一种水声通信中的多阵元Turbo均衡算法

发布时间：2019-01-01 12:26

【摘要】：Turbo均衡应用在水声通信中的问题主要在于水声信道时间扩展长,多接收阵元处理复杂度较高。该文研究了将时间反转与马尔可夫链蒙特卡罗(MCMC)均衡联合优化算法用于实现Turbo均衡。首先进行时间反转实现多接收阵元较长多径时延的压缩,再利用白化滤波器解决时间反转造成的噪声模型失配问题,最后利用复杂度较低的MCMC均衡器结合软迭代信道估计对时间反转合并后得到的信号进行均衡。结合真实实验信道条件对信道响应估计的误差建立模型,通过仿真比较得出,该算法在相同条件下相对于多阵元直接自适应Turbo均衡算法复杂度降低67%,且有1.6 dB的误码率性能增益。通过对湖上试验数据进行处理,进一步验证了该算法的优势。
[Abstract]:The problem of Turbo equalization in underwater acoustic communication is that the time spread of underwater acoustic channel is long and the processing complexity of multi-receiver array elements is high. In this paper, the joint optimization algorithm of time reversal and Markov chain Monte Carlo (MCMC) equalization is studied to realize Turbo equalization. First, the time reversal is used to compress the multipath delay of the multi-receiver array, and then the whitening filter is used to solve the problem of noise model mismatch caused by the time reversal. Finally, the MCMC equalizer with low complexity and soft iterative channel estimation are used to equalize the signals after time reversal. Combined with the error of channel response estimation under the real experimental channel condition, the simulation results show that the complexity of the proposed algorithm is reduced by 67% compared with the multi-element direct adaptive Turbo equalization algorithm under the same conditions. The BER performance gain of 1. 6 dB is also obtained. The advantages of the algorithm are further verified by processing the experimental data on the lake.
【作者单位】：中国科学院声学研究所海洋声学技术实验室;中国科学院大学;中国科学院声学研究所声场声信息国家重点实验室;
【基金】：国家863计划项目(2002AA401004,2009AA093301,2009AA093601)资助课题
【分类号】：TN929.3

【参考文献】