On the Internet delay space dimensionality

Bruno Abrahao, Robert Kleinberg

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

    Abstract

    We investigate the dimensionality properties of the Internet delay space, i.e., the matrix of measured round-trip latencies between Internet hosts. Previous work on network coordinates has indicated that this matrix can be embedded, with reasonably low distortion, into a 4- to 9-dimensional Euclidean space. The application of Principal Component Analysis (PCA) reveals the same dimensionality values. Our work addresses the question: to what extent is the dimensionality an intrinsic property of the delay space, defined without reference to a host metric such as Euclidean space? Is the intrinsic dimensionality of the Internet delay space approximately equal to the dimension determined using embedding techniques or PCA? If not, what explains the discrepancy? What properties of the network contribute to its overall dimensionality? Using datasets obtained via the King [14] method, we study different measures of dimensionality to establish the following conclusions. First, based on its power-law behavior, the structure of the delay space can be better characterized by fractal measures. Second, the intrinsic dimension is significantly smaller than the value predicted by the previous studies; in fact by our measures it is less than 2. Third, we demonstrate a particular way in which the AS topology is reflected in the delay space; subnetworks composed of hosts which share an upstream Tier-1 autonomous system in common possess lower dimensionality than the combined delay space. Finally, we observe that fractal measures, due to their sensitivity to non-linear structures, display higher precision for measuring the influence of subtle features of the delay space geometry.

    Original languageEnglish (US)
    Title of host publicationIMC'08
    Subtitle of host publicationProceedings of the 2008 ACM SIGCOMM Internet Measurement Conference
    Pages157-168
    Number of pages12
    DOIs
    StatePublished - Dec 1 2008
    EventInternet Measurement Conference 2008, IMC'08 - Vouliagmeni, Greece
    Duration: Oct 20 2008Oct 22 2008

    Publication series

    NameProceedings of the ACM SIGCOMM Internet Measurement Conference, IMC

    Conference

    ConferenceInternet Measurement Conference 2008, IMC'08
    CountryGreece
    CityVouliagmeni
    Period10/20/0810/22/08

    Fingerprint

    Internet
    Fractals
    Principal component analysis
    Topology
    Geometry

    Keywords

    • Delay space
    • Dimensionality
    • Internet structure
    • Network embedding

    ASJC Scopus subject areas

    • Software
    • Computer Networks and Communications

    Cite this

    Abrahao, B., & Kleinberg, R. (2008). On the Internet delay space dimensionality. In IMC'08: Proceedings of the 2008 ACM SIGCOMM Internet Measurement Conference (pp. 157-168). (Proceedings of the ACM SIGCOMM Internet Measurement Conference, IMC). https://doi.org/10.1145/1452520.1452541

    On the Internet delay space dimensionality. / Abrahao, Bruno; Kleinberg, Robert.

    IMC'08: Proceedings of the 2008 ACM SIGCOMM Internet Measurement Conference. 2008. p. 157-168 (Proceedings of the ACM SIGCOMM Internet Measurement Conference, IMC).

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

    Abrahao, B & Kleinberg, R 2008, On the Internet delay space dimensionality. in IMC'08: Proceedings of the 2008 ACM SIGCOMM Internet Measurement Conference. Proceedings of the ACM SIGCOMM Internet Measurement Conference, IMC, pp. 157-168, Internet Measurement Conference 2008, IMC'08, Vouliagmeni, Greece, 10/20/08. https://doi.org/10.1145/1452520.1452541
    Abrahao B, Kleinberg R. On the Internet delay space dimensionality. In IMC'08: Proceedings of the 2008 ACM SIGCOMM Internet Measurement Conference. 2008. p. 157-168. (Proceedings of the ACM SIGCOMM Internet Measurement Conference, IMC). https://doi.org/10.1145/1452520.1452541
    Abrahao, Bruno ; Kleinberg, Robert. / On the Internet delay space dimensionality. IMC'08: Proceedings of the 2008 ACM SIGCOMM Internet Measurement Conference. 2008. pp. 157-168 (Proceedings of the ACM SIGCOMM Internet Measurement Conference, IMC).
    @inproceedings{f186abbdc21646049f9e003d89fbaaf1,
    title = "On the Internet delay space dimensionality",
    abstract = "We investigate the dimensionality properties of the Internet delay space, i.e., the matrix of measured round-trip latencies between Internet hosts. Previous work on network coordinates has indicated that this matrix can be embedded, with reasonably low distortion, into a 4- to 9-dimensional Euclidean space. The application of Principal Component Analysis (PCA) reveals the same dimensionality values. Our work addresses the question: to what extent is the dimensionality an intrinsic property of the delay space, defined without reference to a host metric such as Euclidean space? Is the intrinsic dimensionality of the Internet delay space approximately equal to the dimension determined using embedding techniques or PCA? If not, what explains the discrepancy? What properties of the network contribute to its overall dimensionality? Using datasets obtained via the King [14] method, we study different measures of dimensionality to establish the following conclusions. First, based on its power-law behavior, the structure of the delay space can be better characterized by fractal measures. Second, the intrinsic dimension is significantly smaller than the value predicted by the previous studies; in fact by our measures it is less than 2. Third, we demonstrate a particular way in which the AS topology is reflected in the delay space; subnetworks composed of hosts which share an upstream Tier-1 autonomous system in common possess lower dimensionality than the combined delay space. Finally, we observe that fractal measures, due to their sensitivity to non-linear structures, display higher precision for measuring the influence of subtle features of the delay space geometry.",
    keywords = "Delay space, Dimensionality, Internet structure, Network embedding",
    author = "Bruno Abrahao and Robert Kleinberg",
    year = "2008",
    month = "12",
    day = "1",
    doi = "10.1145/1452520.1452541",
    language = "English (US)",
    isbn = "9781605583341",
    series = "Proceedings of the ACM SIGCOMM Internet Measurement Conference, IMC",
    pages = "157--168",
    booktitle = "IMC'08",

