Towards an Efficient Anomaly-Based Intrusion Detection for Software-Defined Networks

  • Majd Latah, Levent Toker
  • IET Networks, August 2018, the Institution of Engineering and Technology (the IET)
  • DOI: 10.1049/iet-net.2018.5080

Supervised machine-learning for SDN-based intrusion detection task.

What is it about?

The integration between software-defined networking and machine learning helps in creating more secure networks. Too many machine learning algorithms can be used for intrusion detection task. Our study focuses on comparing well-known supervised machine learning approaches for building an efficient SDN-based intrusion detection system.

Why is it important?

Our work provides a detailed analysis as the well-known supervised machine-learning algorithms were compared in terms of accuracy, false alarm rate, precision, recall, F1-measure, area under ROC curve, Mc Nemar's test, and the time taken to train and test the model. In addition, we employed the Principle Component Analysis (PCA) approach for feature selection and dimensionality reduction which PCA helped in enhancing the accuracy level from 80.31% to 88.74% in compared with using the basic features provided by the SDN controller.

Perspectives

Majd Latah
Ozyegin Universitesi

This is the first study that provides a detailed analysis for supervised machine-learning-based intrusion detection in SDNs.

Read Publication

http://dx.doi.org/10.1049/iet-net.2018.5080

The following have contributed to this page: Majd Latah