基于大数据聚类的通信网络安全态势预测技术  

Communication Network Security Situation Prediction Technology based on Big Data Clustering

作　　者：陈功平[1] 王红[1] CHEN Gong-ping;WANG Hong(School of Information and Electronic Engineering,Luan Vocational Technical College,Luan Anhui 237158,China)

机构地区：[1]六安职业技术学院信息与电子工程学院,安徽六安237158

出　　处：《淮阴师范学院学报（自然科学版）》2024年第1期20-26,共7页Journal of Huaiyin Teachers College;Natural Science Edition

基　　金：教育部科技发展中心高校产学研创新项目(2020ITA02018);安徽省教育厅高校自然科学研究重点项目(KJ2021A1361);安徽省高等学校省级质量工程项目(2019xfzx04)。

摘　　要：传统通信网络安全态势预测技术缺乏大数据支撑,难以对发生的攻击进行详细分类和追踪,导致在进行长时间的态势预测中收敛过慢,准确度降低.提出一种基于大数据聚类的通信网络安全态势预测技术.分析通信网络的属性以及特点,选择安全态势描述一级指标,将数据标准化处理之后,细分出二级指标;优化大数据聚类算法,计算最优聚类数量、确定聚类中心,建立关联规则库并优化预测流程,完成基于大数据聚类的通信网络安全态势预测技术的设计.通过实验结果表明,与两种传统的安全态势预测技术相比,设计的技术收敛速度更快,全体数据点没有出现残差扩散的现象,并且数据完整度较高.Traditional communication network security situation prediction technology lacks the support of big data,and it was difficult to classify and track the attacks in detail,which leads to slow convergence and low accuracy in long-term situation prediction.A communication network security situation prediction technology based on big data clustering was proposed.The attributes and characteristics of communication network are analyzed,the first level of security situation description index was selected,and the second level index was subdivided after standardized data processing.The big data clustering algorithm was optimized,the optimal clustering number was calculated,the clustering center was determined,the association rule library was established and the prediction process was optimized,and the design of communication network security situation prediction technology based on big data clustering was completed.Experimental results show that compared with the two traditional security situation prediction techniques,the designed technique has a faster convergence rate,and all data points do not appear residual diffusion phenomenon,and the data integrity was high.

关 键 词：大数据聚类 通信网络 安全态势 描述指标 聚类优化 收敛速度 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]