深海不确定环境条件下远程水声通信性能分析与快速预报

陈香梅; 台玉朋; 王海斌; 胡承昊; 汪俊; 王迪雅

doi:10.11999/JEIT251244

深海不确定环境条件下远程水声通信性能分析与快速预报

doi: 10.11999/JEIT251244 cstr: 32379.14.JEIT251244

陈香梅^{1, 2},
台玉朋¹,
王海斌^1, ,,
胡承昊¹,
汪俊¹,
王迪雅¹

1.
中国科学院声学研究所声学与海洋信息实验室北京 100190
2.
中国科学院大学北京 100049

基金项目: 国家自然科学基金 (62301551)

详细信息

作者简介:
陈香梅：女，博士生，研究方向为水声通信

台玉朋：男，研究员，研究方向为水声通信

王海斌：男，研究员，研究方向为水声通信

胡承昊：男，助理研究员，研究方向为水声通信

汪俊：男，研究员，研究方向为水声通信

王迪雅：女，博士后，研究方向为水声通信

通讯作者:
王海斌　whb@mail.ioa.ac.cn

中图分类号: TN929.3
计量
- 文章访问数: 13
- HTML全文浏览量: 6
- PDF下载量: 1
- 被引次数: 0
出版历程
- 修回日期: 2026-01-12
- 录用日期: 2026-01-12
- 网络出版日期: 2026-01-27

Performance Analysis and Rapid Prediction of Long-Range Underwater Acoustic Communications in Uncertain Deep-Sea Environments

1.
State Key Laboratory of Acoustics and Marine Information, Chinese Academy of Sciences, Beijing 100190, China
2.
University of Chinese Academy of Sciences, Beijing 100049, China

Funds: The National Natural Science Foundation of China (62301551)

