Greedy edge-disjoint paths in complete graphs

Paz Carmi, Thomas Erlebach, Yoshio Okamoto

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

8 Scopus citations

Abstract

The maximum edge-disjoint paths problem (MEDP) is one of the most classical NP-hard problems. We study the approximation ratio of a simple and practical approximation algorithm, the shortest-path-first greedy algorithm (SGA), for MEDP in complete graphs. Previously, it was known that this ratio is at most 54. Adapting results by Kolman and Scheideler [Proceedings of SODA, 2002, pp. 184-193], we show that SGA achieves approximation ratio 8F + 1 for MEDP in undirected graphs with flow number F and, therefore, has approximation ratio at most 9 in complete graphs. Furthermore, we construct a family of instances that shows that SGA cannot be better than a 3-approximation algorithm. Our upper and lower bounds hold also for the bounded-length greedy algorithm, a simple on-line algorithm for MEDP.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
EditorsHans L. Bodlaender
PublisherSpringer Verlag
Pages143-155
Number of pages13
ISBN (Electronic)9783540204527
DOIs
StatePublished - 1 Jan 2003

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2880
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Keywords

  • Approximation algorithm
  • Greedy algorithm
  • Shortening lemma

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'Greedy edge-disjoint paths in complete graphs'. Together they form a unique fingerprint.

Cite this