【作者】 杨光；

【导师】 何元安；

【作者基本信息】 哈尔滨工程大学 ， 水声工程， 2022， 博士

【摘要】 水声信道具有可用频带有限,严重的多径效应和多普勒效应以及信道结构随时间变化等特点,这些特点使得水声通信比无线电通信研究更加复杂。为了实现稳健的水声通信性能,通常使用扩频通信体制。直接序列扩频是水声通信中常用的高可靠性通信技术,然而本就有限的水下带宽使直扩频带利用率低的问题愈发凸显。如何在包含众多随机性因素的水声信道中,充分利用有限带宽,在保证水声扩频通信鲁棒性的前提下,面对不同水下通信场景灵活调整系统设计是具有挑战性的问题。本文使用M元扩频代替直扩来提高系统频谱效率和能量效率,并对M元扩频能量检测、双差分编码、多普勒估计和基于判决反馈的时域均衡等关键技术进行了理论分析和软件仿真。以此为基础,针对小多普勒下移动水声通信、变速运动下水声通信、深海远程移动水声通信等问题展开研究,并分别通过哈尔滨、威海和三亚的外场试验进行验证。针对定点通信和小于3m/s匀速相对运动的水声通信场景中,接收数据处理实时性和系统硬件低复杂度的需求,本文提出双差分M元扩频水声通信方案,利用双差分编码消除小多普勒造成的信号相位和频率误差;使用M元扩频相位调制方式,充分利用不同扩频序列的弱互相关性以及尖锐的自相关特性,使每个扩频符号携带更多数据信息,在保证扩频系统通信可靠性的前提下成倍提高直扩系统的通信速率;载波同步和码位同步技术的应用使得系统在未进行多普勒估计和补偿的情况下实现了可靠通信。松花江水声通信试验验证了双差分M元扩频在通信速率高于直扩系统时具有更低的误码率。针对水声信道中多普勒时变性较强的通信场景,接收信号在更短的时间间隔进行多普勒估计、跟踪和符号同步的需求,本文提出优化的逐符号多普勒估计和同步的M元扩频水声通信系统。利用扩频信号时间带宽积大,多普勒敏感的特性,以符号间隔在多普勒域和扩频码域两个维度进行相关搜索,同时实现当前符号的多普勒估计和解扩过程。本文在仿真和试验中发现正交M元扩频系统的多普勒承受能力较强,较低的多普勒估计精度并不影响通信性能。为了削减庞大的相关运算量,本文研究三个优化步骤,即扩大多普勒搜索步长,启发式搜索,以及基于马尔可夫过程的假设。通过优化,将逐符号多普勒估计的接收处理运算量削减至原来的1.24%,只比常规M元扩频系统解扩的相关运算量多了一倍。威海移动通信试验中,本文方法在加速度最高达到47.2m/s²的变化速度条件下实现了4km通信距离低于10^-3的误码率。此外,本文研究一种提高多普勒估计精度的方法,以本地参考信号和接收信号的相关程度构造目标函数,在缩小后的搜索范围内使用粒子群算法寻找目标函数最大值。该方法是以最小二乘法最小化参考信号和接收信号误差构造目标函数的改进,仿真结果表明本文方法对多径和宽带多普勒具有更好的适应能力。针对长多径时延和非最小相位水声信道通信场景中,多径效应造成的严重扩频码片间干扰和符号间干扰抑制需求,本文根据信道时变性不同分别研究两种适用于M元扩频水声通信系统的时域均衡技术。在信道时变性较弱时,以扩频符号间隔更新判决反馈均衡器系数,为了降低反馈不正确判决导致的错误传播概率,本文研究M元扩频双向判决反馈均衡技术,以多一倍计算量为代价,利用正常模式和时反模式下的判决反馈均衡器输出噪声相关性低的特点提高均衡输出信噪比。在信道时变性较强的情况,研究M元扩频假设反馈均衡技术,并研究混合采用码片假设反馈和能量检测的M元扩频通信方法,在不同的噪声环境中优化选择合适的处理方法,有效拓展M元扩频系统在深远海信道噪声环境的适用范围。在三亚进行的103公里深海远程水声扩频通信试验中,M元假设反馈均衡方法在2.4m/s的相对运动速度下实现了少量数据的无误解码。所研究方法的有效性得到验证。更多还原

