在船用网络流量中，持续性隐蔽威胁具有隐蔽性强、持续时间长等特点，传统检测方法难以检测这种长期依赖关系。为了提高深度检测的可靠性，设计基于GAN-LSTM（Generative Adversarial Networks-Long Short Term Memory Networks）的船用网络持续性隐蔽威胁深度检测方法。采用生成对抗网络根据持续性隐蔽威胁攻击特点生成接近真实船用网络的持续性隐蔽威胁攻数据样本。利用长短期记忆网络捕捉船用网络流量中的长期依赖关系，精准识别潜在威胁并输出深度检测结果。实验结果表明，生成样本与真实样本的相似度得分保持在0.9以上，证明了本文方法数据样本生成的质量较高。对于不同船用网络传输距离，攻击链完整度高于70%的阈值，说明本文方法的检测精度较高，能够为船用网络安全防护提供有力的技术支持。
In marine network traffic, persistent covert threats have the characteristics of strong concealment and long duration, and traditional detection methods are difficult to detect such long-term dependencies. In order to improve the reliability of deep detection, a GAN-LSTM (Generative Adversarial Networks-Long Short Term Memory networks) based method for continuous covert threat deep detection in marine networks is designed. Using long short-term memory networks to capture long-term dependencies in marine network traffic, accurately identify potential threats, and output deep detection results. The experimental results show that the similarity score between generated samples and real samples remains above 0.9, demonstrating the high quality of data sample generation using our method. For different ship network transmission distances, the attack chain integrity is above the threshold of 70%, indicating that the detection accuracy of this method is high and can provide strong technical support for ship network security protection.
2025,47(9): 175-179 收稿日期：2025-3-27
DOI：10.3404/j.issn.1672-7649.2025.09.030
分类号：TP391
基金项目：河南省科技攻关项目（24102210084）
作者简介：赵晓华（1990-），女，硕士，讲师，研究方向为智能信息处理及信息安全
参考文献：
[1] 张博文, 马国军, 王亚军. 基于边缘计算的船舶通信网络负载均衡研究[J]. 中国造船, 2024, 65(3): 122-134.
ZHANG B W, MA G J, WANG Y J. Research on load balancing of ship communication network based on edge computing[J]. China Shipbuilding, 2024, 65(3): 122-134.
[2] 吴芬. 改进卷积神经网络的舰船物联网安全风险估计[J]. 舰船科学技术, 2021, 43(14): 205-207.
WU F. Improving convolutional neural networks for ship IoT security risk estimation[J]. Ship Science and Technology, 2021, 43(14): 205-207.
[3] 卢金清. 基于大数据分析的船舶网络入侵检测系统[J]. 舰船科学技术, 2022, 44(7): 166-169.
LU J Q. Ship network intrusion detection system based on big data analysis[J]. Ship Science and Technology, 2022, 44(7): 166-169.
[4] 张晓玲, 李玮. 基于小波变换的舰船主干网络干扰信号防御算法[J]. 舰船科学技术, 2021, 43(18): 163-165.
ZHANG X L, LI W. Anti interference signal algorithm for ship backbone network based on wavelet transform[J]. Ship Science and Technology, 2021, 43(18): 163-165.
[5] 李瑛, 杨丽娟, 朱蓬华. 基于数据挖掘的船用通信网络异常行为分类和识别研究方法[J]. 舰船科学技术, 2023, 45(21): 181-184.
LI Y, YANG L J, ZHU P H. Research method for abnormal behavior classification and recognition of marine communication networks based on data mining[J]. Journal of Naval Science and Technology, 2023, 45(21): 181-184.
[6] 张震, 周一成, 田鸿朋. 基于空间特征和生成对抗网络的网络入侵检测[J]. 郑州大学学报(工学版), 2024, 45(6): 40-47.
ZHANG Z, ZHOU Y C, TIAN H P. Network intrusion detection based on spatial features and generative adversarial networks[J]. Journal of Zhengzhou University (Engineering Edition), 2024, 45(6): 40-47.
[7] 白万荣, 魏峰, 郑广远, 等. 基于TCN-BiLSTM的入侵检测算法研究[J]. 计算机科学, 2023, 50(S2): 941-948.
BAI W R, WEI F, ZHENG G Y, et al. Research on intrusion detection algorithm based on TCN-BiLSTM[J]. Computer Science, 2023, 50(S2): 941-948.
