| 题名： |
Backbone traffic pattern analysing based on joint entropy |
| 作者： |
徐杰,丁伟,龚俭,臧小东 |
| 杂志/会议： |
DASC/PiCom/DataCom/CyberSciTech |
| 卷、期、页码： |
|
| 时间： |
2016-08 |
| 关键词： |
network flow, network measurement, big data, data mining |
| 摘要： |
There are hundreds of billions packets passing through the network, how to analyse these big data is an arduous task. In this paper we proposed a new scheme based on cluster algorithm to explore the backbone network traffic pattern. Two new traffic features are introduced to assist the analysing process. These two novel attributes can represent the flow effectively cooperated with a new flow similarity measurement method proposed in this paper and they are insensitive to the random fluctuation of traffic size. The traffic pattern will be uncovered by a cluster algorithm derived from k-means. Real world traffic captured from a border router is used to test this new method. The result shows that our novel flow attributes and similarity measurement method perform well on the traffic analysis and four traffic patterns are uncovered. |
| 索引： |
EI:20164703031575 |
|
全文链接
导出
|