Research On Variable Structure Blind Equalization Algorithm Based On Underwater Acoustic Channel

Because of the disadvantages such as multipath transmission,time-delay spread,Doppler effect and frequency-related propagation loss in underwater acoustic channel,how to carry out efficient and stable underwater acoustic communication has become an urgent problem to be solved.On the basis of traditional blind equalization algorithm and neural network blind equalization algorithm,a parallel structure blind equalization algorithm based on underwater acoustic channel is proposed in this study.Compared with traditional blind equalization algorithm and neural network blind equalization algorithm,the convergence speed and stability of this algorithm are greatly improved,and the performance of underwater acoustic communication can be effectively improved.Aiming at the complex and changeable underwater acoustic channel,this study deeply analyze the physical characteristics of underwater acoustic channel such as propagation loss,background noise,multipath effect,etc.The noise is eliminated by the blind equalization algorithm.The frequently-used blind equalization algorithms such as Least Mean Square(LMS),Recursive Least Square(RLS)and Constant Modulus Algorithm(CMA)are studied theoretically and simulated experimentally.Experiments show that the equalization effect of linear equalizer is not ideal,so this research adopts the equalization algorithm of nonlinear structure.Neural network shows good performance in the field of learning and prediction.It can approach the given nonlinear function within any precision range.Therefore,the neural network blind equalization algorithm has better performance than the linear equalizer blind equalization algorithm.In this study,the Radial Basis Function Neural Network(RBFNN)blind equalization algorithm is studied.The Phase Transmittance Radial Basis Function Neural Network(PTRBFNN)blind equalization algorithm is proposed to solve the phase rotation problem caused by the RBFNN blind equalization algorithm in underwater acoustic channel.It is verified by experiments that the phase transmittance radial basis function neural network blind equalization can solves the problem of phase rotation and greatly improves the underwater acoustic communication performance.In order to overcome the disadvantages of neural network blind equalization algorithm,this study proposes a CMA-PTRBFNN blind equalization algorithm based on traditional blind equalization algorithm and neural network blind equalization algorithm,which improves the equalization efficiency of the system and reduces the bit error rate.At the same time,the FC-CMA-PTRBFNN blind equalization algorithm is proposed by referring the Fuzzy Controller(FC).The convergence speed is further accelerated and the convergence performance is improved by updating the parameters of CMA and PTRBFNN in parallel structure.Finally,the performance of the algorithm is verified by experimental simulation,and the performances of FC-CMA-PTRBFNN blind equalization algorithm and CMA-PTRBFNN blind equalization algorithm in underwater acoustic channel are compared.Experiments show that FC-CMA-PTRBFNN blind equalization algorithm has better performance,and the mean square error and bit error rate are greatly reduced,which effectively improves the underwater acoustic communication performance.