基于CNN+GRU的网络恶意流量检测算法
Network Malicious Traffic Detection Algorithm Based on CNN+GRU
高新成 1魏壮壮 2王莉利 2李林旭2
作者信息
- 1. 东北石油大学现代教育技术中心,黑龙江 大庆 163318
- 2. 东北石油大学计算机与信息技术学院,黑龙江 大庆 163318
- 折叠
摘要
针对网络恶意流量检测精确度和效率低等问题,提出了一种基于CNN+GRU算法的网络异常流量检测模型(CN-RU).模型使用卷积神经网络和门控循环单元来分别自动化提取流量的空间和时间特征,全方位的收集网络流量特征.模型使用多个小卷积核和少参数的门控循环单元来准确提取流量特征的同时减小模型参数,达到提高检测精度与效率的目的.实验使用ISCX IDS2012、CIC-IDS2017、UNSW-NB15 三种数据集进行效果评估,对比不同算法的网络流量检测模型,实验结果表明所提出的CNN+GRU结构模型解决了神经网络模型梯度消失问题的同时大幅度提高准确率和检测效率.模型具有较高的应用价值,在网络安全管理应用上有更好的普适性.
Abstract
In order to solve the problems that the low accuracy and efficiency of network malicious traffic detec-tion,this paper proposes a network abnormal traffic detection model based on CNN+GRU algorithm(CN-RU).The model uses convolutional neural network and gated recurrent unit to automatically extract the spatial and temporal characteristics of traffic respectively and collect the network traffic characteristics in an all-around way.The model u-ses multiple small convolution kernels and gated recurrent units with few parameters to accurately extract traffic fea-tures and reduce model parameters to achieve the purpose of improving detection accuracy and efficiency.In the ex-periment,ISCX IDS2012,CIC-IDS2017 and UNSW-NB15 were used to evaluate the effect,and network traffic detec-tion models with different algorithms were compared.The experimental results show that the CNN+GRU structure model proposed in this paper solves the problem of gradient disappearance in the neural network model and greatly improves the accuracy and detection efficiency.The model has higher application value and better universality in net-work security management.
关键词
流量检测/特征选择/卷积神经网络/门控循环单元/注意力机制Key words
Malicious traffic detection/Feature selection/Convolutional neural network/Gated recurrent unit/At-tentional mechanism引用本文复制引用
基金项目
国家自然科学基金(61702093)
黑龙江省自然科学基金(F2018003)
东北石油大学引导性创新基金(2020YDL-03)
出版年
2024
NETL
NSTL
万方数据