Pengesanan pencerobohan trafik penerowongan IPV6 berasaskan rangkaian neural perambatan balik

Nazrulazhar, Bahaman (2014) Pengesanan pencerobohan trafik penerowongan IPV6 berasaskan rangkaian neural perambatan balik. Masters thesis, UTeM.

[img]
Preview
Text
Pengesanan pencerobohan trafik penerowongan IPV6 berasaskan rangkaian neural perambatan balik.pdf

Download (138kB) | Preview

Abstract

After implementing of IPv6 transition mechanism (TM) as IP catalyst transition has influences the performance on anti-intrusion system. IDS is one of the greatly affected where there is an increasing rate of false alarms generated due to changing patterns of threat. As an action, this thesis proposes new formulation of attack pattern and model lead by implication of TM named 6to4 tunneling. The outcomes from the model are used to develop SPP-RT that focuses on DoS attack that hiding in encapsulating tunneling packet known as protocol-41. The objectives of the research are (i) analysis the performance of IDS on the testbed model, (ii) formulate SeDAP pattern and model, (iii) ability to extract packet features from IPv4 and IPv6 header, and (iv) capable to increase detection rate and decrease false alarm using BPNN. The development of testbed refers to MIT Licoln Lab model was experimented through several testing methods. This accomplishment was done to ensure that the type of equipment and figure will maintain the accurate results. The SPP-RP formed by 3 major parts containing various modules individually. In first, Information Processing consists of capture module, status labeling module and features extraction module. Secondly, Detection part involved with preprocessing module and classification module. The third part called Output with alert notification module. Furthermore, the BPNN method has been integrated into classification module in determining the normal and anomaly status. Then, this BPNN module applied with 3 dissimilar training algorithms named as Levenberg-Marquardt, Bayesian Regulation and Scaled Conjugate Gradient. As a conclusion the highlighted SPP-RP perform capable to recognize tunneling traffic with normal or anomaly packet and a successful tool in reducing false alarm

Item Type: Thesis (Masters)
Uncontrolled Keywords: TCP/IP (Computer network protocol)
Subjects: T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisions: Library > Tesis > FTMK
Depositing User: Noor Rahman Jamiah Jalil
Date Deposited: 06 Jul 2015 05:58
Last Modified: 06 Jul 2015 05:58
URI: http://eprints.utem.edu.my/id/eprint/14689
Statistic Details: View Download Statistic

Actions (login required)

View Item View Item