Projets de Recherche

In this project, we deal with the QoS of broadband Internet access, the Internet tomography and the routing protocols; all of three oriented to study Internet. They are related as following: The QoS technics will also apply to Internet tomography in order to obtain more information than the topology and to obtain better models, which in turn will be used to test the proposed routing protocols. We will provide a new active method to measure the QoS of broadband Internet access, which measures all the time with a very low load of the channel. In the Internet tomography we will provide a new map with annotations of QoS, and the MPLS tunnels. We also are interested in the IXPs traffic detection. Finally, we will also propose a new External Gateway Protocol for Internet, based on topological considerations and also having a sub-linear table size complexity.

Project ID: 20020130200122BA

Finding new ways to manage the increased data usage and to improve the level of service required by the new wave of applications for smartphones is an essential issue nowadays. The improved understanding of user mobility (i.e. the context they experience) and the content they demand is of fundamental importance when looking for solutions for this problem in the modern communication landscape. The resulting knowledge can help at the design of more adaptable networking protocols or services as well as can help determining, for instance, where to deploy networking infrastructure, how to reduce traffic congestion, or how to fill the gap between the capacity granted by the infrastructure technology and the traffic load generated by mobile users. Building on previous research efforts in the fields of social wireless networking,opportunistic communications, and content networking, the UCOOL project will derive appropriate models for the correlation between user interests and their mobility. Lots of studies have characterized the mobility of mobile nodes based on real world data traces, but knowledge about the interactions with user interests in this context is still missing. Moreover, efficient prediction algorithms will be derived to forecast the node’s mobility as well as interests. Predictions of future events (i.e., related to human mobility or the demanded content) can allow networking protocols or services to anticipate and react upon actions based on patterns learnt from human daily life. The partners involved in the project, located in France, Brazil, Chile, and Argentina, present complementary background and skills to address the context-content correlation and related prediction in mobile wireless networks. The main expected outcome of this project is the establishment and consolidation of a thematic and collaborative research network between the partners to deal with the challenges of leveraging human content and context knowledge. In doing so, we expect new tools and analyses to be developed as well as a new scientific foundation to be established to deal with increasing data usage in wireless mobile networks.

Project ID: 14STIC-08

El presente proyecto tiene como objetivos la medición, modelización y propuesta de nuevos algoritmos de ruteo para Internet. Estas tres actividades están íntimamente relacionadas. Considerando el desarrollo de nuevos algoritmos de ruteo, aquellos capaces de encontrar las rutas dinámicamente a todos los destinos en una red donde se producen cambios, es necesario verificar su funcionamiento en redes artificiales que posean las mismas características que las redes reales. Entonces, para obtener estas redes artificiales es necesario un buen modelo de las reales, siendo que el mismo se basará en la propiedades encontradas en los relevamientos de Internet. Una característica intrínseca de esta actividad es que el modelado de Internet se realiza por aproximaciones sucesivas, es decir se exploran las redes reales, se extraen parámetros, y se proponen modelos que se contrastan con la realidad. Si bien cada modelo tiene un cierto grado de aproximación, es posible refinarlo a través de exploraciones más completas (por ejemplo, agregando a la topología, datos como el ancho de banda o la utilización de los enlaces). En particular, este proyecto proponemos un cierto nivel específico de profundidad: las exploraciones relevarán datos del tráfico de los enlaces además de su conexidad. Claramente estos datos podrán ser utilizados por los algoritmos de ruteo, quienes incluso podrían implementar diferentes rutas según la calidad de servicio (QoS: Quality of Service) deseada. La importancia en la propuesta de nuevos protocolos de ruteo está basada en que el actual protocolo BGP (Border Gateway Protocol) genera tablas de ruteo con un número de entradas similar al número de Sistemas Autónomos (aproximadamente 40000) en los ruteadores centrales. Básicamente las propuestas de protocolos de ruteo existentes generan tablas de ruteo que crecen linealmente con el tamaño de la red. Entonces, proponer protocolos de ruteo cuyas tablas de ruteo crezcan más lentamente, es fundamental para el desarrollo de Internet. Estos protocolos deben ser probados en modelos realistas, con el fin de evaluarlos de manera efectiva. Con el fin de desarrollar capacidades locales en estas temáticas claves para las Tecnologías de Información y Comunicación (TICs), colaboran en este proyecto la Universidad de Buenos Aires, la École Normale Supérieure de Lyon (ENS, Francia), y el Laboratório Nacional de Computação Científica (LNCC, Brasil); capitalizando así los conocimientos sobre sistemas complejos y en redes, líneas de trabajo similares entre los tres grupos que colaboran. Se prevé generar como producto de nuestras investigaciones, además de publicaciones científicas, las implementaciones de los distintos algoritmos desarrollados, que serán puestos a disposición de la comunidad científica utilizando sitios del tipo http://sourceforge.org.

Project ID: 1108

The central role that Internet plays nowadays is undeniable. This project will study Internet's structure through : network tomography (topology and traffic analysis), the proposal of scalable routing protocols, and its analysis from complex systems point of view, mainly oriented to find communities.

Project ID: 20020110200181

The staggering growth of network data technologies at all levels and its applications suggests a wide variety of relevant research lines. In this project we focus on the Internet analysis (its topology and traffic profiling) and on application- oriented routing (how to find paths according to application requirements). In particular, we will analyze Internet from the following three perspectives: the Internet tomography (exploring topology and properties of links), traffic measurements (restricted to a certain application, link or destination), and P2P routing.

Project ID: 20020090200119

This project aims to study computer networks and their data routing in certain areas, with the purpose of improving existing technologies due to their growing influence in actual and future human activity.
In this framework, our project deals with the following areas: the Internet topology and its properties, ad-hoc networks, and the analysis of data traffic from the optics of applications and network resources. Even though all these three areas are huge, our objective is to propose some innovations in each of them, due to that global understanding of the problem helps to find a more robust and realistic solutions.
In particular we will: to continue the topology analysis using k-core decomposition (developing the LaNet-vi software, http://sourceforge.net/projects/lanet-vi), to develop the ad- hoc network protocol ANTop (analyzing and simulating it by QUENAS, http://sourceforge.net/ projects/quenas), and finally to model traffic, sorted by application type, and relating the model to the quality of service QoS.

Project ID: I413