Blind Despreading Of Chaotic Direct Sequence Spread Spectrum Signals Based On Particle Filters

Chaotic signals that result from deterministic nonlinear systems are non-periodic and bounded. Chaotic signals exhibit several special characteristics such as broad Fourier transform spectra, extraordinary sensitivity to initial conditions, long time unpredictable property, and noise-like feature aperiodicity, which coincide with the requirements for signals used in communication systems. The chaotic direct sequence spreading spectrum scheme has been extensively studied as an approach of chaotic secure communication owing to its merits of physical-layer security and low probability of intercept. Therefore, the research on breaking the chaotic direct sequence spreading spectrum signals is fundamental and challenging research topic in chaotic communication field, also it can guide the further improvement of the chaotic direct sequence spreading spectrum scheme’s security. Breaking chaotic direct sequence spreading spectrum signals under the multipath fading channel and non-Gaussian noise is completely nonlinear. For nonlinear problems, many nonlinear filters have been proposed, such as extended Kalman Filter, which has been proved to have bad performance and complexity. Recently, the particle filter has been widely applied to solving the nonlinear and dynamic estimation problems. The particle filter is a sequential Monte Carlo based method which approximates the optimal filtering by representing the probability density function with a swarm of particles. Due to this particle based representation, the particle filter is able to represent a wide range of probability densities, and allow for online, real-time estimation of nonlinear and non-Gaussian dynamic systems.Denoise processing, channel equalization, multi-user signal separation and blind depreading are the key problems in the field of breaking the chaotic direct sequence spreading spectrum signals for practical implementation. In this dissertation, based on particle filtering, blind demodulation, blind channel equalization and blind separation are studied in detail. The main work of the research is as follows:Under the condition that the chaotic system parameters and initial value are unknown, an adaptive particle filter based on the innovation-based adaptive estimation approach is proposed. In addition, the Maximun-A-Posterior method is used to estimation the parameters. Therefore, the adaptive particle filter can alleviate the sample degeneracy problem and the semi-blind despreading is realized. In order to blindly despread the chaotic direct sequence spreading spectrum signals under the colored or non-Gaussian noises condition, a particle-filter-based algorithm is proposed according to chaos fitting. The colored or non-Gaussian noises are formulated by ARMA models, and the range-differentiating factor is imported into the intruder’s chaotic system equation. Simulations show that the proposed algorithm can obtain a good bit-error rate performance when extracting the original binary message from the chaotic direct sequence spreading spectrum signals without any knowledge of the transmitter’s chaotic map, or initial value, even there existing colored or non-Gaussian noises.In order to break chaotic direct sequence spreading spectrum signals under the multipath fading channel, a particle filter based algorithm combining blind channel equalization with chaos fitting is proposed. To implement this algorithm, the intruder substitutes a different chaotic equation into the state-space equations of the channel and the chaos fitting, and then multiple particle filters are used for blind channel equalization and chaos fitting simultaneously by implementing them in reciprocal interaction. As a result, the impact brought about by the multipath fading channel and additive noises can be overcome. Furthermore, the range-differentiating factor is used to make the inevitable chaos fitting error advantageous based on the chaos fitting method. Thus, the CD3 S signals can be broken according to the range of the estimated message. Simulations show that the binary message signal can be extract from the chaotic direct sequence spreading spectrum signals without any knowledge of the chaotic transmitter’s structure, parameters, initial value, or the channel characteristics.According the number of the receiver, the blind separation is divided into single channel blind separation and blind separation under the multiple fading channels in this dissertation. Firstly, the feasibility of single channel signal separation is discussed. Furthermore, the signal separation and despreading are realized simultaneously by using particle filtering based on the state-space equations of the single channel signal separation. And then, a MIMO model of the chaotic direct sequence spreading spectrum system under the multipath fading channels is built, furthermore, the channel coefficients and mix parameters are formulated by AR models. At last, a dual particle filter is proposed to separate and despread the chaotic direct sequence spreading spectrum signals through canceling multipath fading channel distortion, inter-user interference, and channel noise.