【Abstract】 Underwater acoustic（UWA）channel has the characteristics of limited available frequency band,serious multipath effect,Doppler effect and channel structure changing with time.These characteristics make the research of UWA communication more complex than radio communication.To achieve robust UWA communication performance,spread spectrum communication system is usually used.Direct sequence spread spectrum（DSSS）is a high reliability communication technique commonly used in UWA communication.However,due to the limited underwater bandwidth,the problem of low utilization of the DSSS band becomes more and more prominent.How to make full use of the restricted bandwidth in the UWA channel containing many random factors and flexibly adjust the system design in the face of different undersea communication scenarios on the premise of ensuring the robustness of UWA spread spectrum communication is a challenging problem.In this paper,M-ary spread spectrum（MSS）is used instead of DSSS to improve the spectral efficiency and energy efficiency of the system.The key techniques such as MSS energy detection,double differential coding,Doppler estimation and time-domain equalization based on decision feedback are theoretically analyzed and simulated.On this basis,mobile UWA communication under small Doppler,variable speed UWA communication and deepsea long range mobile UWA communication are studied,which are verified by trials in Harbin,Weihai and Sanya.Aiming at the requirements of real-time processing of received data and low complexity of system hardware in the UWA communication scene with fixed end or uniform relative speed less than 3m/s,a scheme of double differential M-ary spread spectrum（DD-MSS）UWA communication system is proposed in this paper.DD coding is used to eliminate the signal phase and frequency errors caused by a small Doppler,and MSS phase modulation mode is used to make full utilization of the weak cross-correlation and sharp autocorrelation characteristics of different spread spectrum sequences,so that each spreading symbol can carry more data.On the premise of ensuring the communication reliability of the spread spectrum system,the communication rate of the DSSS system is doubled.The application of carrier synchronization and code synchronization makes the system realize reliable communication without Doppler estimation,and compensation.The experiment of Songhua River UWA communication verifies that the DD-MSS has a lower bit error rate when the communication rate is higher than that of the DSSS.Aiming at the communication scenario with obvious Doppler time variation in the UWA channel and the requirements of Doppler estimation,tracking and symbol synchronization of the received signal in a shorter time interval,an optimized MSS UWA communication system with a symbol-by-symbol Doppler estimation and synchronization is proposed in this paper.Using the characteristics of large time bandwidth product and Doppler sensitivity of the spread spectrum signal,the correlation search is carried out in two dimensions of Doppler domain and spread spectrum code domain with symbol interval,and the Doppler estimation and de-spreading process of current symbols are realized at the same time.In the simulation and experiment,it is found that the orthogonal MSS system has a strong Doppler tolerance ability,and the low Doppler estimation accuracy does not affect the communication performance.To reduce the huge amount of related operations,this paper proposes three optimization steps,namely,expanding the Doppler search step,heuristic search,and the assumption based on Markov process.Through these optimizations,the amount of receiving and processing operation of the symbol-by-symbol Doppler estimation is reduced to about 1.24%,which is only twice as much as that of de-spreading in a conventional MSS system.In Weihai mobile communication experiment,this method realizes the bit error rate of 4km communication distance less than 10^-3 with the maximum acceleration of 47.2m/s².In addition,this paper proposes a method to improve the accuracy of Doppler estimation,constructs the objective function based on the correlation between the local reference signal and the received signal,and uses the particle swarm optimization algorithm to find the maximum value of the cost function in the reduced search range.This method is an improvement of constructing the goal function by minimizing the error of reference signal and standard signal by the least square method.The simulation results show that this method has a better adaptability to multipath and large Doppler.Aiming at the suppression requirements of serious inter chip interference and inter symbol interference caused by multipath effect in the communication scenario of long time delay and non minimum phase UWA channel,this paper presents two time-domain equalization techniques suitable for the MSS UWA communication system according to the different time variation of the channel.When the channel time variation is weak,the decision feedback equalizer（DFE）coefficients are updated at the spread spectrum symbol interval.To reduce the error propagation probability caused by incorrect feedback decisions,this paper proposes the DSSS bidirectional DFE technique.Which takes an advantage of the low correlation of the DFE output noise in the normal mode and the time reversal mode to improve the output signal-to-noise ratio（SNR）of equalization processing at the cost of doubling the amount of calculations.In the time varied UWA channel,the MSS hypothetical feedback equalization is studied,and the MSS communication method hybrid CHF and energy detection is proposed.The appropriate processing method is optimized in different noise environments,effectively expanding the application scope of the MSS system in the deep sea long range UWA channel.In the 103km deep-sea long-range UWA spread spectrum communication experiment conducted in Sanya,the MSS hypothesis feedback equalization realizes the error-free decoding of a small amount of data at the relative motion speed of 2.4m/s.The effectiveness of the proposed method is verified.更多还原