数据挖掘技术在通信网告警分析中的应用

Application of data mining In Communication Network alarm analyzing

在通信网中每天都会产生大量的告警数据,其中隐藏着很多有价值的信息,可用来过滤各种冗余告警,实现智能化故障定位、诊断和预测,告警关联性分析则是实现前述功能的重要基础。本文通过对通信网告警数据以及告警之间关联性的分析,提出将Apriori关联挖掘算法与事件滑动窗口相结合的告警相关性规则生成算法,实现自动生成告警相关性规则。

Telecommunication networks produce large amounts of alarm daily. This data contains potentially valuable knowledge about the network. This kind of knowledge could be used to filter alarm, guide the intelligent network fault location, diagnosis and forecasting. The basis of alarm filtering and fault location is alarm relativity analysis. This paper analyses the characteristic of alarm data and alarm relativity in detail , then presents an algorithm acquiring alarm time association rule by combining classical Apriori algorithm and time window.