STM Journals Peer-reviewed research from Consortium e-Learning Network

Journal of Telecommunication, Switching Systems and Networks

Convolution Neural Network Model for Intrusion Detection in Network

  1. Shanthi Therese S.
  2. Akash Chhabra
  3. Abhishek Chaturvedi
  4. Shikhar Chaudhary

Abstract

The evolution of the internet has made protecting information a necessity. Network intrusion and prevention plays an integral role in network-based security. The Intrusion technologies primarily used in today’s world deploy various machine learning algorithms and train models based on them resulting in effectively low detection rates. A technical advancement from machine learning, Deep Learning employs complex mechanisms to extract features from samples. As observed that conventional intrusion detection systems utilizing various machine learning algorithms offer a significantly low catching rate, this paper introduces convolution neural network-based network intrusion detection model. In order to classify the intrusion samples in various categories, the proposed model will extract the effective features from the samples offered. Certain experimental runs on the NSL-KDD dataset suggests the immense improvement that can be obtained using CNN methodology.

Keywords

Support