Packet analysis in congested networks

Masaki Fukushima, Shigeki Goto

    Research output: Chapter in Book/Report/Conference proceedingChapter

    1 Citation (Scopus)

    Abstract

    This paper proposes new methods of measuring the Internet traffic. These are useful to analysing the network status, especially when the traffic is heavy, i.e. the network is congested. Our first method realizes a light weight measurement which counts only TCP flags, which occupies 6 bits in a TCP packet. Based on the simple flag counts, we can tell whether the network is congested or not. Moreover, we can estimate the average throughput of a network connection based on the flag count. Our second method analyses a sequence of TCP packets based on an automaton, or a protocol machine. The original automaton has been used in the formal specification of TCP protocol. However, it is not applicable to the real Internet traffic. We have improved the automaton in various ways, and established a modified machine. Using the new machine, we can analyse the Internet traffic even if there are packet losses.

    Original languageEnglish
    Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    Pages600-615
    Number of pages16
    Volume2281
    Publication statusPublished - 2002

    Publication series

    NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    Volume2281
    ISSN (Print)03029743
    ISSN (Electronic)16113349

      Fingerprint

    ASJC Scopus subject areas

    • Computer Science(all)
    • Theoretical Computer Science

    Cite this

    Fukushima, M., & Goto, S. (2002). Packet analysis in congested networks. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2281, pp. 600-615). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2281).