Signal Processing Of Chaotic Communication Based On Particle Filter

Chaotic communication becomes an important research in communications for itshigh security since the electronic circuit in the chaos and chaos synchronization wasfound. At present, the filtering of chaotic signals, chaotic synchronization, noise in thelong-distance chaotic signal transmission, channel distortion and multi-path effects arekey to practical applications of chaotic communication. Because of the chaotic signalsare nonlinear, most of the problems are considered to be the nonlinear filtering. Atpresent, the linear and nearly linear signal processing techniques are often used to solvethose problems, such as extended Kalman filter and unscented Kalman filter. However,these methods can not get a better result of filtering when the chaotic system can notmeet the Gaussian assumption. Particle filter is a nonlinear filtering method based onMonte Carlo simulations under the framework of recursive Bayesian filter. It is animportant research in nonlinear signal processing and suitable for any nonlinear andnon-Gaussian model. This paper studied the chaotic synchronization, demodulation ofinformation signals and parameter estimation of chaotic systems. The main content andinnovation are as follows:（1）This paper studied the chaotic synchronization based on extended Kalman filter,unscented Kalman filter and particle filter from the perspective of filtering. Furthermore, the chaotic synchronization of particle filter is discussed in detail. In order toovercome the particle depletion and improved the performance of chaoticsynchronization based on particle filter, the disturbance method which add theroughening noise to the particles is used to increase the variety of the particles. Theperformance of chaotic synchronization of these three methods is compared. Thesimulations show that the chaotic synchronization can be made effectively throughparticle filter and the chaotic synchronization of particle filter is robust and has lowmean square error compared with the synchronize methods of extended Kalman filterand unscented Kalman filter in low signal to noise ratio.（2）A novel demodulation scheme of antipodal chaos shift keying is proposed. Thescheme solves the problem of high bit error rate caused by coherent demodulationscheme when the chaotic synchronization is not ideal. The extended Kalman filter algorithm and the particle filter algorithm are adopted to track the chaotic signal in thesimulation in this scheme. The result show that the bit error rate performance of thisscheme based on particle filter is better than which based on extended and has strongeranti-noise performance.（3）Chaotic communication with parameter modulated based on particle filteralgorithm is proposed because of the extended Kalman filter algorithm and unscentedKalman filter algorithm have a bad estimation performance of chaotic system state andparameter. In order to get a better estimation performance, initial sampling of theparameter are randomly selected in fixed parameters. The estimation performance ofchaotic parameter of extended Kalman filter, unscented Kalman filter and particle filterare compared in the simulation. The result show that particle filter algorithm in chaoticparameter estimation has shorter convergence time and lower estimation error.