Wavelet Transform Blind Equalization Algorithm Based On The Optimization Of Artificial Fish Swarm Algorithm

In underwater acoustic communication, Inter-Symbol Interference caused by multipath fading time-varying underwater acoustic channels seriously affects communication quality and the limited bandwidth of underwater acoustic channels affects the data of the transmission speed and the reliability. In order effectively overcome ISI and improve the bandwidth utilizations, on the basis of analyzing blind equalization algorithms research situation at home and abroad, a new train of thought is proposed. In this proposed thought, in order to research the blind equalization theory algorithms, simulations, and performance assessment, wavelet theory is regarded as an important tool, artificial fish swarm algorithm, are called as the optimization method, and fractionally spaced, super exponential iteration, and blind equalization algorithm are combined. The main work is as follows:1. Orthogonal wavelet transform blind equalization algorithm based on the optimization of artificial fish swarm algorithm is proposed(1) Blind equalization algorithm based on the optimization of artificial fish school algorithm is proposedIn traditional Constant Modulus blind equalization Algorithm (CMA), the methold of searching optimal weight vector is a gradient descent search method, which can obtain the equation of the equalizer weight vector via using the continuous and derivable cost function, but the methold is easy to fall into local convergence. Aiming at the disadvantage of the searching methold of traditional CMA, on the basis of analyzing the feature of artificial fish swarm algorithm(AFSA), it is introduced to the CMA, so the artificial fish swarm algorithm based CMA(AFSACMA) is proposed. In this proposed algorithm, the fitness function of AFSAis defined by cost function of blind equalization algorithm. It makes full use of the global random searching feature of the AFSA to find the global optimal weight vector of the equalizer. Simulation results with underwater acoustic channel show the superiority of the AFSACMA.(2) Orthogonal wavelet transform blind equalization algorithm based on the optimization of artificial fish swarm algorithmOn the basis of making full use of the strong de-correlation of the orthogonal wavelet transform, the structure of blind equalizer is designed by the equalizer inputs transformed by orthogonal wavelet basis function and the globe optimal weight vector of the equalizer is found by AFSA, so the orthogonal wavelet transform blind equalization algorithm based on the optimization of artificial fish swarm algorithm2. An orthogonal wavelet transform blind equalization algorithm based on hybrid artificial fish swarm optimization of mutation operator and simulated annealingIn order to overcome defect of the poor local search and premature promblem of artificial fish school algorithm (AFSA), AFSA was integrated with mutation operator for the sake of maintain the population diversity and restrain premature problem, while AFSA blend with simulated annealing to strengthen local search capability and hybrid artificial fish swarm algorithm optimization with mutation operator and simulated annealing was present. This algorithm initialize weight vector of blind equalizer, an wavelet transform fractionally spaced blind equalization algorithm based on hybrid artificial fish swarm optimization of mutation operator and simulated annealing was proposed.3. Generalized multi-modulus blind equalization algorithm based on chaos artificial fish swarm optimizationFor improving the equalization performance of high-order QAM signals, orthogonal wavelet transform generalized multi-modulus blind equalization algorithm based on chaos artificial fish swarm optimization is proposed. In this proposed algorithm, according to the prior information of higher-order QAM signal constellations, chaos artificial fish swarm algorithm is fused to generalized multi-modulus blind equalization algorithm. Accordingly, the proposed algorithm use rapid global optimum seeking ability of CAFSA to initialize equalizer weight vector and adaptive adjusts modulus value of objective function in the equalizer vector iterations. The theoretical analyses and computer simulations indicate that the proposed algorithm outperforms GMMA and CAFSA-GMMA in mean square error and convergence rate, which is more efficient for high-order QAM signal.4. Adaptive solution for frequency wavelet transform multi-modulus blind equalizat-ion algorithm based on immune artificial fish swarm optimizationAimed at solving the problem that the adaptive solution for multi-modulus blind equalization algorithm have not enough ability to detect blindly high-order quadrature amplitude modulation(QAM), adaptive solution for multi-modulus blind equalization algorithm based on immune artificial fish swarm optimization algorithm, the theory of fractionally space and fast Fourier transform, which is combination of frequency fractionally-spaced equalizers and immune artificial fish swarm is proposed. The proposed algorithm makes full use of the fast global optimization ability of immune artificial fish swarm to accelerate convergence rate of βMMA, the de-correlation ability of orthogonal wavelet transform for the input signals to reduce the steady state, and utilize circular convolution reduce calculated amount instead of liner convolution. The results from computer simulation show convergence rate, steady-state error and calculated amount of the proposed algorithm are superior to the βMMA.5. Super exponential iteration adaptive minimum entropy blind equalization algorithm Based on quantum artificial fish swarm optimizationIn order to improve equalization performance of high order inconstant modulus signals, Adaptive minimum entropy super-exponential iteration blind equalization algorithm based on quantum artificial fish swarm optimization is proposed, which add less calculated amount. The novel algorithm can accelerate rate of convergence via super-exponential iteration algoritnm, and utilize fast searching ability of quantum artificial fish swarm algorithm to accelerate rate of convergence and decease mean state error further. The simulation results demonstrate that the adaptive minimum entropy blind equalization has different equalization performance to the different modulation system and the new algorithm can speed up convergence and decrease state error.