A Novel Unified Automata Integrated Intrusion Detection Model for Wireless Sensor Network

Authors

  • S. Prithi
  • S. Sumathi

Abstract

To design an efficient intrusion detection system (IDS) various machine learning algorithms such as support vector machine, artificial neural networks, random forest, naïve Bayes and decision trees have been used. In this work, a Hybrid Support Vector Machine – Decision Tree / Random Forest (SVM-DT/RF) based IDS is integrated along with the automata with a view to ameliorate the utilization of energy and lifetime of WSN by detecting the malicious packets and discarding them before the node gets affected. By integrating the IDS with automata, the proposed model improves the detection rate, accuracy as well the energy is efficiently used among the nodes and network lifetime is extended. The proposed automata integrated Hybrid SVM-RF IDS shows an improvement in energy and network lifetime than automata integrated Hybrid SVM-DT, without automata integrated Hybrid SVM-DT IDS and cluster-based IDS.

Downloads

Published

2020-02-10

Issue

Section

Articles