Research On Long Distance Signal Equalization Based On Underwater Acoustic Channel

The exploration and development of marine resources have become increasingly demanding for the quality of underwater real-time communications.The electromagnetic wave due to rapid attenuation in the water,its application is limited in the underwater environment.Underwater acoustic communication can transmit longer distances,but its speed is slow and is hard to be real-time.We propose a method of using multi-band data transmission at the transmitting end,which can remarkably improve the amount of information per unit symbol compared to a common binary modulation scheme.Under the conditions of limited bandwidth,it greatly improves the communication rate and efficiently use the channel bandwidth.We analyze the existing equalizer structure,add the SAG decision-making algorithm based on the structure of the FFE-DFE combinatorial equalizer,and propose a new equalizer structure to solve the error diffusion problem.Then,we analyze the existing iterative algorithm of equalizer weight coefficients.In order to solve the problem of existing algorithms,such as the slow convergence speed,large steady-state error,complicated algorithm and hard to be transplanted in hardware,we propose a fast convergent adaptive variable step size LMS algorithm based on arccosine step function and three parameter adjustment factor.Through MATLAB,the three-parameter factors ?,?,r are simulated and the performance of the algorithm is optimized,and the optimal parameters of the algorithm are obtained.The simulation results are compared with the LMS algorithm based on fixed step size,the variable step size LMS algorithm based on modified arc tangent and the variable step size LMS algorithm based on hyperbolic secant function.The simulation results show that the proposed algorithm has a faster convergence rate of 57.9% and a steady-state error of 5dB than the LMS algorithm with fixed steps,has convergence rate of 26.3% and 15.8% higher than that of hyperbolic secant and modified a-tangent LMS algorithms respectively,and also has 1~2dB lower state error.Finally,the algorithm is transplanted to the ARM signal processing module to build an underwater acoustic channel experiment environment.Experiments show that the signal affected by inter-symbol interference caused by underwater acoustic channels can be recovered after being passed through the equalizer,this system can effectively overcome the inter-symbol interference problem caused by multipath effects.