Research On The Soft-feedback Turbo Equalization Based On Sparse Adaptive Algorithms For Underwater Acoustic Communication

In this paper,the Turbo equalization algorithm is studied for the inter-symbol interference phenomenon in the underwater acoustic communication,and the sparse characteristic is used to improve the performance of the sparse adaptive algorithms,also,the adaptive algorithm is integrated into the Turbo receiver to effectively reduce the bit error rate.Under the framework of Turbo equalization,two kinds of soft feedback algorithms are studied in this paper:soft feedback equalization and soft decision feedback equalization,and the difference between the two algorithms'principles and performance are compared.These two algorithms are based on the same basic principle and the performances are similar,and the choice is based on the SNR and the computational complexity.In adaptive algorithms,we focus on the linear constraint series of normalized LMS algorithm,and the variable step-size algorithm which may speed up the convergence is researched.Sparsely adaptive algorithm has lower mean square error at the cost of computational complexity.₁l-WCNLMS algorithm gets the lowest MSE among linear constraint series of normalized LMS algorithm,the more sparsely the channel is,the more obvious its advantage gets.The MSE of sparse adaptive algorithm can be lower than the original algorithm in the same condition,and the convergence speed is faster too.The sparse adaptive algorithm is applied in Turbo receiver,and the performance is consistent with the independent channel estimation simulation results.Simultaneously,we extend the whole Turbo receiver?including soft equalization algorithm and adaptive sparse channel estimation,etc.?to the higher order PSK modulation,in order to win a high data transmission speed.In QPSK and 8PSK's Turbo equalization,the BER of the same SNR decreases with the increase of the modulating order,but the data transmission rate is times that of BPSK.In sparse underwater acoustic channel,Turbo equalization algorithm can effectively restrain the ISI,reduce the bit error rate,Also,according to the results of the experiment data processing,adaptive sparse algorithm is better than conventional algorithms in the aspect of improve the channel estimation accuracy,the Turbo receiver under the condition of high order PSK modulation meet the same performance characteristics.Through the principle analysis,computer simulation and experimental data processing of underwater acoustic conditions,this paper fully verifies that the Turbo equalization based on sparse adaptive algorithms can effectively reduce the bit error rate.