earticle

논문검색

Data Mining - A Captured Wired Traffic Approach

초록

영어

Data mining is involved in procedures by which patterns are extracted from data. This process has become increasingly important to map data patterns to useful information that can be used to predict future traffic analyses. Other areas were data mining can be used include: fraud detection, marketing, congestion control, and network expansion consideration.
Data mining involves capturing and gathering random data from the flow of information passing through a certain trunk or node and from a statistical point of view, it can only be meaningful if enough samples are gathered. The results achieved from a proper deployment of a data mining method provide valuable insights to how busy a node is, the average oneway and end-to-end delays, and the average size of the packets.. This paper contains the results from capturing live wired traffic and analyzing the statistical values.

목차

Abstract
 1. Introduction
 2. Data mining parameters
  2.1. Packet size
  2.2. Flow Duration
  2.3. Confidence Interval (CI)
 3. Real-time data measures
  3.1. Wired data analysis
  3.2. Scatter plots in data mining
  3.3. Throughput graphs
  3.4. Confidence Interval (CI) calculation
  3.5. TCP traffic versus UDP traffic
 4. Conclusions
 5. References

저자정보

  • Sasan Adibi ECE Department, University of Waterloo

참고문헌

자료제공 : 네이버학술정보
  • 1(Reference title not available)
  • 2(Reference title not available)
  • 3(Reference title not available)
  • 4(Reference title not available)
  • 5(Reference title not available)
  • 6(Reference title not available)
  • 7(Reference title not available)
  • 8(Reference title not available)
  • 9(Reference title not available)
  • 10(Reference title not available)
  • 11(Reference title not available)

※ 원문제공기관과의 협약기간이 종료되어 열람이 제한될 수 있습니다.

0개의 논문이 장바구니에 담겼습니다.