Le routage Internet repose sur le partage d'informations entre les différents équipements constituant le coeur du réseau. Ces équipements communiquent grâce au protocole BGP afin de réagir en cas de modifications du réseau (modification de la structure du réseau, pannes. . . ). Dès lors, il est possible de router le trafic vers n'importe quelle destination de l'Internet. Aujourd'hui le protocole BGP posent un certain nombre de problèmes au sein des réseaux d'opérateurs. Ces problèmes sont généralement dus à une mauvaise diffusion des informations de routages dans le réseau. Dans ce genre de situation le trafic peut être amené à suivre des chemins sous-optimaux ou inconsistants. Qui plus est, ces comportements inattendus peuvent être particulièrement difficiles à détecter ou à anticiper. Pour répondre à cette problématique, on se propose d'aborder le protocole BGP comme suit. – Quels problèmes pose aujourd'hui le protocole BGP? Comment modéliser la diffusion des informations de routage ? – Comment valider la structure BGP d'un réseau ? Comment prévoir le comportement d'un réseau ? – Comment concevoir une structure BGP permettant de garantir que le comportement du réseau est valide, y compris en cas de panne ? – Comment modifier le protocole BGP afin d'avoir un réseau ayant toujours le comportement attendu ? Ce mémoire met en évidence les limites du protocole actuel. Il propose des outils de validation, de simulation et de conception BGP dans ce contexte. Il propose enfin une évolution simple et réalisable du protocole BGP permettant d'aboutir à un réseau fiable, efficace et simple à configurer.

During the last years, MultiProtocol Label Switching (MPLS) has been deployed by most large Service Providers (SP). The main driver for MPLS deployment is the ability to provide new services with stringent Service Level Agreements (SLAs) such as layer-2 and layer-3 Virtual Private Networks (VPNs) as well as Voice and Video over IP. Most of these services are already deployed inside single SP networks. However, customers now require world-wide VPN and VoIP services. Therefore, SPs need to collaborate to offer these services across multiple SP networks. Inside a single SP network, each node usually knows the complete topology of the network with the load and delay of all the links. Based on this information, each router is able to compute constrained paths toward any other router inside the SP network. Then, it can establish a connection and reserve resources along the computed path with the Resource reSerVation Protocol (RSVP-TE). However, when services with stringent requirements must cross multiple SP networks the computation of the path becomes a problem. Routers in different SP networks exchange routing information by using the Border Gateway Protocol (BGP). BGP provides reachability information. It does not distribute complete topology, delay and bandwidth information. One way to provide guaranteed services crossing different SPs is to delegate the computation of the paths to a Path Computation Element (PCE) that learns the topology of the different SPs. However, this requires that SPs reveal information that they usually consider confidential, their topology. In this thesis, we perform active measurements to show the difficulty to engineer the interdomain traffic with BGP. MPLS together with RSVP-TE provide much more control on the traffic. We define extensions to RSVP-TE for the protection of inter-AS MPLS paths. The aim is to be able to provide the same service guarantees as inside a domain while keeping the internal topology of SPs confidential, as required by SPs. We propose and evaluate distributed techniques relying on PCEs for the computation of interdomain constrained paths respecting the latter confidentiality requirement.

The Internet traffic is constantly increasing following the emergence of new network applications like social networks, peer-to-peer, IP phone or IP television. In addition, these new applications request better path availability and path quality. Indeed the efficiency of these applications is strongly related to the quality of the underlying network. In that context network operators make use of traffic engineering techniques in order to improve the quality of the routes inside their network, but also to reduce the network cost of increased traffic handling with a better utilization of existing resources. This PhD thesis covers several topics of Traffic Engineering and Fast Restoration in IP/MPLS networks. Our first contribution is related to the definition of a well-engineered network. In the literature mathematical formulation of Traffic Engineering (TE) requirements are very diverse. We have thus performed a comparative study of many objective functions, in order to differentiate them and choose in a rational way the one that best reflects Traffic Engineering goals. We have also designed a method approaching optimal TE, whereby we divide the traffic matrix in N sub-matrices and route them independently, based on the derivatives of the objective function. The second topic addressed in this work concerns link weight optimizers (LWOs). Link weight optimization is the traffic engineering {it "standard"} technique in networks running link state routing protocols (which are widely used in transit networks). These link weight optimizers suffer from several limitations due to the BGP (Border Gateway Protocol) Hot-Potato rule, which is basically not considered by such optimizers. Therefore we have proposed a BGP-aware link weight optimization method that takes problematic Hot-Potato effects into account, and even turns them into an advantage. We have also studied how LWOs behave in big networks which have to use BGP route reflectors. Finally we have studied whether forwarding loops can appear or not when traffic is split among multiple equivalent egress routers, an optional BGP feature that we did use in our Hot-Potato aware LWO. Our last contribution concerns network resilience. We have proposed a solution for a rapid recovery from a link or node failure in an MPLS network. Our solution allows a decentralized deployment combined with a minimal bandwidth usage while requiring only reduced amount of information to flood in the network. This method is the first that makes possible a decentralized deployment combined with an optimal resource consumption. To easily simulate and test the methods proposed in this work, we have also contributed to the development of TOTEM - a TOolbox for Traffic Engineering Methods.

