Toward intelligent intrusion prediction for wireless sensor networks using three-layer brain-like learning

Jun Wu*, Song Liu, Zhenyu Zhou, Ming Zhan

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

2 Citations (Scopus)

Abstract

The intrusion prediction for wireless sensor networks (WSNs) is an unresolved problem. Hence, the current intrusion detection schemes cannot provide enough security for WSNs, which poses a number of security challenges in WSNs. In many mission-critical applications, such as battle field, even though the intrusion detection systems (IDSs) without prediction capability could detect the malicious activities afterwards, the damages to the WSNs have been generated and could hardly be restored. In addition, sensor nodes usually are resource constrained, which limits the direct adoption of expensive intrusion prediction algorithm. To address the above challenges, we propose an intelligent intrusion prediction scheme that is able to enforce accurate intrusion prediction. The proposed scheme exploits a novel three-layer brain-like hierarchical learning framework, tailors, and adapts it for WSNs with both performance and security requirements. The implementation system of the proposed scheme is designed based on agent technology. Moreover, an attack experiment is done for getting training and test data set. Experiment results show that the proposed scheme has several advantages in terms of efficiency of implementation and high prediction rate. To our best knowledge, this paper is the first to realize intrusion prediction for WSNs.

Original languageEnglish
Article number243841
JournalInternational Journal of Distributed Sensor Networks
Volume2012
DOIs
Publication statusPublished - 2012

ASJC Scopus subject areas

  • Engineering(all)
  • Computer Networks and Communications

Fingerprint

Dive into the research topics of 'Toward intelligent intrusion prediction for wireless sensor networks using three-layer brain-like learning'. Together they form a unique fingerprint.

Cite this