基于网络事件的自智网络探索与应用

Exploration and application of autonomous network based on network events

为落实和达成自智网络的目标,提升网络运维自动化能力,探索了一种自智网络模式,引入多维数据自动聚类、自适应滑动时间窗机制、基于时空序列关联的用户流量损失评估模型、基于图论的故障传播图与时空聚类等算法,从网络事件管理、故障自识别、自定位、自调度和智能质检等多个方面进行自动化能力建设。力求2025年达到L4级高级自智网络的目标。

In order to achieve the goal of autonomous networks and improve the automation of network operation,an autonomous network model was explored.The model introduced algorithms such as multidimensional data automatic clustering,adaptive sliding time window mechanism,user traffic loss assessment model based on spatiotemporal se-quence correlation,fault propagation graph based on graph theory and spatiotemporal clustering,and builded automa-tion capabilities through network event management,fault self-identification,self-positioning,self-scheduling,and intelligent quality inspection.The goal is achieving L4 advanced intelligent network by 2025.