针对探测声呐与通信声呐独立运行导致的频谱利用率低、相互干扰等问题，基于全共享体制和深度学习方法，提出水声正交啁啾复用（Orthogonal Chirp Division Multiplexing，OCDM）信号的联合声呐与通信（Sonar Communication，SonarCom）系统方案。在发射端，采用多载波调制技术发射抗干扰的OCDM共享波形；在接收端，构建融合多尺度卷积（Multi-Scale Convolutional Neural Network，MSCNN）结构和自注意力机制（Self-Attention，SA）的深度残差网络（Residual Network，ResNet）模型MSResNet，完全替代传统OCDM接收机的信号处理模块，充分训练后直接用于在线信息恢复和目标探测。模型精准捕捉到OCDM信号的关键特征，并在保证良好探测性能的前提下集成了通信功能。仿真结果表明，与相关方法相比，基于OCDM的MSResNet-SonarCom可以实现更低的误码率和更高的检测概率，并且对多普勒效应具有较强的鲁棒性。
Aiming at the problems of low spectrum utilization and mutual interference caused by the independent operation of detection sonar and communication sonar, a joint sonar and communication system solution of underwater acoustic orthogonal chirp division multiplexing signal is proposed based on full sharing system and deep learning method. At the transmitting end, multi-carrier modulation technology is used to transmit the anti-interference OCDM shared waveform; at the receiving end, a deep residual network model MSResNet that integrates the multi-scale convolutional neural network structure and the self-attention mechanism is constructed to completely replace the signal processing module of the traditional OCDM receiver. After sufficient training, it is directly used for online information recovery and target detection. The model accurately captures the key features of the OCDM signal and integrates the communication function under the premise of ensuring good detection performance. Simulation results show that compared with related methods, the MSResNet-SonarCom based on OCDM can achieve lower bit error rate and higher detection probability, and has strong robustness to the Doppler effect.
2025,47(18): 149-154 收稿日期：2024-12-20
DOI：10.3404/j.issn.1672-7649.2025.18.024
分类号：TN929.3
基金项目：陕西省教育厅科研计划项目（22JK0454）
作者简介：聂钰蓉（2000 – ），女，硕士，研究方向为水声通信
参考文献：
[1] 卢俊, 张群飞, 史文涛, 等. 探测通信一体化研究现状与发展趋势[J]. 信号处理, 2019, 35(9): 1484-1495.
LU J, ZHANG Q F, SHI W T, et al. Research status and development trend of detection and communication integration[J]. Signal Processing, 2019, 35(9): 1484-1495.
[2] YAN S, ZHOU X, HU J, et al. Low probability of detection communication: Opportunities and challenges[J]. IEEE Wireless Communications, 2019, 26(5): 19-25.
[3] WANG B, WANG Y, LI Y, et al. Underwater acoustic communications based on OCDM for internet of underwater things[J]. IEEE Internet of Things Journal, 2023, 10(24): 22128-22142.
[4] SHI Q, ZHANG T, YU X, et al. Waveform designs for joint radar-communication systems with OQAM-OFDM[J]. Signal Processing, 2022, 195: 108462.
[5] ZHU, XU X, TU X, et al. Anti-multipath orthogonal chirp division multiplexing for underwater acoustic communication[J]. IEEE Access, 2020, 8: 13305-13314.
[6] YIN J, MEN W, HAN X, et al. Integrated waveform for continuous active sonar detection and communication[J]. IET Radar, Sonar & Navigation, 2020, 14(9): 1382-1390.
[7] YAO R, WANG S, ZUO X, et al. Deep learning aided signal detection in OFDM systems with time-varying channels[C]//2019 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing (PACRIM). IEEE, 2019.
[8] WANG X, ZHOU J, YANG S, et al. Signal detection of ofdm systems based on convolutional neural network[C]//2022 41st Chinese Control Conference (CCC). IEEE, 2022.
[9] ZHENG S, CHEN S, YANG X. DeepReceiver: A deep learning-based intelligent receiver for wireless communications in the physical layer[J]. IEEE Transactions on Cognitive Communications and Networking, 2020, 7(1): 5-20.
[10] MEI R, WANG Z, CHEN X. CRNN-ResNet: combined CRNN and ResNet networks for OFDM receivers[J]. IEEE Transactions on Cognitive Communications and Networking, 2024, 10(4): 1294-1306.
[11] DE MAGISTRIS G, STINCO P, BATES J R, et al. Automatic object classification for low-frequency active sonar using convolutional neural networks[C]// MTS/IEEE OCEANS 2019 Seattle. IEEE, 2019.
[12] ZHANG W, FENG M, KRUNZ M, et al. Signal detection and classification in shared spectrum: A deep learning approach[C]//IEEE INFOCOM 2021-IEEE Conference on Computer Communications. IEEE, 2021.
[13] DOAN V S, HUYNH-THE T, KIM D S. Underwater acoustic target classification based on dense convolutional neural network[J]. IEEE Geoscience and Remote Sensing Letters, 2020, 19: 1-5.
[14] HOU S, JIAO D, DONG B, et al. Underwater inspection of bridge substructures using sonar and deep convolutional network[J]. Advanced Engineering Informatics, 2022, 52: 101.
