基于模糊贴近的网络故障数据快速定位方法实现

摘要

A network failure data fast positioning method based on fuzzy closeness algorithm is proposed in this paper. The method includes the establishment of warning correlation analysis model of network fault data,acquisition model of network fault data fuzzy warning closeness index,and network fault data location model based on closeness index reasoning for network busi⁃ness. The current network true warning term was processed in fuzzy mode and converted into fuzzy alarm. All the closeness indexes matching the target is discovered in the closeness index library by fuzzy matching,and the closeness index with highest match⁃ing rate is collected as the reasoning trigger index,from which the correlation information of failure data fuzzy alarm is extract⁃ed. The conclusion for fuzzy alarm of network fault data is reasoned out through fuzzy synthetic strategies. The experimental re⁃sults indicate that the proposed method for network fault data position has high positioning accuracy,and low energy consump⁃tion.%传统的网络故障数据定位方法大都基于固定指标分析告警信息，实现故障信号的定位，一旦指标数量过高则容易出现故障信号定位误差过高的问题。提出一种基于模糊贴近算法的网络故障数据快速定位方法，该方法包括塑造针对网络业务的网络故障数据告警关联性分析模型、采集网络故障数据模糊告警贴近指标，以及实现基于贴近指标推理的网络故障数据定位3个模块。系统将当前的网络事实告警项进行模糊化处理转换成模糊告警，由模糊匹配在贴近指标库中发现与之匹配的全部贴近指标，并采集匹配度最高的贴近指标当成推理触发指标。从该指标中提取故障数据模糊告警关联信息，通过模糊合成策略，推理出网络故障数据的模糊告警结论。实验结果表明，所提方法对网络故障数据进行定位具有较高的定位精度，并且消耗能量较低。