多输入多输出水声信道均衡问题是实现高速水声通信的关键技术难题。为了提高信道的传输性能，通常采用自适应均衡技术，通过递归最小二乘算法进行判决反馈均衡。由于传统均衡方法精度低、性能差，而引入深度学习可显著改善信道传输性能。因此，以传统方法的性能作为基准，对比研究基于深度学习网络（深度神经网络、卷积神经网络以及长短期记忆网络）的信道均衡算法。首先，利用水声信道仿真软件Bellhop构建深度学习数据源，然后训练深度学习网络，开展3种神经网络的信道均衡算法仿真试验。研究结果表明，基于卷积神经网络的信道均衡算法性能最优，且在输入序列为2000码元，等效3.65倍信道长度时误码率最低。
Multi-input multi-output underwater acoustic channel equalization is a key technical challenge in realizing high-speed underwater acoustic communication. In order to improve the transmission performance of the channel, adaptive equalization technology is usually used, and decision feedback equalization is carried out by recursive least square algorithm. Due to the low precision and poor performance of traditional equalization methods, the introduction of deep learning can significantly improve channel transmission performance. Therefore, taking the performance of traditional methods as the benchmark, channel equalization algorithms based on deep learning networks (deep neural networks, convolutional neural networks and long and short-term memory networks) are compared and studied. Firstly, the underwater acoustic channel simulation software Bellhop is used to construct the deep learning data source. Then, the deep learning network is trained and the channel equalization algorithm simulation experiments of three neural networks are carried out. The results show that the channel equalization algorithm based on convolutional neural network has the best performance and the bit error rate is the lowest when the input sequence is2000symbols and the channel length is equivalent to 3.65 times.
2022,44(23): 123-127 收稿日期：2021-11-12
DOI：10.3404/j.issn.1672-7649.2022.23.024
分类号：TN911.7
基金项目：国家重点研发计划（2016YFB0501804）
作者简介：段晨阳(1986-)，女，硕士，研究方向为水声信道均衡技术
参考文献：
[1] 刘雪亭. 水声通信自适应均衡技术研究[J]. 舰船科学技术, 2016, 38(8):109-111
LIU Xue-ting. Research on underwater communication adaptive equalization[J]. Ship Science and Technology, 2016, 38(8):109-111
[2] 韩明秋, 吕振肃, 段磊, 等. 基于RLS的自适应MIMO-DFE算法[C].//2005中国控制与决策学术年会论文集, 2005:1527-1530.
HAN Ming-qiu, LV Zhen-su, DUAN Lei, et al. Adaptive multi-output decision feedback equalizer algorithm based on RLS[C]. Proceedings of 2005 Chinese and Decision Conference, 2005:1527-1530.
[3] 娄松涛. 基于神经网络的舰船通信网络信道均衡控制研究[J]. 舰船科学技术, 2021, 43(18):148-150
[4] CHENG X, LIU D, WANG C, et al. Deep learning-based channel estimation and equalization scheme for FBMC/OQAM Systems[J]. IEEE Wireless Communication Letters, 2019(99):1-1
[5] BHAGWATKAR S A, PATIL B P, SATPUTE B S. Performance of MMSE channel equalization for MIMO OFDM system[C]//2016 International Conference on Computing Communication Control and automation (ICCUBEA), IEEE, 2016.
[6] J C, Patra, R N, et al. Nonlinear channel equalization for QAM signal constellation using artificial neural networks[J]. IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics:a publication of the IEEE Systems, Man, and Cybernetics Society, 1999.
[7] WANG Y, ZHANG H, SANG Z, et al. Modulation classification of underwater communication with deep learning network[J]. Computational Intelligence and Neuroscience, 2019, 2019:1-12
[8] ZHANG Y, LI J, ZAKHAROV Y, et al. Deep learning based underwater acoustic OFDM communications[J]. Applied Acoustics, 2019, 154(NOV.):53-58
[9] 赵亮, 王晓峰, 袁逸涛. 基于深度卷积神经网络的船舶识别方法研究[J]. 舰船科学技术, 2016, 38(15):119-123
ZHAO Liang, WANG Xiao-feng, YUAN Yi-ta. Research on ship recognition method based on deep convolutional neural network[J]. Ship Science and Technology, 2016, 38(15):119-123
