电信科学 ›› 2013, Vol. 29 ›› Issue (11): 94-98.doi: 10.3969／j.issn.1000-0801.2013.11.016

• 电力信息化专栏 • 上一篇    下一篇

基于TiIera平台的网络细粒度应用行为识别

吴舜1,苏丹1,吴佳1,李坤1,许大卫1,刘昀1,魏征2   

  1. 1 国网冀北电力有限公司信通分公司 北京100053
    2 西安交通大学 西安710049
  • 出版日期:2013-11-20 发布日期:2017-07-04

Fine-Grained Network Traffic Identification Based on Tilera Platform

Shun Wu1,Dan Su1,Jia Wu1,Kun Li1,Dawei Xu1,Yun Liu1,Zheng Wei2   

  1. 1 Information ＆ Telecommunication Company of State Grid Jibei Electric Power Co., Ltd., Beijing 100053, China
    2 Xi'an Jiaotong University, Xi'an 710049, China
  • Online:2013-11-20 Published:2017-07-04

在线阅读

3

PDF下载

438

摘要：

网络管理员需要根据用户对应用的使用情况决定如何合理地分配带宽，提供区别化服务，进行安全监测和管理，以进行更加细致和准确的IT运维。研究了网络细粒度应用行为识别系统，能根据应用产生的流量精准识别用户在应用上的行为。与x86架构相比，为了实现更高的吞吐率和更低的功耗，应着重在Tilera平台上对应用识别进行并行设计和实现。最后对两种平台下系统的各种性能测试结果进行了对比，并得出结论。

关键词: 网络行为识别, Tilera平台, 并行化

Abstract:

It is essential for network managers to decide how to use a reasonable allocation of bandwidth, providing differentiated services, as well as conducting security monitor and management so as to do fine-grained and precise IT operation. The fine-grained traffic identification method proposed was able to identify the user's behavior accurately according to the traffic that the application generates. In addition, compared with the traditional x86 platform, in order to achieve higher throughput and lower power consumption, the method was redesigned and implemented in the Tilera platform with parallelization. Finally, a variety of system performance test results were compared under the two platforms.

Key words: network traffic identification, Tilera platform, parallelization

No Suggested Reading articles found!