GMM based semi-supervised learning for channel-based authentication scheme

Nikhil Gulati, Rachel Greenstadt, Kapil R. Dandekar, John M. Walsh

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

    Abstract

    Authentication schemes based on wireless physical layer channel information have gained significant attention in recent years. It has been shown in recent studies, that the channel based authentication can either cooperate with existing higher layer security protocols or provide some degree of security to networks without central authority such as sensor networks. We propose a Gaussian Mixture Model based semi-supervised learning technique to identify intruders in the network by building a probabilistic model of the wireless channel of the network users. We show that even without having a complete apriori knowledge of the statistics of intruders and users in the network, our technique can learn and update the model in an online fashion while maintaining high detection rate. We experimentally demonstrate our proposed technique leveraging pattern diversity and show using measured channels that miss detection rates as low as 0.1% for false alarm rate of 0.4% can be achieved.

    Original languageEnglish (US)
    Title of host publication2013 IEEE 78th Vehicular Technology Conference, VTC Fall 2013
    DOIs
    StatePublished - Dec 1 2013
    Event2013 IEEE 78th Vehicular Technology Conference, VTC Fall 2013 - Las Vegas, NV, United States
    Duration: Sep 2 2013Sep 5 2013

    Publication series

    NameIEEE Vehicular Technology Conference
    ISSN (Print)1550-2252

    Conference

    Conference2013 IEEE 78th Vehicular Technology Conference, VTC Fall 2013
    CountryUnited States
    CityLas Vegas, NV
    Period9/2/139/5/13

    Fingerprint

    Semi-supervised Learning
    Supervised learning
    Authentication
    Sensor networks
    Statistics
    Network protocols
    Security Protocols
    False Alarm Rate
    Gaussian Mixture Model
    Probabilistic Model
    Sensor Networks
    Update
    Model-based
    Demonstrate
    Statistical Models

    ASJC Scopus subject areas

    • Computer Science Applications
    • Electrical and Electronic Engineering
    • Applied Mathematics

    Cite this

    Gulati, N., Greenstadt, R., Dandekar, K. R., & Walsh, J. M. (2013). GMM based semi-supervised learning for channel-based authentication scheme. In 2013 IEEE 78th Vehicular Technology Conference, VTC Fall 2013 [6692216] (IEEE Vehicular Technology Conference). https://doi.org/10.1109/VTCFall.2013.6692216

    GMM based semi-supervised learning for channel-based authentication scheme. / Gulati, Nikhil; Greenstadt, Rachel; Dandekar, Kapil R.; Walsh, John M.

    2013 IEEE 78th Vehicular Technology Conference, VTC Fall 2013. 2013. 6692216 (IEEE Vehicular Technology Conference).

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

    Gulati, N, Greenstadt, R, Dandekar, KR & Walsh, JM 2013, GMM based semi-supervised learning for channel-based authentication scheme. in 2013 IEEE 78th Vehicular Technology Conference, VTC Fall 2013., 6692216, IEEE Vehicular Technology Conference, 2013 IEEE 78th Vehicular Technology Conference, VTC Fall 2013, Las Vegas, NV, United States, 9/2/13. https://doi.org/10.1109/VTCFall.2013.6692216
    Gulati N, Greenstadt R, Dandekar KR, Walsh JM. GMM based semi-supervised learning for channel-based authentication scheme. In 2013 IEEE 78th Vehicular Technology Conference, VTC Fall 2013. 2013. 6692216. (IEEE Vehicular Technology Conference). https://doi.org/10.1109/VTCFall.2013.6692216
    Gulati, Nikhil ; Greenstadt, Rachel ; Dandekar, Kapil R. ; Walsh, John M. / GMM based semi-supervised learning for channel-based authentication scheme. 2013 IEEE 78th Vehicular Technology Conference, VTC Fall 2013. 2013. (IEEE Vehicular Technology Conference).
    @inproceedings{acf365d3f00f456889bacbb03de5ae00,
    title = "GMM based semi-supervised learning for channel-based authentication scheme",
    abstract = "Authentication schemes based on wireless physical layer channel information have gained significant attention in recent years. It has been shown in recent studies, that the channel based authentication can either cooperate with existing higher layer security protocols or provide some degree of security to networks without central authority such as sensor networks. We propose a Gaussian Mixture Model based semi-supervised learning technique to identify intruders in the network by building a probabilistic model of the wireless channel of the network users. We show that even without having a complete apriori knowledge of the statistics of intruders and users in the network, our technique can learn and update the model in an online fashion while maintaining high detection rate. We experimentally demonstrate our proposed technique leveraging pattern diversity and show using measured channels that miss detection rates as low as 0.1{\%} for false alarm rate of 0.4{\%} can be achieved.",
    author = "Nikhil Gulati and Rachel Greenstadt and Dandekar, {Kapil R.} and Walsh, {John M.}",
    year = "2013",
    month = "12",
    day = "1",
    doi = "10.1109/VTCFall.2013.6692216",
    language = "English (US)",
    isbn = "9781467361873",
    series = "IEEE Vehicular Technology Conference",
    booktitle = "2013 IEEE 78th Vehicular Technology Conference, VTC Fall 2013",

    }

    TY - GEN

    T1 - GMM based semi-supervised learning for channel-based authentication scheme

    AU - Gulati, Nikhil

    AU - Greenstadt, Rachel

    AU - Dandekar, Kapil R.

    AU - Walsh, John M.

    PY - 2013/12/1

    Y1 - 2013/12/1

    N2 - Authentication schemes based on wireless physical layer channel information have gained significant attention in recent years. It has been shown in recent studies, that the channel based authentication can either cooperate with existing higher layer security protocols or provide some degree of security to networks without central authority such as sensor networks. We propose a Gaussian Mixture Model based semi-supervised learning technique to identify intruders in the network by building a probabilistic model of the wireless channel of the network users. We show that even without having a complete apriori knowledge of the statistics of intruders and users in the network, our technique can learn and update the model in an online fashion while maintaining high detection rate. We experimentally demonstrate our proposed technique leveraging pattern diversity and show using measured channels that miss detection rates as low as 0.1% for false alarm rate of 0.4% can be achieved.

    AB - Authentication schemes based on wireless physical layer channel information have gained significant attention in recent years. It has been shown in recent studies, that the channel based authentication can either cooperate with existing higher layer security protocols or provide some degree of security to networks without central authority such as sensor networks. We propose a Gaussian Mixture Model based semi-supervised learning technique to identify intruders in the network by building a probabilistic model of the wireless channel of the network users. We show that even without having a complete apriori knowledge of the statistics of intruders and users in the network, our technique can learn and update the model in an online fashion while maintaining high detection rate. We experimentally demonstrate our proposed technique leveraging pattern diversity and show using measured channels that miss detection rates as low as 0.1% for false alarm rate of 0.4% can be achieved.

    UR - http://www.scopus.com/inward/record.url?scp=84893274194&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84893274194&partnerID=8YFLogxK

    U2 - 10.1109/VTCFall.2013.6692216

    DO - 10.1109/VTCFall.2013.6692216

    M3 - Conference contribution

    SN - 9781467361873

    T3 - IEEE Vehicular Technology Conference

    BT - 2013 IEEE 78th Vehicular Technology Conference, VTC Fall 2013

    ER -