A Study On The Detection Of Chaotic Direct Sequence Spread Spectrum Signal And Chaos Synchronization

Nonlinear science is a basic subject which studies the common characteristics of nonlinear phenomena. Nonlinear method can describe the objective fact more reasonable. Chaos is the most important one in nonlinear science, which has generated a huge influence to modern science and technologies. Chaotic communication is one of the most excellence applications in engineering. This thesis studies the chaos theory and its application in non-cooperative chaos signal detection, chaos synchronization and parameters estimation of chaotic systems. Chaos detection is the precondition of the modeling of chaotic communication signal, which includes chaos identification and the signal detection based on chaos system. Chaos identification can identify the chaotic characters from the observed time series, and the signal detection based on chaotic system exploits a chaotic system to detect weak signal. Chaos synchronization is the powerful guarantee to implement chaotic communication. And just the achievement of the synchronization of two chaotic systems with unidirectional couple in electronic circuit, the chaotic communication becomes possible. The generalized chaos synchronization is one of the main methods to achieve the analysis of the chaotic communication signals based on chaos synchronization. Chaotic parameter modulation is the main way of chaotic communication, which uses the chaotic signal from different parameter set of the chaotic system to hide the transmit information and achieve secure communication. The parameters estimation of chaotic systems is the necessary precondition to information demodulate. The main results of this thesis are:1,An efficient method of distinguishing chaos from noise is proposed. The changing trend of the average distance between phase points in phase space is studied with the increase of the embedding dimension. According to the different characteristics between chaos and noise in phase space, this thesis deduces a method which can distinguish chaos from noise. This method can do decision quickly even when the data are short and contaminated. This method not only overcomes the shortage of the subjectivity of parametric choice and time consuming of traditional method, but also improves the shortcoming of both the requirement for a large number of data and the noise sensitivity of the invariant of chaotic system to some extent.2,The detection of baseband chaotic direct sequence spread spectrum signal(CDSSS) is studied. In allusion to the real number chaotic spread spectrum mode in CDSSS communication, this thesis proposes to do detection by false neighbor method. Essentially, the detection of the real number CDSSS signal is the problem of distinguishing chaos from noise. Based on the analysis of the influence of information code to chaotic dynamics, this paper uses the identification algorithm which based on false neighbor to detect the baseband real number CDSSS signal. This method not only overcomes the shortage of both the requirement for a large number of data and the noise sensitivity of the invariant of chaos system, but also can distinguish traditional DSSS signal to some extent.3,For quantization CDSSS signal, this thesis proposes a quantitative signal detection method by using Duffing oscillator based on Hamiltonian. Duffing oscillator exploits the characteristics that additional perturbation signals make the dynamics of the system change obviously and immune to noise to detect weak signal. This improves the performance of the weak signal detection greatly. According to the analysis of the dynamics characteristic, the time-frequency characteristic, the immune to noise characteristic, the state transition characteristic and the influence of the frequency and phase difference with the perturbation signal of Duffing oscillator, this paper proposes to use state transition of the phase portrait and intermittent chaos of the time series diagram to detect signal respectively. Considering most of the existing weak signal detection methods based on Duffing oscillator are stay on qualitative and simulative analysis also, this paper proposes to use Hamiltonian to construct statistic to do quantitative decision. Hamiltonian can depict the dynamics real time. This scheme can detect weak signal quantitatively in lower signal-noise-ratio. 4,An adaptive hybrid generalized synchronization of two different chaotic system with unknown parameters is proposed. This scheme can achieve synchronization and track of two different chaotic systems. Considering the influence of unknown parameters to the stability of the synchronization manifold, based on adaptive theory, this paper exploits the Lyapunov stability theory to design the control signals and parameters update laws analytically. This scheme can achieve complete synchronization, anti-synchronization, projective synchronization, simultaneously, and it can be extended to more general generalized chaos synchronization. The modeling circumstance is closer to the actual non-cooperative signal analysis. Numerical simulations verify the effectiveness of the proposed scheme.5,The anticipating chaotic synchronization based parameter estimation of chaotic system is proposed. Considering the communication systems are at difference places, the signal transmit has time delay to each other, the estimation of the unknown parameters of the delay coupled chaotic system from the observed time series is significant. This paper investigates the parameter estimation of time delay coupled chaotic systems. Although the Krasovskii-Lyapunov functional method often results in delay-independent stability condition, time delay closely relates to the parameter estimation. The numerical method is used to study the influence of the parameters of the proposed scheme to unknown parameter estimation scheme. This scheme can estimate the unknown parameter correctly even when the time delay exists, and improve the parameter estimation based on complete synchronization.