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

    Internet Traffic
    Internet
    Automata
    Count
    Light measurement
    Network protocols
    Weighing
    Packet loss
    Formal Specification
    Packet Loss
    Throughput
    Traffic
    Estimate
    Formal specification

    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).

    Packet analysis in congested networks. / Fukushima, Masaki; Goto, Shigeki.

    Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 2281 2002. p. 600-615 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2281).

    Research output: Chapter in Book/Report/Conference proceedingChapter

    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, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 2281, pp. 600-615.
    Fukushima M, Goto S. 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. 2002. p. 600-615. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
    Fukushima, Masaki ; Goto, Shigeki. / Packet analysis in congested networks. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 2281 2002. pp. 600-615 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
    @inbook{7cb9c031f599420aa04c72f7278ba98f,
    title = "Packet analysis in congested networks",
    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.",
    author = "Masaki Fukushima and Shigeki Goto",
    year = "2002",
    language = "English",
    isbn = "3540433384",
    volume = "2281",
    series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
    pages = "600--615",
    booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

    }

    TY - CHAP

    T1 - Packet analysis in congested networks

    AU - Fukushima, Masaki

    AU - Goto, Shigeki

    PY - 2002

    Y1 - 2002

    N2 - 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.

    AB - 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.

    UR - http://www.scopus.com/inward/record.url?scp=23044533726&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=23044533726&partnerID=8YFLogxK

    M3 - Chapter

    AN - SCOPUS:23044533726

    SN - 3540433384

    SN - 9783540433385

    VL - 2281

    T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

    SP - 600

    EP - 615

    BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

    ER -