【作者】 李岩；

【导师】 刘大利； 王岁；

【作者基本信息】 天津工业大学 ， 电子信息（专业学位）， 2025， 硕士

【摘要】 水声通信的研究伴随着无线通信技术的发展和人类对海洋开发需求的提升越来越受到重视。水声信道环境复杂且恶劣,具有多径效应严重、带宽有限等特点。正交频分复用技术(OFDM)凭借其良好的抗多径效应能力以及频谱效率高等特点在水声通信领域获得广泛应用,然而水声通信中由于OFDM技术对同步非常敏感,当前的符号定时同步算法中存在着定时度量函数正确定时点处有“平台效应”、副峰占据主峰比例较大的问题,其次当前的一些频域信道估计算法无法利用水声信道的稀疏性,估计性能较差,且信道估计的时效性得不到保障,当信道估计的信息反馈到发送端时水声信道可能已经发生较大变化。针对以上问题,本文提出了改进型符号定时同步算法与基于信道校正的改进型估计算法,主要研究工作如下:(1)对三种传统符号定时同步算法S&C算法、Minn算法以及Park算法进行细致分析并进行仿真,针对定时同步性能不足的问题,对算法训练序列的构造、定时度量函数M(d)以及算法复杂度方面分别做出改进,提出了改进型符号定时同步算法,并且与Park算法仿真比较,结果表明,改进型算法的定时同步的准确度在相同信噪比下更高。(2)研究了非自适应信道的估计算法,对比了频域信道估计最小二乘(LS)算法与压缩感知正交匹配追踪(OMP)算法、正则化正交匹配追踪(ROMP)算法的估计性能,其次对自适应信道的估计算法进行研究,分析时域校正的原理以及优势,接着分析了几种经典的校正算法,综合考虑算法的性能与计算复杂度,确定了递归最小二乘(RLS)算法更有优势。综合分析提出了基于信道校正的改进型估计算法,并进行仿真分析,结果表明改进型估计算法较ROMP算法性能更有优势。(3)基于对符号定时同步与信道估计两种OFDM水声通信重要技术的分析,搭建了基于BELLHOP水声信道的仿真实验,对所提出的改进型符号定时同步算法、基于信道校正的改进型估计算法进行验证分析。通过以上研究,改进型符号定时同步算法相比于传统算法提升了定时同步精度,降低了系统误码率。基于信道校正的改进型估计算法在初步的信道估计之后,根据追踪到的信道变化对估计的信道信息作进一步的校正处理,提升估计精度,解决了时效性不足的问题。更多还原

【Abstract】 Research on underwater acoustic communication has garnered increasing attention with the advancement of wireless communication technologies and the growing demand for marine exploration.The underwater acoustic channel is characterized by a complex and harsh environment,featuring severe multipath effects and limited bandwidth.Orthogonal Frequency Division Multiplexing(OFDM)has been widely adopted in underwater acoustic communication due to its robustness against multipath effects and high spectral efficiency.However,OFDM-based systems in underwater environments are highly sensitive to synchronization.Existing symbol timing synchronization algorithms suffer from critical limitations,including a"plateau effect"at the correct timing point in the timing metric function and a significant proportion of subsidiary peaks relative to the main peak,leading to timing synchronization deviations.Furthermore,conventional frequency-domain channel estimation algorithms fail to exploit the sparsity of underwater acoustic channels,resulting in poor estimation performance and insufficient timeliness.When estimated channel information is fed back to the transmitter,the channel state may have already changed substantially.To address these challenges,this paper proposes an improved symbol timing synchronization algorithm and a channel calibration-based enhanced channel estimation algorithm.The main contributions are as follows:(1)A comprehensive analysis and simulation of three traditional symbol timing synchronization algorithms—Schmidl&Cox(S&C),Minn,and Park algorithms—were conducted.To mitigate synchronization performance deficiencies,improvements were made in training sequence design,timing detection error function M(d),and algorithmic complexity.The proposed enhanced symbol timing synchronization algorithm was compared with the Park algorithm via simulations,demonstrating higher accuracy under identical signal-to-noise ratio(SNR)conditions.(2)Non-adaptive channel estimation algorithms were investigated,including the frequency-domain Least Squares(LS)algorithm and the compressed sensing-based Regularized Orthogonal Matching Pursuit(ROMP)algorithm.Adaptive channel estimation algorithms were further explored,emphasizing the principles and advantages of time-domain calibration.After evaluating classical calibration algorithms,the Recursive Least Squares(RLS)algorithm was identified as superior in balancing performance and computational complexity.A channel calibration-based improved estimation algorithm was proposed and validated through simulations,outperforming the ROMP algorithm in estimation accuracy.(3)A simulation platform based on the BELLHOP underwater acoustic channel model was established to evaluate the proposed improved symbol timing synchronization and channel estimation algorithms.Experimental results confirmed the effectiveness of the proposed methods.The improved symbol timing synchronization algorithm enhances synchronization accuracy and reduces system bit error rates compared to conventional approaches.The enhanced channel estimation algorithm incorporates post-calibration processing based on tracked channel variations,improving estimation precision and addressing timeliness limitations.更多还原