一種新的基于事件觸發(fā)的網(wǎng)絡(luò)流量采樣測量方法及其應(yīng)用研究
A Novel Network Traffic Sampling Measurement Method Based on Event-Driven and Its Application Research
-
摘要: 該文主要對采樣測量方法進(jìn)行了研究，提出了基于事件觸發(fā)的分層隨機(jī)雙采樣方法(ESRDS)。通過與傳統(tǒng)采樣方法的仿真實(shí)驗(yàn)比較，證明該方法在分析包長、包到達(dá)時(shí)間間隔時(shí)性能都比較高，尤其該方法采樣結(jié)果在估計(jì)包到達(dá)時(shí)間間隔時(shí)方面和實(shí)際比較接近。同時(shí)作者還對采樣測量方法在計(jì)算Hurst參數(shù)方面進(jìn)行了研究,應(yīng)用ESRDS對網(wǎng)絡(luò)流量的Hurst參數(shù)進(jìn)行了估計(jì)，估計(jì)值和實(shí)際值之間的誤差非常小。
-
關(guān)鍵詞:
- 采樣測量; Hurst系數(shù); 雙采樣方法
Abstract: This paper researches mainly 011 the network traffic sampling measurement methods. The Stratified Random Double Sampling method based on Event-driven (ESRDS) is proposed. Its performance is proved to be better than traditional methods in analyzing the length and interval of packets according to simulation experiments. Especially, estimated interval time of packages by this method is nearly real. It also researches on the measurement methods in the calculating Hurst coefficient. According to the estimation of Hurst coefficient using ESRDS method, the error is much less between estimated and real values. -
Cceres R, Duffield N G, Feldmann A, et al.. Measurement and analysis of IP network usage and behavior. IEEE Communications Magazine, 2000, 38(5): 144-151.[2]Duffeld N G, Grossglauser M. Trajectory sampling for direct traffic observation, Computer Communication Review, 2000, 30(4): 271-282.[3]Estan C, Varghese G. New directions in traffic measurement and accounting. In Proceedings of ACM SIGCOMM, San Francisco, CA, 2002: 323-336.[4]Duffield N, Lund C, Thorup M. Charging from sampled network usage. ACM SIGCOMM Internet Measurement Workshop 2001, San Francisco, CA, November 1-2, 2001: 101-112.[5]Claffy K C, Polyzos G C, Braun H W. Application of sampling methodologies to network traffic characterization[J].Computer Communication Review.1993, 23(4):194-203 -
計(jì)量
- 文章訪問數(shù): 2278
- HTML全文瀏覽量: 137
- PDF下載量: 871
- 被引次數(shù): 0
下載:
