On the characteristics of Internet traffic variability: Spikes and elephants

Tatsuya Mori, Shozo Naito, Ryoichi Kawahara, Shigeki Goto

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    24 Citations (Scopus)

    Abstract

    Analysing and modeling of traffic play a vital role in designing and controlling of networks effectively. To construct a practical traffic model that can be used for various networks, it is necessary to characterize aggregated traffic and user traffic. This paper investigates these characteristics and their relationship. Our analyses are based on a huge number of packet traces from five different networks on the Internet. We found that: (1) marginal distributions of aggregated traffic fluctuations follow positively skewed (non-Gaussian) distributions, which leads to the existence of "spikes", where spikes correspond to an extremely large value of momentary throughput, (2) the amount of user traffic in a unit of time has a wide range of variability, and (3) flows within spikes are more likely to be "elephant flows", where an elephant flow is an IP flow with a high volume of traffic. These findings are useful in constructing a practical and realistic Internet traffic model.

    Original languageEnglish
    Title of host publicationProceedings - International Symposium on Applications and the Internet
    Pages99-106
    Number of pages8
    DOIs
    Publication statusPublished - 2004
    EventProceedings - 2004 International Symposium on Applications and the Internet (Saint 2004) - Tokyo
    Duration: 2004 Jan 262004 Jan 30

    Other

    OtherProceedings - 2004 International Symposium on Applications and the Internet (Saint 2004)
    CityTokyo
    Period04/1/2604/1/30

    Fingerprint

    Internet
    Throughput

    ASJC Scopus subject areas

    • Engineering(all)

    Cite this

    Mori, T., Naito, S., Kawahara, R., & Goto, S. (2004). On the characteristics of Internet traffic variability: Spikes and elephants. In Proceedings - International Symposium on Applications and the Internet (pp. 99-106) https://doi.org/10.1109/SAINT.2004.1266104

    On the characteristics of Internet traffic variability : Spikes and elephants. / Mori, Tatsuya; Naito, Shozo; Kawahara, Ryoichi; Goto, Shigeki.

    Proceedings - International Symposium on Applications and the Internet. 2004. p. 99-106.

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Mori, T, Naito, S, Kawahara, R & Goto, S 2004, On the characteristics of Internet traffic variability: Spikes and elephants. in Proceedings - International Symposium on Applications and the Internet. pp. 99-106, Proceedings - 2004 International Symposium on Applications and the Internet (Saint 2004), Tokyo, 04/1/26. https://doi.org/10.1109/SAINT.2004.1266104
    Mori T, Naito S, Kawahara R, Goto S. On the characteristics of Internet traffic variability: Spikes and elephants. In Proceedings - International Symposium on Applications and the Internet. 2004. p. 99-106 https://doi.org/10.1109/SAINT.2004.1266104
    Mori, Tatsuya ; Naito, Shozo ; Kawahara, Ryoichi ; Goto, Shigeki. / On the characteristics of Internet traffic variability : Spikes and elephants. Proceedings - International Symposium on Applications and the Internet. 2004. pp. 99-106
    @inproceedings{5314a140f5f0419f979fcb9cd04c3c6b,
    title = "On the characteristics of Internet traffic variability: Spikes and elephants",
    abstract = "Analysing and modeling of traffic play a vital role in designing and controlling of networks effectively. To construct a practical traffic model that can be used for various networks, it is necessary to characterize aggregated traffic and user traffic. This paper investigates these characteristics and their relationship. Our analyses are based on a huge number of packet traces from five different networks on the Internet. We found that: (1) marginal distributions of aggregated traffic fluctuations follow positively skewed (non-Gaussian) distributions, which leads to the existence of {"}spikes{"}, where spikes correspond to an extremely large value of momentary throughput, (2) the amount of user traffic in a unit of time has a wide range of variability, and (3) flows within spikes are more likely to be {"}elephant flows{"}, where an elephant flow is an IP flow with a high volume of traffic. These findings are useful in constructing a practical and realistic Internet traffic model.",
    author = "Tatsuya Mori and Shozo Naito and Ryoichi Kawahara and Shigeki Goto",
    year = "2004",
    doi = "10.1109/SAINT.2004.1266104",
    language = "English",
    isbn = "0769520685",
    pages = "99--106",
    booktitle = "Proceedings - International Symposium on Applications and the Internet",

    }

    TY - GEN

    T1 - On the characteristics of Internet traffic variability

    T2 - Spikes and elephants

    AU - Mori, Tatsuya

    AU - Naito, Shozo

    AU - Kawahara, Ryoichi

    AU - Goto, Shigeki

    PY - 2004

    Y1 - 2004

    N2 - Analysing and modeling of traffic play a vital role in designing and controlling of networks effectively. To construct a practical traffic model that can be used for various networks, it is necessary to characterize aggregated traffic and user traffic. This paper investigates these characteristics and their relationship. Our analyses are based on a huge number of packet traces from five different networks on the Internet. We found that: (1) marginal distributions of aggregated traffic fluctuations follow positively skewed (non-Gaussian) distributions, which leads to the existence of "spikes", where spikes correspond to an extremely large value of momentary throughput, (2) the amount of user traffic in a unit of time has a wide range of variability, and (3) flows within spikes are more likely to be "elephant flows", where an elephant flow is an IP flow with a high volume of traffic. These findings are useful in constructing a practical and realistic Internet traffic model.

    AB - Analysing and modeling of traffic play a vital role in designing and controlling of networks effectively. To construct a practical traffic model that can be used for various networks, it is necessary to characterize aggregated traffic and user traffic. This paper investigates these characteristics and their relationship. Our analyses are based on a huge number of packet traces from five different networks on the Internet. We found that: (1) marginal distributions of aggregated traffic fluctuations follow positively skewed (non-Gaussian) distributions, which leads to the existence of "spikes", where spikes correspond to an extremely large value of momentary throughput, (2) the amount of user traffic in a unit of time has a wide range of variability, and (3) flows within spikes are more likely to be "elephant flows", where an elephant flow is an IP flow with a high volume of traffic. These findings are useful in constructing a practical and realistic Internet traffic model.

    UR - http://www.scopus.com/inward/record.url?scp=2642544060&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=2642544060&partnerID=8YFLogxK

    U2 - 10.1109/SAINT.2004.1266104

    DO - 10.1109/SAINT.2004.1266104

    M3 - Conference contribution

    AN - SCOPUS:2642544060

    SN - 0769520685

    SN - 9780769520681

    SP - 99

    EP - 106

    BT - Proceedings - International Symposium on Applications and the Internet

    ER -