www.uhasselt.be
DSpace

Document Server@UHasselt >
Education >
Faculty of Engineering Technology >
Master theses >

Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/19522

Title: Software-Defined Networking for Multi-Camera Systems
Authors: Rymen, Martijn
Advisors: CLAESEN, Luc
CHEN, Chien
Issue Date: 2015
Publisher: UHasselt
Abstract: Intensively interacting multi-camera systems can have the requirement to be dynamically reconfigurable and to have an easy control of the network. Therefor, this work investigates the possibilities of Software-Defined Networking (SDN). SDN offers a flexible way of man- aging networks while proving a great scalability. In SDN the data plane and control plane are decoupled, enabling this dynamic control. First, multi-camera networks and SDN have been researched. Afterwards, a use case is defined to validate the concept of using SDN in a multi-camera network. This use case is a soccer field, where it can be employed in demand driven 3D image calculations. For this reason, links needs to be created between adjacent cameras, allowing them to work jointly on common tasks. These links need to be created dynamically, because a tracked object can move over the entire field. This demands a 3D image from other locations on the field. To create this use case, a network controller is used. This allows Java programming and the development of a REST API. With this API it is possible to alter the location of a tracked item, while the controller module can connect to a SQL database to retrieve the cameras that have to work together on the task. To prove that the controller is only forwarding the necessary traffic and blocking the other traffic, the network is emulated with Mininet. In this way, superfluous traffic is avoided on the network. This concludes that SDN is a dynamic way to control a multi-camera network.
Notes: master in de industriĆ«le wetenschappen: elektronica-ICT
URI: http://hdl.handle.net/1942/19522
Category: T2
Type: Theses and Dissertations
Appears in Collections: Master theses

Files in This Item:

Description SizeFormat
N/A6.93 MBAdobe PDF
N/A972.16 kBAdobe PDF

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