基于多源数据驱动的5G通信网络安全自动化实时预测系统  

Real time automation prediction system for 5G communication network security based on multi-source data-driven

作　　者：雷琳琳 喻朝新 李丹 LEI Linlin;YU Chaoxin;LI Dan(China Mobile GBA(Greater Bay Area)Innovation Institute,Guangzhou 510623,China)

机构地区：[1]中国移动大湾区(广东)创新研究院,广州510623

出　　处：《自动化与仪器仪表》2024年第7期208-212,共5页Automation & Instrumentation

摘　　要：单一数据源可能无法提供全面的信息,导致5G通信网络安全预测结果存在偏差。针对上述问题,设计基于多源数据驱动的5G通信网络安全自动化实时预测系统。基于Model-View-Controller架构设计系统框架,设计系统数据库,用于存储多源数据;利用探针采集器、Flume的采集器采集多源数据,并通过中间件将数据传输到数据库中;通过灰度关联度法进行多源数据降维,从中提取关键数据;量化5G通信网络安全并以此为因变量,以关键多源数据为自变量,构建多元回归预测模型,实现5G通信网络安全自动化实时预测。测试结果表明:回归预测模型的决定系数在0.9以上,由此证明所设计系统的预测准确性更高。A single data source may not be able to provide comprehensive information,leading to bias in 5G communication network security prediction results.To solve the above problems,an automated real-time prediction system for 5G communication network security driven by multi-source data is designed.Design system framework based on Model-View-Controller architecture,design system database for storing multi-source data;The probe collector and Flume collector are used to collect multi-source data,and the data is transferred to the database through the middleware.The dimensionality of multi-source data is reduced by gray relational degree method to extract key data.The security of 5G communication network is quantified and taken as the dependent variable,and the key multi-source data is taken as the independent variable,and the multiple regression prediction model is constructed to realize the automatic real-time prediction of 5G communication network security.The test results show that the determination coefficient of the regression prediction model is above 0.9,which proves that the prediction accuracy of the designed system is higher.

关 键 词：多源数据驱动 5G通信网络 网络安全 预测系统 

分 类 号：TP65.99[自动化与计算机技术—控制理论与控制工程]