1. UKM Learning and Research Repository
2. Master Thesis / Tesis Master
3. Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat

Title:	Data traffic measurement and analysis based on scalable features using Hadoop
Authors:	Lena T. Ibrahim (P74156)
Supervisor:	Rosilah Hassan, Assoc. Prof. Dr.
Keywords:	Data traffic Hadoop Internet Network Dissertations, Academic -- Malaysia
Issue Date:	13-Aug-2016
Description:	The potential of Internet traffic measurement and analysis (ITMAA) approaches is to manage the online data transferring processes efficiently. Various approaches and frameworks have been developed for the purpose of estimating and analysing the internet traffic, such as Hadoop. Hadoop is mainly designed to measure network traffic load and segment the data into packet. It also used to analyse network traffic packets based on network layer protocols of different protocols. In order to properly estimate the dynamicity of the transferred data, transferring protocols, and the dynamicity of the traffic, it is essential to consider the characteristics of the transferred data. The problem lays in Hadoop consists of measuring and analysing the Internet traffic using a fixed configuration. Thus, this study aims at designing a scalable ITMAA approach by changing the operation of Hadoop to real time networks. NS2 was used to develop and validate the ITMAA based Hadoop in three network modes (low, medium, and heavy) using client-server network topology on the Internet. The simulation used two main transferring protocol approaches of Transmission Control Protocol (TCP) and User Datagram protocol (UDP). The obtained result showed that the scalable ITMAA approach based Hadoop performed better in TCP than UDP protocol in which the throughput value was higher in TCP than UDP. The same was true for the collision rate in which ITMAA enabled TCP to estimate the client ports in shorter period as compared to UDP. However, delay time of the ITMAA in UDP was found to be better than in TCP. In addition, the scalable ITMAA approach with fixed transferring properties is not recommended. This was reasoned to that the speed of data transfer was affected by transferring properties such as level of network traffic, data segmentation, and the type of transfer protocol.Thus, the Hadoop should be deployed in real time networks.,Certification of Master's/Doctoral Thesis" is not available
Pages:	117
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.