Generation, Analysis And Processing Of Wideband Chaotic Signals

Wideband signals have a lot of performance advantages in radar and communication applications, such as high range-resolution, good electromagnetic compatibility, anti-jamming ability, and low-interception probablity. Since 1960, much attention has been taken on the signal research and generated great advances on signal generation, analyses and processing. With the advance on the new theroy and techniques, the new results appear as sequal. Chaotic signal, generated by deterministic systems, can be easily implemented by a simple circuit and controlled to the desired states. Since its founding, the singal has attaracted wide interests in applications. With the development of chaotic theory and appeal in applications, this dissertation mainly studies the generation, analysis and processing of chaotic signals.Main results are concluded as follows:1. New generating system and implementation of chaotic sourcesA Clopitts circuit drived by a chaotic binary code is implemented. Then we propose a chaotic system consisting of digital-analog mixed circuits. The digital circuit provides a pulse-driving signal for the generation of chaotic signals from the analog circuits. The analog circuit consists of the oscillator and the filters. The filters are used to shape the structure of the chaotic singals. The SPICE simulation and circuit implementation show that the new source has controllable spectra and easy to control and synchronize. It is also found that thecircuit can generate high-frequency chaotic signals with low-frequency driving pulse.2. Time-frequency analysis of chaotic signalsTime-scaling signals are key operation in the time-frequency analysis. For chaotic signals, we reveal the relation between the time-scaling and system parameters. With the relation, the time-scaling chaotic signals can be generated with the chaotic generating systems. We apply the generating methods to the analysis of chotic singals and develop the corresponding techniques for calculating ambiguity functions of the chaotic signals. For the generation of the time-scaling singals, we develop the pule-synchaonization-based and generalized-synchronization-based techniques with the corresponding autonomous and non-autonomous systems. The calculation of the amnbiguity functions can be implemented by analog-digital mixed structure and provides new strategy for the analysis processing of chaotic radar signals. 3. Partially-blind extraction and separation of chaotic singalsWe develop the synchronization-based techniques for the extraction and separation of chaotic signals. In the extraction of chaotic signals, we transform the estimation of the extracting vectors into the parameter estimation problem of the chaotic systems. The signal extraction is implemented by a synchronization-based parameter estimation of chaotic systems. In the separation of chaotic signals, we integrate the singal channel and multiple sources into a system. The developed parameter estimation method is used to implement the source separation. The separation technique fully utilizes the information of the chaotic generating systems and enhances the robustness of the separation. It is found that the new technique is still applicable under signal level below 0dB.