基于无监督学习算法的5G网络安全态势感知系统设计  

Design of 5G network security situation awareness system based on unsupervised learning algorithm

作　　者：任若冰 李启文[1] 王丹弘 REN Ruobing;LI Qiwen;WANG Danhong(China Mobile Communications Group Guangdong Co.,Ltd.,Guangzhou 510623,China)

机构地区：[1]中国移动通信集团广东有限公司,广东广州510623

出　　处：《电子设计工程》2025年第5期188-191,196,共5页Electronic Design Engineering

摘　　要：5G网络的特性使其更容易受到攻击和威胁,网络安全问题日益突出。为此,设计基于无监督学习算法的5G网络安全态势感知系统。在系统采集模块中利用三个采集器从各种网络设备和系统收集数据,态势要素筛选模块从收集的数据中筛选态势要素,态势感知模块中利用无监督学习算法,根据态势要素感知安全态势等级。结果表明,通过系统应用,安全态势在一段时间内存在波动和起伏,但总体上保持在一个较高的水平,5G网络在大部分时间都保持了较高的安全性,但在某些时间段内可能存在一些安全风险。无监督学习算法运行下内存占用率为22.63%,说明该系统在系统资源占用性能方面表现更好。The characteristics of 5G networks make them more vulnerable to attacks and threats,and network security issues are becoming increasingly prominent.Therefore,a 5G network security situation awareness system based on unsupervised learning algorithm is designed.In the system acquisition module,three collectors are used to collect data from various network devices and systems.In the situation factor screening module,situation factors are screened from the collected data.In the situation awareness module,unsupervised learning algorithm is used to perceive the security situation level according to situation factors.The results show that:through the system application,the security situation fluctuates and fluctuates for a period of time,but generally remains at a high level,indicating that the 5G network maintains a high security in most of the time,but there may be some security risks in some time periods.The memory usage of unsupervised learning algorithm is 22.63%,which indicates that the system has better performance in system resource usage.

关 键 词：无监督学习算法 5G网络 聚类算法 安全态势感知系统 

分 类 号：TN98[电子电信—信息与通信工程]