OFDM Iterative Receiver Design In Time-varying Underwater Acoustic Channels

Orthogonal frequency division multiplexing(OFDM)is widely used in high speed underwater acoustic communications as it can effectively deal with the frequency selective fading of underwater acoustic channel and has high spectral efficiency.However,the time-varying characteristic of underwater acoustic channel will cause large Doppler spread,leading to severe inter-carrier interference(ICI)among subcarriers in OFDM underwater acoustic communication system,which degrades the system performance significantly.To improve the robustness of OFDM underwater acoustic communication system,we adopt the structure of iterative receiver and focus on its key technologies in this paper.The research work of this paper is as follows:First,the model of OFDM underwater acoustic communication system is presented,and the influence of Doppler spread caused by time-varying underwater acoustic channel is analyzed.Second,nonbinary QC-LDPC code is used as the coding scheme of OFDM underwater acoustic communication system.Meanwhile,three decoding algorithms QSPA,log-SPA and max-log-SPA are described.The performance of the three decoding algorithms is compared by simulation.Third,in order to reduce the complexity of time-varying channel equalization,the banded frequency domain channel matrix is adopted in channel equalization.We compared the performance loss of channel equalization caused by the banded frequency domain channel matrix with different approximations by simulation.Fourth,taking advantage of the fact that the underwater acoustic channels are sparse,channel estimation methods based on compressed sensing are used to obtain the frequency domain channel matrix.To deal with the problem of time domain leakage in discrete channel impulse response of OFDM underwater acoustic communication system,we propose Modified SP and Modified SAMP algorithm based on Subspace Pursuit(SP)algorithm and Sparsity Adaptive Matching Pursuit(SAMP)algorithm respectively.The Modified SAMP algorithm can get accurate channel estimation while the channel sparsity is unknown.In simulation,we found that the performance of Modified SAMP algorithm is better than that of OMP algorithm.Last,the experiment was conducted near the Wuyu Island in Xiamen.The performance of the proposed Modified SAMP algorithm is verified.Meanwhile the experimental data shows that,the iterative receiver can achieve lower BER and has higher robustness compared with the noniterative receiver.