1. NCCU Academic Hub
  2. 學術產出
  3. College of Commerce
  4. Proceedings
  5. Item

學術產出-Proceedings

Article View/Open

Publication Export

Google ScholarTM

政大圖書館

Citation Infomation

題名 A Fuzzy-based Algorithm to Remove Clock Skew and Reset from One-Way Delay Measurement
作者 郭更生
Kuo, Geng-Sheng
貢獻者 IEEE
日期 2004-11
上傳時間 6-Oct-2010 09:18:41 (UTC+8)
摘要 One-way delay (OWD) traces are important measurements for analyzing end-to-end performance on the Internet. It is still a great challenge to provide a scalable solution for large-scale OWD measurement. Because the clocks at end systems are usually not synchronized, the OWD measurements are often inaccurate. For the more challenging case with clock resets to some reference times during the measurement, the OWD measurements are more inaccurate. Furthermore, the measurement data often exhibit considerable network-induced noise when the network is heavily loaded. All the existing OWD measurement techniques, such as linear programming and convex-hull approach (CHA), try to solve this problem by deterministic mathematics model. However, they often fail to distinguish clock resets from temporary Internet congestion. Based on the fuzzy-clustering analysis, this paper proposes a new algorithm to estimate and remove the clock skews and resets from measurement results. This algorithm has been implemented as a tool called fuzzy-based OWD corrector (FOC). The paper then presents OWD measurements of several Internet paths using FOC. Numerical experiments demonstrate that FOC is more accurate and robust than the existing techniques. FOCs computation complexity O(N) is similar to that of CHA and its computing time is much less than that of convex-hull technique.
關聯 Neural Networks, IEEE Transactions on (Volume:16 , Issue: 5 )
     Page(s):
     1125 - 1135
資料類型 conference
DOI http://dx.doi.org/10.1109/GLOCOM.2004.1378218
dc.contributor IEEEen_US
dc.creator (作者) 郭更生zh_TW
dc.creator (作者) Kuo, Geng-Sheng-
dc.date (日期) 2004-11en_US
dc.date.accessioned 6-Oct-2010 09:18:41 (UTC+8)-
dc.date.available 6-Oct-2010 09:18:41 (UTC+8)-
dc.date.issued (上傳時間) 6-Oct-2010 09:18:41 (UTC+8)-
dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/45302-
dc.description.abstract (摘要) One-way delay (OWD) traces are important measurements for analyzing end-to-end performance on the Internet. It is still a great challenge to provide a scalable solution for large-scale OWD measurement. Because the clocks at end systems are usually not synchronized, the OWD measurements are often inaccurate. For the more challenging case with clock resets to some reference times during the measurement, the OWD measurements are more inaccurate. Furthermore, the measurement data often exhibit considerable network-induced noise when the network is heavily loaded. All the existing OWD measurement techniques, such as linear programming and convex-hull approach (CHA), try to solve this problem by deterministic mathematics model. However, they often fail to distinguish clock resets from temporary Internet congestion. Based on the fuzzy-clustering analysis, this paper proposes a new algorithm to estimate and remove the clock skews and resets from measurement results. This algorithm has been implemented as a tool called fuzzy-based OWD corrector (FOC). The paper then presents OWD measurements of several Internet paths using FOC. Numerical experiments demonstrate that FOC is more accurate and robust than the existing techniques. FOCs computation complexity O(N) is similar to that of CHA and its computing time is much less than that of convex-hull technique.-
dc.language.iso en_US-
dc.relation (關聯) Neural Networks, IEEE Transactions on (Volume:16 , Issue: 5 )
     Page(s):
     1125 - 1135		en_US
dc.title (題名) A Fuzzy-based Algorithm to Remove Clock Skew and Reset from One-Way Delay Measurementen_US
dc.type (資料類型) conferenceen
dc.identifier.doi (DOI) 10.1109/GLOCOM.2004.1378218en_US
dc.doi.uri (DOI) http://dx.doi.org/10.1109/GLOCOM.2004.1378218 en_US