Greedy edge-disjoint paths in complete graphs

    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