时空尺度下大规模域间路由事件检测方法

发布时间：2018-10-24 07:37

【摘要】：针对现有方法无法对大规模协同性域间路由事件进行有效检测的问题,提出一个基于时空张量的分布式域间路由分析框架.首先从BGP路由表中提取域间路由的时空特征,并使用MapReduce框架生成时空张量,用于表征域间路由的动态变化,然后使用张量分解的方法从时空张量中提取二维事件,最后运用矩阵聚类的方法生成三维域间路由事件.将该分析框架运用于不同时间跨度和时间粒度的路由表数据集,实验结果表明,该方法不仅能够准确检测局部域间路由事件,而且能够从时间和空间尺度上对全球域间路由事件进行有效剖析,揭示了域间路由事件具有影响范围广、持续时间短等一般规律.
[Abstract]:A distributed inter-domain routing analysis framework based on space-time Zhang Liang is proposed to solve the problem that existing methods can not effectively detect large-scale cooperative inter-domain routing events. Firstly, the temporal and spatial characteristics of inter-domain routing are extracted from the BGP routing table, and the spatio-temporal Zhang Liang is generated by using the MapReduce framework, which is used to represent the dynamic changes of inter-domain routing. Finally, three-dimensional inter-domain routing events are generated by matrix clustering. The analysis framework is applied to routing table data sets with different time span and time granularity. The experimental results show that the proposed method can not only accurately detect the routing events between local domains. Moreover, the global inter-domain routing events can be effectively analyzed from the scale of time and space, which reveals the general rules of inter-domain routing events, such as wide range of influence, short duration and so on.
【作者单位】：电子工程学院网络系;
【分类号】：TP393.06

【参考文献】