摘要

摘要: 在复杂且动态变化的海洋环境中，通信性能起伏显著且难以预估，传统依赖反馈链路进行信道状态估计与参数调整的方法难以适用于深海远程水声通信。为此，该文提出一种基于深度学习声场不确定预估的水声通信性能分析与快速预报方法，在无反馈条件下实现通信参数与信道状态的高效匹配。该方法基于深度学习快速预测的传播损失概率分布，构建了从传播损失到信噪比，再到统计信道容量与中断容量的链式映射模型，实现环境不确定性与通信性能的量化映射。进一步结合典型深海单载波通信系统在特定信道条件下的链路性能与传播损失的统计特性，提出通信“速率—可靠性”预报方法，评估不同速率下的可靠通信概率，从而为复杂动态环境下的系统参数匹配提供依据。海上试验结果表明，所提方法在复杂信道环境下对通信“速率—可靠性”的预报与实测结果高度一致：会聚区与影区各速率点上的可靠概率偏差分别为0.9%～4%和1%～9%；以90%可靠通信概率为阈值时，预报的最大可靠速率与实测结果一致，验证了该方法在深海远程水声通信中的准确性和实用性。
- 水声通信 /
- 性能预报 /
- 深度学习 /
- 不确定性 /
- 信道容量
Abstract: Objective In complex and dynamically varying deep-sea environments, underwater acoustic communication performance exhibits significant fluctuations. Feedback-based methods for channel state estimation and parameter adaptation become impractical in long-range deep-sea communications, as platform constraints hinder the establishment of reliable feedback channels and the slow propagation speed of sound introduces substantial feedback latency. In typical long-range underwater acoustic communication scenarios, the dynamic variability of the ocean environment is often ignored, and communication parameters are selected heuristically. As a result, these parameters are frequently mismatched with actual channel conditions, leading to communication failures or reduced efficiency. Therefore, it is essential to develop predictive methodologies capable of assessing communication performance in advance and enabling feed-forward adjustment of system parameters. The primary objective of this paper is to develop a deep-learning-based framework for performance analysis and rapid prediction of long-range underwater acoustic communications under uncertain deep-sea environmental conditions, thereby enabling efficient and reliable parameter–channel matching without relying on feedback. Methods A feed-forward underwater acoustic communication performance analysis and rapid forecasting method enabled by deep-learning-based sound-field uncertainty prediction is proposed in this paper. First, a neural network is employed to estimate the probability distributions of transmission loss (TL PDFs) at the receiver under dynamically varying ocean environments. Based on the predicted TL PDFs, the corresponding probability distributions of the signal-to-noise ratio (SNR PDFs) is derived, forming the basis for evaluating communication performance without relying on real-time feedback. Subsequently, the statistical channel capacity and outage capacity in dynamic environments are analyzed, enabling the characterization of the theoretical upper limits of communication rates in dynamic environments. Finally, by integrating the SNR distribution with the bit-error-rate characteristics of a representative deep-sea single-carrier communication system under specific channel, a rate–reliability prediction model is developed. This model estimates the probability of reliable communication at various data rates, offering a practical tool for forecasting link performance in highly dynamic and feedback-limited underwater acoustic scenarios. Results and Discussions The proposed method is validated using both simulation data and sea trial data The TL PDFs predicted by the deep learning model exhibit strong agreement with traditional Monte Carlo(MC) method across multiple receiver locations (Fig.6). Under identical computational settings, the deep-learning-based prediction of the TL PDFs reduces the computation time by 2 to 3 orders of magnitude compared with MC method. The chained mapping from TL PDFs to SNR PDFs and further to channel capacity metrics accurately captures the probabilistic nature of communication performance under uncertain environmental conditions (Fig.7 and 8). The rate–reliable communication probability curves derived from the TL PDFs predicted by the deep learning model are highly consistent with the results obtained using the MC-based approach. In the high sound-intensity region, the prediction errors for the reliable communication probabilities across different data rates range from 0.1% to 3% compared with the MC-based results, while in the low sound-intensity region the errors are approximately 0.3% to 5% (Fig.12).Furthermore, the results from the sea trials indicate that the predicted rate–reliability performance is in good agreement with the measured data. In the convergence zone, the deviations between the predicted and measured reliability probabilities at each data rate range from 0.9% to 4%, while in the shadow zone, the deviations range from 1% to 9%(Fig.18). Moreover, under a 90% reliability requirement, the maximum achievable reliable communication rates predicted by the method are consistent with the measurements in both the convergence zone and shadow zone, demonstrating the accuracy and practical applicability of the proposed approach in complex channel environments. Conclusions A deep-learning-based framework for performance analysis and rapid prediction of long-range underwater acoustic communications under uncertain deep-sea environments is proposed and validated. The framework establishes a chained mapping from environmental parameters to TL PDFs, SNR PDFs, and communication performance metrics, enabling quantitative assessment of capacity under dynamic ocean conditions. Accurate prediction of communication “rate–reliability” profiles is enabled by the integration of the performance of a representative deep-sea single-carrier communication system under specific channel conditions and probabilistic propagation characteristics, providing guidance for parameter selection without requiring feedback. Sea trial results confirm the high consistency between predicted and measured outcomes, demonstrating the method’s accuracy and practical applicability. The proposed approach provides a new technical pathway for feed-forward performance analysis, dynamic system adaptation of deep-sea long-range underwater acoustic communication systems, and it can be extended to other communication scenarios in dynamic ocean environments.
- Underwater Acoustic Communication /
- Performance Forecasting /
- Deep Learning /
- Uncertainty /
- Channel Capacity

