北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2017, Vol. 40 ›› Issue (s1): 24-28.doi: 10.13190/j.jbupt.2017.s.006

• 论文 • 上一篇    下一篇

面向高速网络流量的硬件测量装置

武昊, 黎阳, 张楚文, 刘斌   

  1. 清华大学 计算机科学与技术系, 北京 100084
  • 收稿日期:2016-03-23 出版日期:2017-09-28 发布日期:2017-09-28
  • 作者简介:武昊(1988-),男,博士生,E-mail:wu-h11@mails.tsinghua.edu.cn;刘斌(1964-),男,教授,博士生导师.
  • 基金资助:
    国家自然科学基金项目(61432009,61373143);教育部博士点基金项目(20131019172)

The Hardware Measurement System for High-Speed Network Flow

WU Hao, LI Yang, ZHANG Chu-wen, LIU Bin   

  1. Department of Computer Science and Technology, Tsinghua University, Beijing 100084, China
  • Received:2016-03-23 Online:2017-09-28 Published:2017-09-28

摘要: 为解决高速网络测量中面临的如何平衡测量速度、精度与存储器容量的问题,采用自适应非线性采样方法,基于高速现场可编程门阵列(FPGA)硬件平台,研究了针对高速网络流量的测量装置FAST.通过控制状态机设计、存储器功能划分和并行回导设计,FAST原型系统可根据实时流量动态调整采样概率,在较高的计数器压缩率下,保持99%以上的平均测量精度,分组测量吞吐量达27.4 Mpps.

关键词: 网络测量, 硬件测量装置, 采样方法

Abstract: Aiming at the problem that how to trade off the access throughput, accuracy and capacity of counter memory in high-speed network measurement, the non-linear sampling was adopted to develop an FPGA-based network measurement system FPGA-based adaptive sampling measurement (FAST). The key designs of FAST include the centralized state machine, the separate function banks in counter memory and the parallel export operations. Through numerous evaluations, this FAST prototype can adaptively tune sampling rate according to real-time flow volume and reach a high counter compression rate with more than 99% average accuracy rate. The throughput attains to 27.4Mpps.

Key words: network measurement, hardware-based measurement system, sampling method

中图分类号: