. Because the number of observed signals is smaller than the number of source signals and general communication signals don’t meet the sparsity, so use empirical mode decomposition to construct virtual receiver array to resolve this problem and propose a way to realize underdetermined blind extraction of harmonic signal. Take advantage of low complexity of smoothing algorithm to reduce interference of noise signal on extracted signals. The effectiveness and reliability of this algorithm has been verified by a large number of simulations. Quantitative analysis of related factors that have impact on the algorithm have proven algorithm’s stability and universality.