模糊关联规则在网络告警分析中的应用研究

发布时间：2018-07-10 18:22

本文选题：数据挖掘 + 关联规则　；参考：《北京邮电大学》2016年硕士论文

【摘要】：当今网络技术发展非常迅速,互联网逐渐变得的越来越普及,人们学习生活的各个方面都已经离不开网络。而对于日渐复杂的网络系统来说,如果网络处于不正常的运行状态,难免会产生海量的告警事件,对人们的学习生活带来重大的影响。因此,如何应对网络告警,从复杂冗余的告警数据中发现告警之间隐藏的关联关系,使网络正常稳定的运行,这正在逐渐成为一个需要解决的问题。目前在告警相关性分析的研究中,科研工作者们已经采用了多种方法来进行研究,因为数据挖掘技术在数据处理方面具有一定的优势,近年来正成为告警相关性分析中一个主要的研究方向,取得了一定的研究成果,但仍然存在一些不足之处。比如,故障与告警之间往往存在一种模糊的对应关系,而在传统方法中,往往考虑不到这种模糊的关系,以至于影响故障的准确推断。对于上述存在的问题,本文将模糊理论的思想应用于告警数据的分析中,对告警信息依据模糊理论的方法进行处理,以解决故障与告警之间划分过硬,不能准确推断主要故障源点的问题。在论文中,首先研究了模糊聚类算法以及关联规则挖掘算法,在此基础上,通过仿真实验来探究模糊理论的思想在告警相关性分析中的应用效果,经过试验证明,该方法可以有效地应用于告警相关性分析中。
[Abstract]:Nowadays, with the rapid development of network technology, the Internet is becoming more and more popular. For the increasingly complex network system, if the network is in an abnormal state, it will inevitably produce a large number of alarm events, which will have a significant impact on people's learning and life. Therefore, how to deal with the network alarm, discover the hidden correlation between the alarms from the complex redundant alarm data, and make the network run smoothly is becoming a problem that needs to be solved gradually. At present, in the research of alarm correlation analysis, researchers have adopted a variety of methods to study, because data mining technology has some advantages in data processing, In recent years, it is becoming a main research direction in the analysis of alarm correlation, and some research results have been obtained, but there are still some shortcomings. For example, there is a fuzzy correspondence relationship between fault and alarm, but in the traditional method, the fuzzy relationship is often not considered, which affects the accurate inference of fault. For the problems mentioned above, this paper applies the idea of fuzzy theory to the analysis of alarm data, and deals with the alarm information according to the method of fuzzy theory, in order to solve the problem of perfect division between fault and alarm. The problem of the main fault source cannot be inferred accurately. In this paper, firstly, the fuzzy clustering algorithm and association rule mining algorithm are studied. On this basis, the application effect of fuzzy theory in alarm correlation analysis is explored through simulation experiments. This method can be effectively applied to alarm correlation analysis.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP311.13;TP393.06

【参考文献】