Wavelet Transform Blind Equalization Algorithm Based On The Optimization Of Fruit Fly Algorithm

At present, the research of fruit fly optimization algorithm and its application is very few at home and abroad.And it is no report abaout the research combined with the three algorithms in the field of underwater acoustic communication, the wavelet transformation, single fruit fly optimization algorithm and the blind equalization algorithm. So it is a great significance to research on underwater acoustic communication signal processing by combing the three algorithm. SFOA and improved SFOA are used to optimize blind equalizer, and the wavelet transformation theory is used to decrease the autocorrelation of the input signals and improve balance performance in blind equalization algorithm.The major achievements are as follows:1wavelet transform blind equalization algorithm based on the optimization of single fruit fly algorithmIn 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. Single fruit fly algorithm and orthogonal wavelet basis function are used to find the equalizer initialization vector quickly to avoid producing local limit that caused by the gradient descent method to find the most optimal vector.2. Orthogonal wavelet transform adaptive constant modulus blind equalization algorithm based on the optimization of single fruit fly algorithmIn order to overcome the slow convergence rate and big mean square error of Soft Constraint Satisfaction Constant Modulus Algorithm (SCS-CMA), an orthogonal wavelet transform adaptive constant modulus blind equalization algorithm based on single fruit fly optimization algorithm is proposed. The proposed algorithm makes full use of the de-correlation ability of orthogonal wavelet transform to accelerate convergence rate of SCS-CMA and reduce the steady state error and the global fast optimization ability of single fruit fly algorithm to optimize performance of equalizer further. Simulation results in underwater acoustic channels show that the raised algorithm has outstanding performance.3. Orthogonal wavelet transform generalized split multi-modulus blind equalization algorithm based on hybrid single fruit fly algorithm optimization and simulated annealing For strengthening the equalization performance of high-order QAM signals, improving the convergence speed of blind equalization and reducing the mean square error further, an orthogonal wavelet transform generalized split multi-modulus blind equalization algorithm based on hybrid single fruit fly optimization algorithm and simulated annealing is proposed. SFOA was combined with simulated annealing to avoid premature phenomena of single fruit fly optimization algorithm (SFOA) and simulated annealing and single fruit fly algorithm mixed is used to optimize the generalized split multi-modulus blind equalization algorithm in the proposed algorithm. The global accurate optimization ability of hybrid algorithm accelerates convergence rate of generalized split multi-modulus blind equalization algorithm. The strong de-correlation of orthogonal wavelet transform can reduce the autocorrelation of the input signals so that it can reduce the steady-state error. The best equilibrium performance of the new algorithm was proved by computer simulations.4. Orthogonal wavelet transform blind equalization algorithm based on the optimization of many swarms cooperative fruit fly algorithm(1) Many swarms cooperative fruit fly optimization algorithmAimed at solving the problem that fruit fly optimization algorithm may fall into local optimization and be poor in performance of precision when it search the complex area, Many swarms cooperative fruit fly optimization algorithm is proposed, which introduced collaborative evolutionary into fruit fly optimization algorithm. In many swarms cooperative fruit fly optimization algorithm, parallel topology and postive feedback of information communication were used, and many fruit fly swarms are regarded as many independent swarms to search optimum and tracked the other swarms global optimum. Guided by using cooperative evolution, positive feedback strategy and present optimum, many swarms cooperative fruit fly optimization algorithm has good property in rate of convergence and the precision of solution.(2) Orthogonal wavelet transform blind equalization algorithm based on the optimization of many swarms cooperative fruit fly algorithmMany swarms cooperative fruit fly optimization algorithm is applied to blind equalization algorithm and orthogonal wavelet transform blind equalization algorithm based on the optimization of many swarms cooperative fruit fly algorithm is proposedOn 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 many swarms cooperative fruit fly algorithm.(3) Orthogonal wavelet transform multi-modulus blind equalization algorithm based on the optimization of many swarms cooperative fruit fly algorithmIn order to overcome the draw back of blind equalization algorithm, orthogonal wavelet transform multi-modulus blind equalization algorithm based on the optimization of many swarms cooperative fruit fly algorithm is proposed. The proposed algorithm makes full use of the fast global optimization ability of many swarms cooperative fruit fly algorithm to accelerate convergence rate of MMA, the de-correlation ability of orthogonal wavelet transform for the input signals to reduce the steady state and it is more applicable to equilibrium high-order QAM signal.