Study On The Signal Processing Method Based On Chaotic Theory

Chaos in nonlinear circuits and systems is one of the advanced topics in nonlinearscientific field. Among many topics about chaos study, the generation, prediction andthe application of chaos are main three hotspots. The use of chaos in signalprocessing, communication and control has been a very active area of research inrecent years.Since chaos is a deterministic complex phenomenon, it opens up a new field forsignal processing. This dissertation focuses on developing efficient algorithm forsignal detection, parameter estimation of radar and also on radar signal generator ?chaotic radar waveform for better performance. This study belongs to fundamentalresearch aiming to engineering application.The main contents of this dissertation are follows:The mathematical model of chaos is discussed deeply from mathematical map,statistical viewpoint for the purpose of analysis and synthesis chaotic signal.The dissertation analysis chaotic signals deeply, mainly about the correlationproperty(including auto-correlation and cross correlation) and Lyapunov exponent.The study shows the ideal auto correlation and cross correlation of the chaotic signalsby using theoretical derives and computer simulations. Simulations reveals that theresults is common, this property is the main fundamental of the following study. Andthe sign of maximum Lyapunov exponent is the flag of chaos.Radar is used to detect and identify different targets. As the name has shows its mainpurpose, one of the main task of radar is to detect the target in vary background.Among which the detection in sea clutter is the most difficult. Recent researches haveshowed that sea clutter can be modeled as chaotic dynamic. Local predictiontechnique is used for detection and estimation problem in chaotic environment forradar signal processing.DO A estimation is often encountered in vary application of radar and communicationsystem. The dissertation then discusses the estimation of interested signals in chaoticnoise. We analyzed three different algorithm for this purpose. For all the threemethods , we give out the accomplish steps and simulation results.focuses on the application of chaos signal for radar waveform generation. Based on the special property of chaotic signals, we presented the chaotic radar signals. Additional, we introduce the chaotic sequence to the digital watermark technique. To sum up, the contribution of this dissertation are as following: The main progress and some statistical results of chaotic signals are summed up, the similarity and distinguish between chaos and noise are discussed and shown out. For the detection problem in chaotic noise (clutter), this paper presents to use the RBF neural network to solve it. Analysis of the algorithm is given out. Simulation results and real-life data are used to show the feasibility of the proposed method. LS-AR spectrum method and GA-MPSV are presented to solve the estimation problems in chaotic noise. As GA-MPSV algorithm considering the dynamical property of chaos, GA-MPSV algorithm is theoretically better than LS-AR method. However MPSV is much more complex the LS-AR algorithm. For these reasons, Neural network predicting method combined with Autoregressive spectrum estimation(NN-AR for short) is presented to solve the problems. This dissertation proposed and recommended the NN-AR method for real application to have good trade-off between computation burden and performance improvement. After enough analysis and computer simulation, we sum up that as one potential waveform for "Low Intercept probability" radar system, chaotic signals provided with the following properties:(l)In performance, contrast to normal random signal and noise signals, chaotic radar signal has ideal function, it doesn't exist the step phenomena ,which means it can work in any environment; (2) from signal generation and processing aspects, chaotic signals in easy to signal generation a...