Publication Type:

Conference Paper

Source:

Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012, Liverpool, p.1264-1269 (2012)

ISBN:

9780769547497

URL:

http://www.scopus.com/inward/record.url?eid=2-s2.0-84870433494&partnerID=40&md5=1023a88556508b756b707b368017ac6d

Keywords:

Adverse result, Complete system, Computer crime, Critical parameter, Denial of Service, DoS attacks, Embedded software, Machine learning techniques, Medium access control, Medium access control layer, Neural networks, probability, Real time systems, Security, Security enhancements, Security flaws, Security of wireless sensor networks, Support vector machines, Transmission control protocol, Wireless sensor networks

Abstract:

Ensuring the security of wireless sensor networks (WSNs) is vital for monitoring real-time systems. One of the major security flaws experienced by WSNs is denial of service (DoS) which can even lead to the breakdown of the complete system or to wrong decisions being made by the system that can cause adverse results. This research work focuses on two techniques for detecting a DoS attack at a medium access control (MAC) layer. Our research compares and evaluates the performance of two major machine learning techniques: neural network (NN) and support vector machine (SVM). Vanderbilt Prowler is used for simulating the scenarios. In the simulations, normalized critical parameters and their corresponding probabilities of DoS attack are computed in 50 trial runs. These normalized critical parameters and their corresponding probabilities of DoS attack are used as training inputs in NN and SVM approaches. The simulation results clearly show that SVM provides better accuracy compared to NN, 97% accuracy by SVM and 91% accuracy by NN. The simulation also shows that SVM takes much less time to detect and determine the probability of a DoS attack, 0.25 seconds by SVM and 0.75 seconds by NN. All these results clearly show that SVM performs better than NN when used for detecting the probability of DoS attack in WSNs. © 2012 IEEE.

Notes:

cited By (since 1996)0; Conference of org.apache.xalan.xsltc.dom.DOMAdapter@46c3e8a ; Conference Date: org.apache.xalan.xsltc.dom.DOMAdapter@5e7dd84a Through org.apache.xalan.xsltc.dom.DOMAdapter@180312bd; Conference Code:94116

Cite this Research Publication

A. B. Raj, Dr. Maneesha V. Ramesh, Kulkarni, R. V., and Hemalatha, T., “Security enhancement in wireless sensor networks using machine learning”, in Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012, Liverpool, 2012, pp. 1264-1269.

It appears your Web browser is not configured to display PDF files. Download adobe Acrobat or click here to download the PDF file.

207
PROGRAMS
OFFERED
5
AMRITA
CAMPUSES
15
CONSTITUENT
SCHOOLS
A
GRADE BY
NAAC, MHRD
9th
RANK(INDIA):
NIRF 2017
150+
INTERNATIONAL
PARTNERS