HTML全文

图 1 基于声场不确定性预估的通信前馈式性能分析和快速预报方法

下载: 全尺寸图片幻灯片

图 2 基于DL的声场不确定性快速预测

下载: 全尺寸图片幻灯片

图 3 DL模型结构

下载: 全尺寸图片幻灯片

图 4 用于测试的基础环境

下载: 全尺寸图片幻灯片

图 5 基础环境对应的TL场

下载: 全尺寸图片幻灯片

图 6 两个方法在不同接收点得到的TL-PDF

下载: 全尺寸图片幻灯片

图 7 不同接收深度和距离处统计信道容量

下载: 全尺寸图片幻灯片

图 8 不同深度下统计容量和中断容量随距离的变化

下载: 全尺寸图片幻灯片

图 9 不同接收点处的仿真信道

下载: 全尺寸图片幻灯片

图 10 不同接收点处的SNR-PDF

下载: 全尺寸图片幻灯片

图 11 不同接收点处在各调制编码方案下的误码率曲线

下载: 全尺寸图片幻灯片

图 12 不同接收点处可靠通信概率

下载: 全尺寸图片幻灯片

图 13 实验海区的地形、声速剖面和沉积层参数

下载: 全尺寸图片幻灯片

图 14 声源和接收点在TL场中的位置

下载: 全尺寸图片幻灯片

图 15 两个接收点处的TL-PDF和SNR-PDF

下载: 全尺寸图片幻灯片

图 16 A点处的信道和误码率曲线

下载: 全尺寸图片幻灯片

图 17 B点处的信道和误码率曲线

下载: 全尺寸图片幻灯片

图 18 两个接收点处不同通信速率的可靠通信概率

下载: 全尺寸图片幻灯片

表 1 通信系统采用的调制编码方案

方案m	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
调制阶数	2	2	2	2	2	4	4	4	4	8	8	8	16	16	16
编码码率	1/4	1/2	2/3	3/4	5/6	1/2	2/3	3/4	5/6	2/3	3/4	5/6	2/3	3/4	5/6
速率(bps)	17	34	45	50	56	67	89	100	112	135	150	167	177	200	222

下载: 导出CSV

表 2 试验参数

试验参数	接收机采样频率 (kHz)	信号带宽 (Hz)	中心频率 (Hz)	通信距离 (km)	符号宽度 (ms)	发射声源级 (dB)	发射深度 (m)	信号长度 (bit)	通信速率 (bps)
取值	4	100	500	72.6km	15	190	500	192	20、25、 50、100、200

下载: 导出CSV

参考文献(22)

