Noise Reduction In Chaotic Signal And Its Application In The Communication

Chaos is a frontal science and research hotspot, which is in the extensive attention. Chaotic signals are non-periodic and bounded signals resulting from deterministic nonlinear system. Chaotic signals are sensitive to initial conditions and look like"noise"in time domain. Because of the noise-like feature, chaotic signals occupy a wide bandwidth in frequency domain. Those characteristics mentioned above make it have important application significance in safety communication. At present, a lot of countries are dedicated to researching chaotic communicational secrecy.However, most researches on chaos-based communication systems are based on the assumption of a rather ideal transmission channel, which neglects nonideal channel's influence on chaotic synchronization. The disturbance of real communicational environment influences performance of system based on chaotic communication, which may lead that system doesn't work. In all disturbances ,the most serious disturb is noise disturb. Real physics of the transmitted broadband chaotic signals, the multipath effect in wireless communication channel and the time-varying feature of channel can also distort the transmitted chaotic signals. These factors affect the realization of the proposed chaos-based communication system. As done in conventional wireless communication systems, channel equalization is an important technique for combating these unwanted channel interference in chaos-based communication systems.Because of the important significance of researches based on chaotic communication, this thesis will focus on chaos-based communication systems, and address two basic issues ,one is the filtering for noisy contaminated chaotic signals, another is the blind channel equalization for chaos-based communication systems. The start point of this thesis is to make use of adaptive filter algorithms and the modeling technique for signals.We do these works as follows:At first, we introduce two chaotic systems, i.e., one dimensional logistic map, and three- dimensional Lorenz map. After confirm the parameter of system, we simulate. For logistic map, its parameter is 4, simulation indicate system brings chaotic sequence. For Lorenz map, its parameters are 10,28,8/3, we get three dimension map of three states, and two dimension map of any two states, simulation indicate system brings chaotic sequence.Then, the extended Kalman filter algorithm will be used to filter noise contaminated chaotic signals. After expressing the filtering problem into the estimation issue of the mixed state-space model, we consider two chaotic systems, i.e., one dimensional logistic map, and three- dimensional Lorenz map. It indicates by computer simulation that this filtering method can effectively reduce noise, but for Lorenz map ,effect is not good.The unscented Kalman filter algorithm will be used to filter noise contaminated chaotic signals. After expressing the filtering problem into the estimation issue of the mixed state-space model, we consider two chaotic systems, i.e., one dimensional logistic map, and three- dimensional Lorenz map. It indicates by computer simulation that this filtering method can effectively reduce noise. By comparison, filtering effect of the unscented Kalman filter is better than that of the extended Kalman filter.Finally, the performance of the two adaptive filtering algorithms, the extended Kalman filter, the unscented Kalman filter, will be compared when they are applied in blind equalization for the chaos-based communication systems. The results demonstrate by computer simulation that these filtering algorithms can realize the blind equalization. By comparison, the convergence ability of the unscented Kalman filter algorithm outperform that extended Kalman filter algorithm. The blind equalization performance of the unscented Kalman filter outperforms that of the extended Kalman filter.