OAK

Evaluation of steady-state probability of Pareto/M/1/K experiencing tail-raising effect

Metadata Downloads
Author(s)
Koh, Y.; Kim, Ki Seon
Type
Conference Paper
Citation
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), pp.561 - 570
Issued Date
2003-07
Abstract
To resolve the problem of evaluating for the steady-state probability of Pareto/M/1/K showing the tail-raising effect, we propose an empirical model of the steady-state probability and justified by evaluating the mean queue length and delay by simulations.
Publisher
Springer-Verlag Berlin Heidelberg
Conference Place
GE
URI
https://scholar.gist.ac.kr/handle/local/29074
공개 및 라이선스
  • 공개 구분공개
파일 목록
  • 관련 파일이 존재하지 않습니다.

Items in Repository are protected by copyright, with all rights reserved, unless otherwise indicated.