APP下载

基于深度学习的网络入侵防御技术研究

2020-12-23周路明郑明才

微型电脑应用 2020年11期
关键词:深度学习

周路明 郑明才

摘 要:随着互联网技术的快速发展,网络用户的数量激增,仅在国内就有着接近一半人口的用户。如此大规模的网络给网络攻击者带来了巨大的潜在利益,也给网络入侵的防御提出了更高的要求。传统的网络防御手段因其仅能针对特定的网络入侵行为进行甄别,无法智能化、动态化的应对复杂的网络入侵行为已经逐渐难以满足当下需求。因此,针对网络入侵防御的问题,提出了一种基于深度学习的入侵检测手段,并阐述了入侵防御系统的设计方法。首先,介绍了目前网络入侵防御所面临的严峻形势;其次,阐述了网络入侵检测与网络入侵防御中的框架性问题;再次,详细阐述了基于深度学习的入侵检测算法的设计方法,并阐述了入侵防御设计的要点,最后,入侵检测算法的有效性和准确性通过仿真进行了验证。仿真结果表明所设计的算法能够对复杂的入侵数据具有较高的威胁检测准确度,测试数据集对按照公式计算最终测得的检测率为95.22%和误报率为0.67%。

关键词:入侵防御;深度学习;神经网络;入侵检测

中图分类号:TP242.3

文献标志码:A

文章编号:1007-757X(2020)11-0093-05

Abstract:With the rapid development of Internet technologies, the number of network users has increased rapidly. In China, there are nearly half of the population of users. Such a large scale network has brought huge potential benefits to network attackers, and puts forward higher requirements for the defense of network intrusion. Traditional network defense means can only screen specific network intrusions and cannot deal with complex network intrusions intelligently and dynamically. Therefore, it is increasingly difficult to meet current needs. Aiming at the problem of network intrusion prevention, this paper proposes an intrusion detection method based on deep learning, and expounds the design method of intrusion prevention system. Firstly, the paper introduces the severe situation of network intrusion prevention. Secondly, the framework of network intrusion detection and network intrusion prevention is discussed. Thirdly, the design method of intrusion detection algorithm based on deep learning is elaborately explained, and the key points of intrusion prevention design are expounded. Finally, the effectiveness and accuracy of intrusion detection algorithm are verified through the simulation. Simulation results show that the proposed algorithm has a high threat detection accuracy for complex intrusion data, and the detection rate of the test data set calculated according to the formula is 95.22% and the false alarm rate is 0.67%.

Key words:intrusion prevention;deep learning;neural networks;intrusion detection

0 引言

2000年以后,我國的互联网技术在政策支持以及实际需求的双重助推下迎来了发展的高峰。物联网、大数据以及云计算等网络技术随之出现,通过技术变革的力量改善着人民大众的生产生活,有效提高社会资源的利用效率[1]。互联网的优势来自于其互联性与开放性,以网络服务为核心将现有的社会资源进行优化配置[2]。……

登录APP查看全文

猜你喜欢

深度学习
从合坐走向合学:浅议新学习模式的构建
面向大数据远程开放实验平台构建研究
基于自动智能分类器的图书馆乱架图书检测
搭建深度学习的三级阶梯
有体验的学习才是有意义的学习
电子商务中基于深度学习的虚假交易识别研究
利用网络技术促进学生深度学习的几大策略
MOOC与翻转课堂融合的深度学习场域建构
大数据技术在反恐怖主义中的应用展望
深度学习算法应用于岩石图像处理的可行性研究