电子学报 ›› 2022, Vol. 50 ›› Issue (3): 652-662.DOI: 10.12263/DZXB.20210183

• 学术论文 • 上一篇    下一篇

水声信道频域Turbo均衡的期望传播改进算法

姜斌1,2(), 唐禹1, 包建荣1, 唐向宏1, 刘超1()   

  1. 1.杭州电子科技大学通信工程学院,浙江 杭州 310018
    2.杭州电子科技大学电子信息学院,浙江 杭州 310018
  • 收稿日期:2021-01-28 修回日期:2021-04-19 出版日期:2022-03-25 发布日期:2022-03-25
  • 通讯作者: 姜斌,刘超
  • 作者简介:姜 斌 男,1980年生,浙江衢州人.现为杭州电子科技大学通信工程学院高级实验师,在职博士生.主要研究方向为无线通讯、信号处理、无线传感网络等.
    唐 禹 男,1997年生,安徽马鞍山人. 现为杭州电子科技大学通信工程学院硕士研究生,研究方向为水声Turbo迭代技术.E-mail:181080015@hdu.edu.cn
    包建荣(通讯作者) 男,1978年生,浙江杭州人.杭州电子科技大学教授、博导,研究方向为空间与海洋无线通信、协同信息论与编码、分布式多天线联合迭代检测等.
    唐向宏 男,1962年生,四川蓬溪人.杭州电子科技大学教授,主要研究方向为图像处理与传输、小波理论及应用及通信与信息系统等.E-mail:tangxh@hdu.edu.cn
    刘 超 男,1977年生,湖北武汉人.杭州电子科技大学副教授,主要研究方向为无线通信、计算机通信网等.
  • 基金资助:
    国家自然科学基金(U1809201);浙江省自然科学基金(LY20F010010);浙江省属高校基本科研业务费专项资金项目(GK20 9907299001-003)

Improved Expectation Propagation Algorithm of Frequency Domain Turbo Equalization for Underwater Acoustic Channels

JIANG Bin1,2(), TANG Yu1, BAO Jian-rong1, TANG Xiang-hong1, LIU Chao1()   

  1. 1.School of Communication Engineering,Hangzhou Dianzi University,Hangzhou,Zhejiang 310018,China
    2.School of Electronic Information,Hangzhou Dianzi University,Hangzhou,Zhejiang 310018,China
  • Received:2021-01-28 Revised:2021-04-19 Online:2022-03-25 Published:2022-03-25
  • Contact: JIANG Bin,LIU Chao

摘要:

针对水声信道带限,稀疏及长时延特性,并考虑水声实时高质量通信需求,提出了一种基于期望传播(Expectation Propagation, EP)的迭代信道估计和频域Turbo均衡(EP based Iterative Channel Estimation-Frequency of Domain Turbo Equalization, EP-ICE-FDTE)算法.其利用EP算法迭代估计传输符号的先验及后验分布,并将后验概率软映射符号提升信道估计精度.故该EP通过优化符号概率估计和信道估计,使Turbo均衡的干扰消除性能得到提高.此外,为了降低系统时延,还利用低复杂度频域Turbo均衡处理接收信号.仿真表明:所提EP-ICE-FDTE算法能实现更优的接收信号恢复效果.在静止和时变水声信道,该算法较目前较好的迭代信道估计-频域判决反馈的频域Turbo均衡(Iterative Channel Estimation-Frequency Domain Decision Feedback-Frequency Domain Turbo Equalization, ICE-FDDF-FDTE)算法,分别有3.4dB,1.3dB性能增益.

关键词: 水声通信, 期望传播算法, 软迭代信道估计, 频域Turbo均衡

Abstract:

According to the band-limited, sparse and long-delay characteristics of the underwater acoustic (UWA) channels, and considering the requirements for real-time high-quality UWA communication, an expectation propagation (EP) based iterative channel estimation and frequency domain turbo equalization (EP-ICE-FDTE) algorithm is proposed. The EP is used to iteratively estimate the a priori and a posterior distribution of the transmitted symbols, and theposterioriprobability soft-mapped symbols are used to improve the accuracy of channel estimation. Therefore, EP improves the performance of Turbo equalization interference cancellation by optimizing symbol probability and channel estimation. Simulations indicate that the proposed algorithm obtains better received signal recovery than those of traditional ones. The proposed EP-ICE-FDTE obtains 3.4dB, 1.3dB performance gains, when compared with those of the currently well iterative channel estimation-frequency domain decision feedback-frequency domain Turbo equalization (ICE-FDDF-FDTE) in static and time-varying UWA channels, respectively.

Key words: underwater acoustic communications, EP algorithm, soft iterative channel estimation, frequency domain Turbo equalization

中图分类号: