We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Network abnormal flow grouping method for cloud computing.
- Authors
LI Chun-lin; HUANG Yue-jiang; NIU Chang-xi
- Abstract
The question of how to manage the abnormal flows among the enormous user flows is issued to cloud computing. Considering that traditional grouping methods are not suitable for cloud computing, this paper proposed a method of grouping these abnormal flows into group. The main idea was clustering the flow data of each user by BIRCH algorithm at first. And then it merged these clustering results into new groups. The merging step overcame the deficiency of BIRCH's lacking of soft clustering. It applied this method to a scenario with abrupt abnormal flows. The result shows that the method can successfully distinguishing these abnormal users by group from normal user groups. And the result also proves that drop packets randomly within groups generated by this method having better performance than dropping packet randomly or dropping randomly weighted by service type.
- Subjects
CLOUD computing; DATA flow computing; COMPUTER networks; GROUP theory; DATA packeting; CLUSTER analysis (Statistics)
- Publication
Application Research of Computers / Jisuanji Yingyong Yanjiu, 2014, Vol 31, Issue 12, p3704
- ISSN
1001-3695
- Publication type
Article
- DOI
10.3969/j.issn.1001-3695.2014.12.047