一种水声通信中的多阵元Turbo均衡算法

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。

An Algorithm of Multi-array Turbo Equalization of Underwater Acoustic Communication

The main problems of the application of the Turbo equalizer in underwater acoustic communication are long time spread of channel and the multi-array processing. The union algorithm of time reversal and Markov Chain Monte Carlo (MCMC) equalization is proposed. Time reversal compresses the long time spread by combining multi-array signal, then the whitening filter is adopted to the solution of the noise model mismatch, at last the MCMC equalizer under optimal Maximum A Posteriori (MAP) criterion realizes the soft-in soft-out equalizer with the channel information obtained by channel estimation of soft iteration. The simulation based on the real experimental condition is conducted for the error model of truncated channel estimation. Simulation results denote that, this algorithm gets 1.6 dB Bit Error Rate (BER) performance gain, and 67% complexity loss over adaptive Turbo equalization. In the real experiment conducted in a lake, result of data processing denotes that the union algorithm of time reversal and MCMC equalizer have a superior performance over the algorithm of multichannel adaptive Turbo equalizer.