Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Bayesian neural networks for internet traffic classification.

Tom Auld1, Andrew W Moore, Stephen F Gull

  • 1Department of Physics, University of Cambridge, Cambridge CB3 OHE, UK. tauld@mrao.cam.ac.uk

IEEE Transactions on Neural Networks
|February 7, 2007
PubMed
Summary

This study introduces a novel internet traffic classifier. It accurately identifies traffic types using only packet headers, enhancing network management and security without needing packet content.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

From photons to big-data applications: terminating terabits.

Philosophical transactions. Series A, Mathematical, physical, and engineering sciences·2016
Same author

Hydrocodone in postoperative personalized pain management: pro-drug or drug?

Clinica chimica acta; international journal of clinical chemistry·2013
Same author

Reconstruction of Calmodulin Single-Molecule FRET States, Dye-Interactions, and CaMKII Peptide Binding by MultiNest and Classic Maximum Entropy.

Chemical physics·2013
Same author

Classic maximum entropy recovery of the average joint distribution of apparent FRET efficiency and fluorescence photons for single-molecule burst measurements.

The journal of physical chemistry. B·2012
Same author

Algorithms for rapid outbreak detection: a research synthesis.

Journal of biomedical informatics·2005
Same author

Automated syndromic surveillance for the 2002 Winter Olympics.

Journal of the American Medical Informatics Association : JAMIA·2003

Area of Science:

  • Computer Science
  • Network Engineering
  • Machine Learning

Background:

  • Accurate internet traffic identification is crucial for network management, demand prediction, security anomaly detection, and realistic traffic modeling.
  • Current methods often require access to packet content or specific host/port information, limiting their applicability.

Purpose of the Study:

  • To develop a highly accurate internet traffic classifier that does not require source/destination host-address, port information, or packet content.
  • To enable wider application of traffic identification techniques by utilizing only packet header information.

Main Methods:

  • Utilized supervised machine learning, specifically a Bayesian trained neural network.
  • Trained and tested the classifier using features derived from packet streams, focusing on packet headers.

Related Experiment Videos

  • Leveraged categorized training data derived from packet content for feature engineering.

    • Main Results:

    • Achieved high accuracy in classifying diverse internet traffic types.
    • Demonstrated effective classification using only packet header information, without inspecting packet content.
    • The developed technique offers a significant advantage over methods requiring full packet inspection.

    Conclusions:

    • The proposed traffic classifier provides a powerful and broadly applicable solution for network management and security.
    • By relying solely on packet headers, this method overcomes limitations of content-dependent classification techniques.
    • This approach facilitates more efficient and privacy-preserving internet traffic analysis.