| Underwater acoustic communication technology is of great significance in marine observation,deep-sea exploration and development,and military use.The underwater acoustic channel is a time / frequency dual selective fading channel,the inter-symbol interference is extremely large,which hinders the development of high-speed underwater acoustic communication.Currently,the widely used underwater acoustic high-speed communication method is orthogonal frequency division multiplexing(OFDM)technology.Compared with OFDM technology,underwater acoustic single carrier frequency domain equalization(SC-FDE)technology has many advantages,such as low peak-to-average ratio,Resistant to frequency selective fading,and insensitivity to frequency offset have attracted many researchers,etc.Based on the SC-FDE system,this paper focuses on the synchronization technology and channel equalization technology under the underwater acoustic time-varying channel.First,it introduces the frame structure selection method,compares the performance of several unique words(UW),selects the optimal payload frame format and unique word sequence,and compares the effect of length of the UW sequence on the bit error rate(BER)performance of the underwater SC-FDE system,also introduces the key technologies in the single carrier communication system.Then,Research on synchronization technology in underwater acoustic SC-FDE system.First,a UW sequence-based synchronization method commonly used in single-carrier communication is given,but this method is affected by the underwater acoustic Doppler effect and Multipath spreading delay,witch is not conducive to the estimation of synchronization points.So a two-dimensional autocorrelation time synchronization and Doppler estimation method based on UW sequences is studied,which uses UW sequences before and after each data block of the underwater acoustic SC-FDE system to go for a two-dimensional autocorrelation search to get the precise time synchronization and Doppler factor estimation.At the end of this part,a numerical simulation analysis is performed,and it is concluded that the two-dimensional autocorrelation method based on UW sequences is significantly better than the cross-correlation method based on UW sequences in multipath channels and mostDoppler frequency offsets channel.Finally,in order to reduce the impact of signal-to-symbol interference,three frequency-domain equalization algorithms were studied,namely time-frequency domain decision feedback equalization(FD-DFE)algorithm,based noise prediction decision feedback equalization algorithm,and block iterative decision feedback equalization(IB-DFE)algorithm.The FD-DFE algorithm is a traditional equalization algorithm,that its feedforward part is in the frequency domain,and its feedback part is in the time domain to go for decision,which is made symbol by symbol.Based on noise prediction equalization method's feedforward is used the minimum mean square error equalization(MMSE)algorithm and feedback part is used the known characteristics of the UW sequence to estimated the noise of the received UW sequence and then to estimate the data noise,which Its complexity is lower than FD-DFE.Finally,the IB-DFE technology that incorporates LDPC codes in the iterative part is introduced.The feedback part is judged block by block,the computational complexity is lower than the first two equalization methods,and the BER performance is better than the two algorithms.The simulation and pool experiment results show that the BER performance of the FD-DFE algorithm is better than the BER performance of the decision feedback equalization algorithm based on noise prediction,and the BER performance of the IB-DFE algorithm is the best of the three algorithms,and the system BER performance related to UW sequence length and SNR. |
PDF Full Text Request