Made available in DSpace on 2015-04-11T14:02:50Z (GMT). No. of bitstreams: 1 ricardo.pdf: 1126811 bytes, checksum: 85736cdaf1e7cf53e93c63df8abb6b16 (MD5) Previous issue date: 2013-07-29 / CNPq - Conselho Nacional de Desenvolvimento Científico e Tecnológico / The Internet is organized on network groups managed by administrative domains known as Autonomous Systems (ASes). Each AS employs its own routing policies and has autonomy in comparison to other ASes. The comunication and coordination between these ASes is made possible by the interdomain routing protocol. The Border Gateway Protocol (BGP)is the interdomain routing protocol currently used on Internet. However, the Internet s interdomain routing architecture has undergone only minor changes since its inception. It presents issues difficult to solve, because of the barrier to deploy new features, the hardness to understand its behavior and dynamics, and complexity to identify and correct faults. The deployment on Internet architecture is a tough task, because the need to deploy it directly on routers. Besides that, there is a requirement of global acceptance of new protocols and modifications. We observed that Software Defined Networking (SDN) could be used to provide innovation on interdomain routing, but it lacked mechanisms support this approach. SDN is an emerging paradigm composed by a data plane, a control plane, and an open protocol. On this work we present a mechanism able to perform interdomain routing with domains that deploys SDN paradigm, called Inter-SDN Routing Component. We point out the main issues on current interdomain routing, describes the Inter-SDN component, its behavior, and its experimental evaluation. Besides that, we show how prototyping and building of an interdomain mechanism on SDN are tasks relatively simple, and explain how our solution takes advantage of the SDN features to address the issues of the interdomain routing. / A Internet está organizada em grupos de redes que são gerenciados por domínios ad- ministrativos conhecidos como Sistemas Autônomos (ASes Autonomous Systems). Cada AS emprega suas próprias políticas de roteamento e tem autonomia em relação a outros ASes. A comunicação e coordenação entre estes ASes acontece por meio do protocolo de roteamento interdomínio. O Protocolo de Roteamento de Borda (BGP Border Gateway Protocol) é o protocolo de roteamento interdomínio atualmente utilizado na Internet. Entretanto, a arquitetura do roteamento interdomínio da Internet tem sofrido poucas mudanças desde sua criação. Ela apresenta problemas complexos de serem resolvidos, devido à dificuldade para a implantação de novas soluções, à dificuldade de entender seu comportamento e dinâmica, e à complexidade de identificar e corrigir falhas. A implantação de novas funcionalidades na arquitetura da Internet é uma tarefa difícil, devido à necessidade de manipular diretamente todos os roteadores. Além disso, é necessário aceitação global de novos protocolos e modificações nos existentes. Nós observamos que a abordagem Redes Definidas por Software (SDN Software-Defined Networking) poderia ser usada para prover inovação no roteamento interdomínio, mas que faltavam mecanismos para suportar esse tipo de roteamento. Neste trabalho, nós apresentamos um mecanismo capaz de realizar o roteamento interdomínio entre domínios que utilizam o paradigma SDN, chamado Inter-SDN. Nós ressaltamos os principais problemas no modelo atual de roteamento interdomínio, descrevemos o componente Inter-SDN, seu comportamento, e apresentamos uma análise experimental. Além disso, nós mostramos como o projeto e a construção de um mecanismo de roteamento interdomínio em SDN são tarefas relativamente simples, e explicamos como nossa solução utiliza as vantagens do SDN para resolver problemas do roteamento interdomínio.

Diploma thesis deals with creation of laboratory exercises in ns-3 environment. Each one of three exercises consists of theoretical introduction and instructions to carry out the simulation. The first exercise´s topic is routing protocol BGP. The second exercise is focused on transport protocols TCP, UDP, SCTP. In the last exercise, the basic network devices and topologies are simulated. The ARP and RIPv2 protocols are simulated as well.

This document specifies a simple signalling mechanism for inter-domain QoS marking using a BGP Extended Community QoS Attribute. Class based packet forwarding for delay and loss critical services is currently performed in an individual AS internal manner. The new QoS marking attribute makes the QoS class setup within the IP prefix advertising AS known to all access and transit ASes. This enables individual (re-)marking and forwarding treatment adaptation to the original QoS class setup of the respective IP prefix. The attribute provides the means to signal QoS markings on different layers, which are linked together in QoS class sets. It provides inter-domain and cross-layer insight into the QoS class mapping of the source AS with minimal signalling traffic.