[1]	KILFOYLE D B and BAGGEROER A B. The state of the art in underwater acoustic telemetry[J]. IEEE Journal of Oceanic Engineering, 2000, 25(1): 4–27. doi: 10.1109/48.820733.
[2]	STOJANOVIC M and PREISIG J. Underwater acoustic communication channels: Propagation models and statistical characterization[J]. IEEE Communications Magazine, 2009, 47(1): 84–89. doi: 10.1109/mcom.2009.4752682.
[3]	PREISIG J C. Performance analysis of adaptive equalization for coherent acoustic communications in the time-varying ocean environment[J]. The Journal of the Acoustical Society of America, 2005, 118(1): 263–278. doi: 10.1121/1.1907106.
[4]	杨健敏, 王佳惠, 乔钢, 等. 水声通信及网络技术综述[J]. 电子与信息学报, 2024, 46(1): 1–21. doi: 10.11999/JEIT230424. YANG Jianmin, WANG Jiahui, QIAO Gang, et al. Review of underwater acoustic communication and network technology[J]. Journal of Electronics & Information Technology, 2024, 46(1): 1–21. doi: 10.11999/JEIT230424.
[5]	罗亚松, 许江湖, 胡洪宁, 等. 正交频分复用传输速率最大化自适应水声通信算法研究[J]. 电子与信息学报, 2015, 37(12): 2872–2876. doi: 10.11999/JEIT150440. LUO Yasong, XU Jianghu, HU Hongning, et al. Research on self-adjusting OFDM underwater acoustic communication algorithm for transmission rate maximization[J]. Journal of Electronics & Information Technology, 2015, 37(12): 2872–2876. doi: 10.11999/JEIT150440.
[6]	ZHANG Yonglin, TAI Yupeng, WANG Diya, et al. Online machine learning-based channel estimation for underwater acoustic communications[J]. The Journal of the Acoustical society of American, 2024, 155(S3): A88–A89. doi: 10.1121/10.0026909.
[7]	HU Yunfeng, TAO Jun, and TONG Feng. Estimation of time-varying underwater acoustic channels via an improved sparse adaptive orthogonal matching pursuit algorithm[J]. Applied Acoustics, 2025, 233: 110624. doi: 10.1016/j.apacoust.2025.110624.
[8]	刘志勇, 金子皓, 杨洪娟, 等. 基于深度学习的水声信道联合多分支合并与均衡算法[J]. 电子与信息学报, 2024, 46(5): 2004–2010. doi: 10.11999/JEIT231196. LIU Zhiyong, JIN Zihao, YANG Hongjuan, et al. Deep learning-based joint multi-branch merging and equalization algorithm for underwater acoustic channel[J]. Journal of Electronics & Information Technology, 2024, 46(5): 2004–2010. doi: 10.11999/JEIT231196.
[9]	刘凇佐, 韩雪, 马璐, 等. 基于排序码本的水声自适应OFDM通信中信道状态信息反馈研究[J]. 电子与信息学报, 2024, 46(5): 2095–2103. doi: 10.11999/JEIT230878. LIU Songzuo, HAN Xue, MA Lu, et al. Research on channel state information feedback in underwater acoustic adaptive OFDM communication based on sequenced codebook[J]. Journal of Electronics & Information Technology, 2024, 46(5): 2095–2103. doi: 10.11999/JEIT230878.
[10]	JING Lianyou, DONG Chaofan, HE Chengbing, et al. Adaptive modulation and coding for underwater acoustic OTFS communications based on meta-learning[J]. IEEE Communications Letters, 2024, 28(8): 1845–1849. doi: 10.1109/lcomm.2024.3418192.
[11]	HUANG Jianchun and DIAMANT R. Adaptive modulation for long-range underwater acoustic communication[J]. IEEE Transactions on Wireless Communications, 2020, 19(10): 6844–6857. doi: 10.1109/twc.2020.3006230.
[12]	BUSACCA F, GALLUCCIO L, PALAZZO S, et al. Adaptive versus predictive techniques in underwater acoustic communication networks[J]. Computer Networks, 2024, 252: 110679. doi: 10.1016/j.comnet.2024.110679.
[13]	笪良龙, 过武宏, 赵建昕, 等. 海洋-声学耦合模式捕捉水声环境不确定性[J]. 声学学报, 2015, 40(3): 477–486. doi: 10.15949/j.cnki.0371-0025.2015.03.016. DA Lianglong, GUO Wuhong, ZHAO Jianxin, et al. Capture uncertainty of underwater environment by ocean-acoustic coupled model[J]. Acta Acustica, 2015, 40(3): 477–486. doi: 10.15949/j.cnki.0371-0025.2015.03.016.
[14]	LV Zhichao, DU Libin, LI Huming, et al. Influence of temporal and spatial fluctuations of the shallow sea acoustic field on underwater acoustic communication[J]. Sensors, 2022, 22(15): 5795. doi: 10.3390/s22155795.
[15]	GAO Fei, XU Fanghua, LI Zhenglin, et al. Acoustic propagation uncertainty in internal wave environments using an ocean-acoustic joint model[J]. Chinese Physics B, 2023, 32(3): 034302. doi: 10.1088/1674-1056/ac89dc.
[16]	FENG Xiao, CHEN Cheng, and YANG Kunde. Fast estimation algorithm of sound field characteristics under the disturbance of sound speed profile in the marine environment[J]. Ocean Engineering, 2024, 297: 117197. doi: 10.1016/j.oceaneng.2024.117197.
[17]	KHAZAIE S, WANG Xun, KOMATITSCH D, et al. Uncertainty quantification for acoustic wave propagation in a shallow water environment[J]. Wave Motion, 2019, 91: 102390. doi: 10.1016/j.wavemoti.2019.102390.
[18]	LEE B M, JOHNSON J R, and DOWLING D R. Predicting acoustic transmission loss uncertainty in ocean environments with neural networks[J]. Journal of Marine Science and Engineering, 2022, 10(10): 1548. doi: 10.3390/jmse10101548.
[19]	CHEN Xiangmei, LI Chao, WANG Haibin, et al. A spatially informed machine learning method for predicting sound field uncertainty[J]. Journal of Marine Science and Engineering, 2025, 13(3): 429. doi: 10.3390/jmse13030429.
[20]	GIBBS A L and SU F E. On choosing and bounding probability metrics[J]. International Statistical Review, 2002, 70(3): 419–435. doi: 10.2307/1403865.
[21]	WENZ G M. Acoustic ambient noise in the ocean: Spectra and sources[J]. The Journal of the Acoustical Society of America, 1962, 34(12): 1936–1956. doi: 10.1121/1.1909155.
[22]	JIANG Dongge, LI Zhenglin, QIN Jixing, et al. Characterization and modeling of wind-dominated ambient noise in South China Sea[J]. Science China Physics, Mechanics & Astronomy, 2017, 60(12): 124321. doi: 10.1007/s11433-017-9088-5.