融合CNN-BiGRU和注意力机制的网络入侵检测模型

信息安全研究 ›› 2024, Vol. 10 ›› Issue (3): 202-.

融合CNN-BiGRU和注意力机制的网络入侵检测模型

杨晓文张健况立群庞敏

(中北大学计算机科学与技术学院太原030051)
(机器视觉与虚拟现实山西省重点实验室太原030051)
(山西省视觉信息处理及智能机器人工程研究中心太原030051)

出版日期:2024-03-23 发布日期:2024-03-08
通讯作者: 张健硕士.主要研究方向为深度学习、网络安全. 1724055732@qq.com
作者简介:杨晓文博士，副教授.主要研究方向为人工智能、网络安全. wenyang1314@nuc.edu.cn 张健硕士.主要研究方向为深度学习、网络安全. 1724055732@qq.com 况立群博士，教授.主要研究方向为人工智能与计算机视觉. kuang@nuc.edu.cn 庞敏博士，讲师.主要研究方向为计算机视觉、虚拟仿真. 28994896@qq.com

A Network Intrusion Detection Model Integrating CNN-BiGRU and Attention Mechanism

Yang Xiaowen, Zhang Jian, Kuang Liqun, and Pang Min#br#

#br#

(School of Computer Science and Technology, North University of China, Taiyuan 030051)
(Shanxi Key Laboratory of Machine Vision and Virtual Reality, Taiyuan 030051)
(Shanxi Province’s Vision Information Processing and Intelligent Robot Engineering Research Center, Taiyuan 030051)

Online:2024-03-23 Published:2024-03-08

摘要/Abstract

摘要： 为提高网络入侵检测模型特征提取能力和分类准确率，提出了一种融合双向门控循环单元(CNNBiGRU)和注意力机制的网络入侵检测模型.使用CNN有效提取流量数据集中的非线性特征；双向门控循环单元(BiGRU)提取数据集中的时序特征，最后融合注意力机制对不同类型流量数据通过加权的方式进行重要程度的区分，从而整体提高该模型特征提取与分类的性能.实验结果表明：其整体精确率比双向长短期记忆网络(BiLSTM)模型提升了2.25%.K折交叉验证结果表明：该模型泛化性能良好，避免了过拟合现象的发生，印证了该模型的有效性与合理性.

关键词: 网络入侵检测, 卷积神经网络, 双向门控循环单元, 注意力机制, 深度学习

Abstract: To enhance the feature extraction capabilities and classification accuracy of the network intrusion detection model, a network intrusion detection model integrating CNNBiGRU (Convolutional Neural NetworkBidirectional Gated Recurrent Unit) and attention mechanism is proposed. CNN is employed to effectively extract nonlinear features from traffic datasets,while BiGRU extracts timeseries features. The attention mechanism is then integrated to differentiate the importance of different types of traffic data through weighted means, thereby improvingthe overall performance of the model in feature extraction and classification. The experimental results indicate that the overall accuracy rate is 2.25% higher than that of the BiLSTM (Bidirectional Long ShortTerm Memory) model. Kfold crossvalidation results demonstrate that the proposed model's good generalization performance, avoiding the occurrence of overfitting phenomenon, and affirming its effectiveness and rationality.

Key words: network intrusion detection, convolutional neural network, bidirectional gated recurrent unit, attention mechanism, deep learning

中图分类号:

TP391

杨晓文, 张健, 况立群, 庞敏, . 融合CNN-BiGRU和注意力机制的网络入侵检测模型[J]. 信息安全研究, 2024, 10(3): 202-.

