http://hdl.handle.net/2268.2/1963 Thu, 05 Mar 2026 22:29:38 GMT 2026-03-05T22:29:38Z http://hdl.handle.net/2268.2/8397 Title: Making applications aware of IPv6 multi-prefix networking Abstract: The IPv6 protocol provides, besides its 128-bit locator, a lot of interesting new networking principles to the Internet. One very promising concept is called Provisioning Domains (PvD). It allows to regroup different networking configuration elements together. The goal of this paper is to make applications aware of PvD principles in order to show its usefulness throughout different use cases. Two open-source projects, the VideoLan VLC media player as well as the GNOME Web browser, have been successfully extended with features, which allow the user to interact with PvDs in different ways. But, in order to modify these applications, the appropriate tools and development infrastructure are needed. Most of the tools are already provided through previous works, but the collection has been expanded by a new tool allowing the collection of network statistics in function of Provisioning Domains. A new development environment had to be created as well, to effectively probe the new developed features. But, in order to understand the functioning of those features, a good understanding of IPv6 and PvD principles is needed. Thus, this document starts off by giving a general description of some of their concepts. Sun, 08 Sep 2019 22:00:00 GMT http://hdl.handle.net/2268.2/8397 2019-09-08T22:00:00Z http://hdl.handle.net/2268.2/6757 Title: Implementation of a Platform for Computer Security and Network Labs Abstract: Networking is a domain with a strong growth due to the interest of both industries and individuals. It leads to an high increase of the demand of expert in the domain. More and more courses at the University of Liège tackles parts this subject from basic concepts to advanced ones such as network security. To be prepared for active life, the students have to attend practical labs. However, it is not possible to provide physical devices for each student, therefore, software are used to emulate the networks. As the currently used software are quite old and new courses with new requirements are created, the purpose of this thesis is to find a software with additional features to handle both old and new labs. The CORE emulator has be selected from a pool of fifty software because it is an lightweight and up-to-date emulator, which comes with multiple features. It has been installed in a virtual environment to be port easily on various machines, then a plugin has been implemented to increase its user-friendliness. Finally, the network scenarios of several courses had been ported on the CORE software to identify its advantages, drawbacks and limitations and to provide a pool of application examples for further scenarios. This work has highlighted the panel of possibilities and features offered by the CORE emulator which can be compared to the ones of the software that are currently used. Moreover, it has shown how easy it is to create a plugin for the software in order to modify and customise it. Tue, 25 Jun 2019 22:00:00 GMT http://hdl.handle.net/2268.2/6757 2019-06-25T22:00:00Z http://hdl.handle.net/2268.2/5513 Title: Transport protocol for machine/deep learning Abstract: Datasets for machine learning can be very large in size, and are usually transferred to dedicated servers that have the computing power necessary for complex model training. This transfer is done losslessly by TCP, but could allow some losses without affecting too much the quality of the trained model, which would make the transfer faster. The first part of this work looks at how packet loss on different datasets affect the accuracy of the corresponding model, and how the different ways to deal with packet loss compare to one another. It finds that it is indeed possible to lose packets during the transfer without affecting the model too much, but it requires some data preparation to repair the files if parts of it are lost, and some parts cannot be lost and are absolutely necessary to the files. Using those results, a protocol is designed and implemented to transfer datasets partly by TCP and partly by UDP depending on which parts of the files are required or can be lost without too many consequences. Focus is mainly set on JPEG files. Tests then show that the protocol can be indeed be faster, but parameters need to be set to the right values or the ratio speed/loss becomes too bad. Wed, 05 Sep 2018 22:00:00 GMT http://hdl.handle.net/2268.2/5513 2018-09-05T22:00:00Z http://hdl.handle.net/2268.2/4582 Title: Web-scale Network Telemetry for live user traffic: In-band OAM for the Linux Kernel Abstract: The measurement methods are important in the network to monitor it.Companies having a hypergiant CDN or a big data center need some mecha-nisms to detect an issue and quickly find its origin to be able to fix it. Indeed,due to some issue, it is possible that a whole part of the service proposed by acompany will be unavailable during a certain amount of time. Consequently,it is normal to see the apparition of new methods which fix some issue likeParis traceroute for traceroute, that are more efficient or allow to collect moreinformation. One of them which allows to collect more information, is thesubject of this thesis. It is iOAM.This work tempts to put the basics needed by iOAM in the Linux kernel.Indeed, not all the possibilities bring by this method are implemented duringthis thesis but now, there is a starting point to continue the development of itin the Linux kernel. The code of the implementation made during this masterthesis can be retrieved on Git Hub (https://github.com/FrancartJ/iOAM-IPv6-LinuxKernel). It allows a node having Linux as anoperating system to understand and perform the appropriate actions whenit receives an IPv6 packet containing an iOAM pre-allocate trace inside theHop-By-Hop options header.The performances of this implementation are quite promising. They high-light the link between the percentage of packet which will receive the iOAMtrace and the throughput. Also they put on the way of possible improve-ments for the recording of data. Sun, 24 Jun 2018 22:00:00 GMT http://hdl.handle.net/2268.2/4582 2018-06-24T22:00:00Z