    }

    TY - GEN

    T1 - On the Internet delay space dimensionality

    AU - Abrahao, Bruno

    AU - Kleinberg, Robert

    PY - 2008/12/1

    Y1 - 2008/12/1

    N2 - We investigate the dimensionality properties of the Internet delay space, i.e., the matrix of measured round-trip latencies between Internet hosts. Previous work on network coordinates has indicated that this matrix can be embedded, with reasonably low distortion, into a 4- to 9-dimensional Euclidean space. The application of Principal Component Analysis (PCA) reveals the same dimensionality values. Our work addresses the question: to what extent is the dimensionality an intrinsic property of the delay space, defined without reference to a host metric such as Euclidean space? Is the intrinsic dimensionality of the Internet delay space approximately equal to the dimension determined using embedding techniques or PCA? If not, what explains the discrepancy? What properties of the network contribute to its overall dimensionality? Using datasets obtained via the King [14] method, we study different measures of dimensionality to establish the following conclusions. First, based on its power-law behavior, the structure of the delay space can be better characterized by fractal measures. Second, the intrinsic dimension is significantly smaller than the value predicted by the previous studies; in fact by our measures it is less than 2. Third, we demonstrate a particular way in which the AS topology is reflected in the delay space; subnetworks composed of hosts which share an upstream Tier-1 autonomous system in common possess lower dimensionality than the combined delay space. Finally, we observe that fractal measures, due to their sensitivity to non-linear structures, display higher precision for measuring the influence of subtle features of the delay space geometry.

    AB - We investigate the dimensionality properties of the Internet delay space, i.e., the matrix of measured round-trip latencies between Internet hosts. Previous work on network coordinates has indicated that this matrix can be embedded, with reasonably low distortion, into a 4- to 9-dimensional Euclidean space. The application of Principal Component Analysis (PCA) reveals the same dimensionality values. Our work addresses the question: to what extent is the dimensionality an intrinsic property of the delay space, defined without reference to a host metric such as Euclidean space? Is the intrinsic dimensionality of the Internet delay space approximately equal to the dimension determined using embedding techniques or PCA? If not, what explains the discrepancy? What properties of the network contribute to its overall dimensionality? Using datasets obtained via the King [14] method, we study different measures of dimensionality to establish the following conclusions. First, based on its power-law behavior, the structure of the delay space can be better characterized by fractal measures. Second, the intrinsic dimension is significantly smaller than the value predicted by the previous studies; in fact by our measures it is less than 2. Third, we demonstrate a particular way in which the AS topology is reflected in the delay space; subnetworks composed of hosts which share an upstream Tier-1 autonomous system in common possess lower dimensionality than the combined delay space. Finally, we observe that fractal measures, due to their sensitivity to non-linear structures, display higher precision for measuring the influence of subtle features of the delay space geometry.

    KW - Delay space

    KW - Dimensionality

    KW - Internet structure

    KW - Network embedding

    UR - http://www.scopus.com/inward/record.url?scp=63049112027&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=63049112027&partnerID=8YFLogxK

    U2 - 10.1145/1452520.1452541

    DO - 10.1145/1452520.1452541

    M3 - Conference contribution

    AN - SCOPUS:63049112027

    SN - 9781605583341

    T3 - Proceedings of the ACM SIGCOMM Internet Measurement Conference, IMC

    SP - 157

    EP - 168

    BT - IMC'08

    ER -