Clustering network traffic utilization
Classification of network traffic using distinctive characteristic application is not ideal for P2P and HTTP protocols. This is for the case when a user intercepts the application from other proxy or dynamic port, then the bytes utilization can be manipulated. In this paper, we present a clustering...
Saved in:
| Main Authors: | , , , , |
|---|---|
| 格式: | Article |
| 語言: | English |
| 出版: |
Praise Worthy Prize
2013
|
| 在線閱讀: | http://psasir.upm.edu.my/id/eprint/30577/1/Clustering%20network%20traffic%20utilization.pdf http://psasir.upm.edu.my/id/eprint/30577/ |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | Classification of network traffic using distinctive characteristic application is not ideal for P2P and HTTP protocols. This is for the case when a user intercepts the application from other proxy or dynamic port, then the bytes utilization can be manipulated. In this paper, we present a clustering approach for network traffic classification using information from one particular port. The clustering experiments were conducted using three different clustering algorithms, which are K-Means, DBScan and AutoClass. The analysis discussed on the quality of resulting clusters from all the algorithms.
|
|---|