参考文献

［1］Saljoughi A S, Mehrvarz M, Mirvaziri H. Attacks and intrusion detection in cloud computing using neural networks and particle swarm optimization algorithms［J］. Emerging Science Journal, 2017, 1(4): 179191［2］Ahsan M K. Increasing the predictive potential of machine learning models for enhancing cybersecurity［D］. North Dakota State: North Dakota State University, 2021［3］Kim J, Shin Y, Choi E. An intrusion detection model based on a convolutional neural network［J］. Journal of Multimedia Information System, 2019, 6(4): 165172［4］Kim J, Kim J, Thu H L T, et al. Long short term memory recurrent neural network classifier for intrusion detection［C］ Proc of 2016 Int Conf on Platform Technology and Service (PlatCon). Piscataway, NJ: IEEE, 2016: 15［5］Fu Y, Du Y, Cao Z, et al. A deep learning model for network intrusion detection with imbalanced data［J］. Electronics, 2022, 11(6): 898［6］Krishnan D. Detection of denialofservice attacks using stacked LSTM networks［C］ Proc of Data Analytics and Management. Berlin: Springer, 2022: 229239［7］Koroniotis N, Moustafa N, Slay J. A new intelligent satellite deep learning network forensic framework for smart satellite networks［J］. Computers and Electrical Engineering, 2022, 99: 107745［8］Qazi E U H, Almorjan A, Zia T. A onedimensional convolutionalneural network (1DCNN) based deep learning system for network intrusion detection［J］. Applied Sciences, 2022, 12(16): 7986［9］舒豪, 王晨, 史崯. 基于BiLSTM和注意力机制的入侵检测［J］. 计算机工程与设计, 2020, 41(11): 30423046［10］Vinayakumar R, Soman K, Poornachandran P. Applying convolutional neural network for network intrusion detection［C］ Proc of 2017 Int Conf on Advances in Computing, Communications and Informatics (ICACCI). Piscataway, NJ: IEEE, 2017: 12221228［11］Liu Y, Liu S, Zhao X. Intrusion detection algorithm based on convolutional neural network［J］. Transaction of Beijing Institute of Technology, 2017, 37(12): 12711275［12］王伟. 基于深度学习的网络流量分类及异常检测方法研究［D］. 合肥: 中国科学技术大学, 2018［13］Zainel H, Koak C. LAN intrusion detection using convolutional neural networks［J］. Applied Sciences, 2022, 12(13): 6645［14］Szegedy C, Vanhoucke V, Ioffe S, et al. Rethinking the inception architecture for computer vision［C］ Proc of the IEEE Conf on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2015: 19［15］Woo S, Park J, Lee JY, et al. Cbam: Convolutional block attention module［C］ Proc of the European Conf on Computer Vision (ECCV). Berlin: Springer, 2018: 319［16］Do N Q, Selamat A, Lim K C, et al. An improved ensemble deep learning model based on CNN for malicious website detection［C］ Proc of Int Conf on Industrial, Engineering and Other Applications of Applied Intelligent Systems. Berlin: Springer, 2022: 497504［17］Seth S, Singh G, Kaur K. Smart intrusion detection system using deep neural network gated recurrent unit technique［C］ Proc of ICCCE 2021. Berlin: Springer, 2022: 285293［18］Zhao P, Fan Z, Cao Z, et al. Intrusion detection model using temporal convolutional network blend into attention mechanism［J］. International Journal of Information Security and Privacy, 2022, 16(1): 120［19］何春蓉, 朱江. 基于注意力机制的GRU神经网络安全态势预测方法［J］. 系统工程与电子技术, 2020, 43(1): 258266［20］Mohammadpour L, Ling T C, Liew C S, et al. A survey of CNNbased network intrusion detection［J］. Applied Sciences, 2022, 12(16): 8162［21］Manoharan G, Sivakumar K. A study on outlier detection using hybrid convolution neural network and novel bidirectional gated recurrent unit (CNNBiGRU)［J］. PalArch’s Journal of Archaeology of EgyptEgyptology, 2020, 17(7): 48024808［22］Halbouni A, Gunawan T S, Habaebi M H, et al. CNNLSTM: Hybrid deep neural network for network intrusion detection system［J］. IEEE Access, 2022, 10: 9983799849

[1]	罗乐琦, 张艳硕, 王志强, 文津, 薛培阳, . 基于BERT模型的源代码漏洞检测技术研究[J]. 信息安全研究, 2024, 10(4): 294-.
[2]	赵荻, 尹志超, 崔苏苏, 曹中华, 卢志刚, . 基于图表示的恶意TLS流量检测方法[J]. 信息安全研究, 2024, 10(3): 209-.
[3]	张淑慧, 胡长栋, 王连海, 徐淑奖, 邵蔚, 兰田, . 基于GHM可视化和深度学习的恶意代码检测与分类[J]. 信息安全研究, 2024, 10(3): 216-.
[4]	江荣旺, 魏爽, 龙草芳, 杨明, . 基于增量学习的车联网恶意位置攻击检测研究[J]. 信息安全研究, 2024, 10(3): 277-.
[5]	黄灵, 何希平, 贺丹, 杨楚天, 旷奇弦, . 融合卷积神经网络和Transformer的人脸欺骗检测模型[J]. 信息安全研究, 2024, 10(1): 25-.
[6]	王耀辉, 王可, 宫良一, 付豫豪, 王跃达, 李婧, . 基于异构图的恶意域名检测方法研究[J]. 信息安全研究, 2023, 9(E1): 38-.
[7]	闫一非, 文斌, 张逢, . 基于图神经网络的智能合约源码漏洞检测[J]. 信息安全研究, 2023, 9(E1): 55-.
[8]	郑丽娜, 杜彦辉, . 基于深度学习的HTTP慢速DoS攻击检测研究[J]. 信息安全研究, 2023, 9(E1): 72-.
[9]	叶水欢, 葛寅辉, 陈波, 于泠, . 基于ELMoTextCNN的网络欺凌检测模型[J]. 信息安全研究, 2023, 9(9): 868-.
[10]	王志强, 王姿旖, 王庆德, 徐华福, . 基于LightGBM的区块链异常交易检测技术研究[J]. 信息安全研究, 2023, 9(9): 877-.
[11]	李敬. 基于卷积神经网络的加密代理流量识别方法[J]. 信息安全研究, 2023, 9(8): 722-.
[12]	张鹏飞. 基于机器学习的入侵检测模型对比研究[J]. 信息安全研究, 2023, 9(8): 739-.
[13]	杜林, 许传淇. 基于BERT的漏洞文本特征分类技术研究[J]. 信息安全研究, 2023, 9(7): 687-.
[14]	蒋明, 张宗凯, 刘熙尧, 郭标, 胡家馨, 张硕, . 基于多注意力机制的孪生网络图像隐写分析方法[J]. 信息安全研究, 2023, 9(6): 573-.
[15]	刘亦纯, 张光华, 宿景芳. 基于多级度量差值的神经网络后门检测方法[J]. 信息安全研究, 2023, 9(6): 587-.