The New Jersey Institute of Technology's
Electronic Theses & Dissertations Project

Title:	Soft fault detection using MIBs in computer networks
Author:	Arora, Sachin
View Online:	njit-etd2003-034 (ix, 33 pages ~ 2.8 MB pdf)
Department:	Department of Electrical and Computer Engineering
Degree:	Master of Science
Program:	Computer Engineering
Document Type:	Thesis
Advisory Committee:	Manikopoulos, Constantine N. (Committee chair) Antoniou, George (Committee member) He, Bin (Committee member)
Date:	2003-05
Keywords:	Computer networks
Availability:	Unrestricted
Abstract:	To improve network reliability and management in today's high-speed communication system, a statistical anomaly network intrusion detection system (NIDS) has been investigated, for network soft faults using the Management Information Base (Mm) traffic parameters provided by Simple Network Management Protocol (SNMP), for both wired and wireless networks. The work done would be a contribution to a system to be designed MIB Anomaly Intrusion Detection, a hierarchical multi-tier and multiobservation-window Anomaly Intrusion Detection system. The data was derived from many experiments that had been carried out in the test bed that monitored 27 MIB traffic parameters simultaneously, focusing on the soft network faults. The work here has been focused on early detection, i.e., detection at low values of the ratio of fault to background traffic. The performance of this system would be measured using traffic intensity scenarios, as the fault traffic decreased from 10% to 0.5% of the background.