Algorithm And DSP Simulation Of Blind Equlization Based On Renyi Entropy Using Rayleigh Distribution And Wavelet Transformation

Blind equalization technique is one of key technology in communication system. Under the premise of no increasing the bandwidth of the system, it can improve bandwidth utilization rate, and avoid inter-symbol interference (ISI) introduced by limited bandwidth and multi-path transmission. With the rapid development of wavelet transform theory, a new way is supplied for transform domain blind equalization algorithm study. Autocorrelation of input signal various components can be reduced effectively after orthogonal wavelet transform, and the convergence rate of equalization algorithm can be speeded up effectively. In recent years, Renyi entropy is successfully applied in the Adaptive filtering, Pattern recognition, Source separation, Deconvolution nuclear balance based on the large number of training sequence and practical applications. By minimizing the entropy, the more information of the output signal can be provided and the performance of the equalizer will be improved. But it is a very significant topic to research on channel blind equalization technology by means of wavelet theory, Renyi entropy and making full use of their superiority. Aiming at the disadvantage of CMA (Constant Modulus Algorithm), the blind equalization algorithm based on Renyi entropy are studied deep in this dissertation. The main contributions are as follows:1. In order to overcome the contradiction between the convergence rates and accuracy for using fixed steps in traditional Constant Modulus Algorithm (CMA), a hyperbolic tangent function is established the between the step factor and the error signal. Then the wavelet theory was introduced into the design of the equalizer and the orthogonal wavelet expression of the equalizer is derived. Finally, the wavelet blind equalization algorithm based on hyperbolic tangent function is proposed. Simulation results with an underwater acoustic channel model demonstrate that the proposed algorithm has faster convergence rate and lower residual error.2. The method of the Parzen window probability density estimation is introduced, and Renyi entropy of Rayleigh distribution is derived using Rayleigh function. Finally, a new algorithm for wavelet based on hyperbolic tangent step-size and Renyi entropy using Rayleigh distribution was proposed. Theoretic analysis and computer simulations demonstrate that the presented algorithm has fast convergence rates, small steady-state error and good noise immunity.3. According to simulation parameters from Matlab, WT-Tanh-CMA algorithm is simulated by using TMS320VC5502 chip. CCS is used as development environment of DSP. In the CCS, code of the algorithm is compiled using C language, and is optimized. Eye diagram of signal is showed on the host successfully.Figure [30] table [1] reference [81]...