北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2005, Vol. 28 ›› Issue (3): 89-92.doi: 10.13190/jbupt.200503.89.zhangf

• 论文 • 上一篇    下一篇

改进型链路负载自适应抽样测量算法

张峰,雷振明   

  1. 北京邮电大学 信息工程学院, 北京 100876
  • 出版日期:2005-06-28 发布日期:2005-06-28

Improved Adaptive Sampling Technique for Link Load Measurement

ZHANG Feng,LEI Zhenming   

  1. School of Information Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China
  • Online:2005-06-28 Published:2005-06-28

摘要:

为了提高网络负载测量的效益和灵活性,在对自适应控制理论及实际流量特征进行研究后,提出了一种改进型的网络负载自适应抽样测量算法, 并对该算法的优越性及改进前后算法的性能对比进行了仿真研究. 结果表明,改进型算法不仅降低了测量开销,而且还可将测量精度提高40%以上. 另外,该算法可提供不同可视化粒度输出,使网络负载测量具有更大的灵活性和可操作性,更适宜于高速、高突发程度的网络测量环境.

关键词: 报文抽样, 链路负载测量, 自适应控制

Abstract:

An improved adaptive sampling algorithms to improve the efficiency and flexibility of network load measurement is proposed after the detailed research on the theory of adaptive control and traffic characteristics. Performance comparison ,shows that there is improvement by 40% on the measurement precision. It can also provide several visual outputs with different granularity to support the measurement efficiency and flexibility. So the improved algorithm suits for the environments of measuring highspeed network load with high burst arrival rate.

Key words: packet sampling, link load measurement, adaptive control

中图分类号: