Research On Iterative Equalization For Underwater Acoustic Modem

With the increasing in demand and ability of human activities under marine environ-ments,people pay more attention on the researches of underwater acoustic communication.Due to the particularity of underwater acoustic channels,namely the large delay-spread and time-varying characteristics,high-speed and reliable data transmission has been diffi-cult to realize.Considering the multipath characteristics of underwater acoustic channels.the application of iterative equalization and decoding system is studied based on the Turbo coding related iterative equalization and decoding theory.Firstly,the channel coding and equalization techniques for realizing reliable data transmission were studied.The performance of block code and convolution code,together with various decoding methods were compared.The performance of liner equalizer,de-cision feedback equalizer and the optimization algorithm for updating filter coefficients were analyzed.Simulations were performed with proper coding and equalization methods for underwater acoustic channels.Secondly,the structure of Turbo coding cascaded decision feedback equalizer and Turbo/RSC coding based liner equalizer and decision feedback equalizer were studied.Special attention were paid to the structure of SISO(Soft-input and Soft-output,SISO)Turbo coding iterative decision feedback equalizer and symbol-by-symbol decoder.Sim-ulation using QPSK modulation under time-variant channel was performed.Results show that the performance of Turbo code and SDFE iterative equalization decoding structure is the best,RSC code and SDFE iterative equalization decoding structure performance is the second,and RSC code and LMMSE based linear equalizer iterative equalization decodingstructure performance is the worst.Thirdly,the symbol-by-symbol iterative decoding structure was analyzed.RLS al-gorithm was used to adaptively update the filter coefficients for better equalization per-formance in time-variant acoustic channels.Simulation results indicate that in low SNR(Signal-to-noise Ratio,SNR)conditions,symbol-by-symbol iterative decoding offers low-er BER(Bit Error Rate,BER)and decreased ISI(Inter-symbol-interference,ISI).At last,the reliability of iterative equalization decoding method is analyzed by pool experiment.