The size and speed of computer networks continue to expand at a rapid pace, as do the corresponding errors, failures, and faults inherent within such extensive networks. This thesis introduces a novel approach to interface Border Gateway Protocol (BGP) computer networks with neural networks to learn the precursor connectivity patterns that emerge prior to a node failure. Details of the design and construction of a framework that utilizes neural networks to learn and monitor BGP connection states as a means of detecting and predicting BGP peer node failure are presented. Moreover, this framework is used to monitor a BGP network and a suite of tests are conducted to establish that this neural network approach as a viable strategy for predicting BGP peer node failure. For all performed experiments both of the proposed neural network architectures succeed in memorizing and utilizing the network connectivity patterns. Lastly, a discussion of this framework's generic design is presented to acknowledge how other types of networks and alternate machine learning techniques can be accommodated with relative ease.

本篇論文建立一般化（與 Bond et al. (1994)比較）兩部門內生性成長模型，其中實質資本與人力資本之生產技術是固定規模報酬。利用傳統靜態國貿理論２Ｘ２模型的特性及結合跨期無套利條件，我們証明出在下列三種情形下(i) 0<α<1 (ii)α=1 (iii) α=0，均衡成長是馬鞍均衡穩定，要素密集度之大小不會影響這項推論。最後，我們仔細研究動態調整的過程並且討論時間偏好及支出比改變的長期效果。 / In this paper, we develop a "more" (compared with Bond et al. (1994)) general two-sector endogenous growth model with constant-return-to-scale production technologies governing the evolution of physical and human capital. 　　By utilizing the properties of traditional static 2x2 inter- national trade model and combining an intertemporal no-arbitrage condition which links the capital gain on human capital to the difference between the rentals on capital and wage rate, we prove that the balanced growth equilibrium is saddle-path stable in three cases (i) 0<α<1 (ii)α=1 (iii) α=0 regardless of the factor intensity ranking. 　　Finally, we provide a detail characterization of transitional dynamics and discuss the long-run effects of changes in time pre-ference and the expenditure share,α.

<p>This thesis aims to investigate a method and tool for emulation of the General Packet Radio Service (GPRS) core network needed as an environment to test the routing functionality. GPRS is the most widely adopted mobile packet data delivery technology in the world. It utilizes an Intranet Protocol (IP)-based core network and involves significant changes to the way the Global System for Mobile communications (GSM) air interface is structured. It also forms the basis of the future structure of mobile network transmission and switching.</p><p>The Serving GPRS Support Node (SGSN) is the most fundamental node in GPRS. Ericsson produces and manages an increasing number of SGSN nodes in the world. One of main functionalities of SGSN node is to forward IP packets according to the destination address in the IP header on IP core network.</p><p>In each new release of SGSN, or when implementation or upgrades have been done on routing application on SGSN, design and test engineers at Ericsson need to emulate the IP core network. This must be done with use of many routers to generate huge amounts of data that can simulate the real world IP core network.</p><p>The major goal of this thesis was to analyze and verifying the use of a suitable and economical solution to emulating IP Core Network of the GPRS system for testing of different functionality of the routing application running in SGSN , instead of building up a physical Core Network with different infrastructure and many routers.</p><p>The method chosen for emulating the IP core network with many routers, and investigated in the thesis, is based on a Cisco simulator called “Dynamips”, which runs many actual Cisco Internetwork Operating Systems (IOS) with many different models of Cisco products in a virtual environment on Windows or Linux platforms. With this simulator, engineers at Ericsson will be able to use this simulator to emulate IP core network easily and efficiently to accomplish system test cases.</p><p>A conclusion of this work is that Dynamips could be used to emulate many complicated IP core network scenarios, with many routers to generate huge amounts of data to simulate the real world IP core network. The emulated system fulfils its purpose for testing of the routing application of SGSN regarding different functionality and characteristics. This is done to ensure and verify that SGSN routing application meets its functional and technical requirements, and also helps to find undiscovered errors as well as helps to ensure that the individual components of routing application on SGSN are working correctly.</p>

This thesis aims to investigate a method and tool for emulation of the General Packet Radio Service (GPRS) core network needed as an environment to test the routing functionality. GPRS is the most widely adopted mobile packet data delivery technology in the world. It utilizes an Intranet Protocol (IP)-based core network and involves significant changes to the way the Global System for Mobile communications (GSM) air interface is structured. It also forms the basis of the future structure of mobile network transmission and switching. The Serving GPRS Support Node (SGSN) is the most fundamental node in GPRS. Ericsson produces and manages an increasing number of SGSN nodes in the world. One of main functionalities of SGSN node is to forward IP packets according to the destination address in the IP header on IP core network. In each new release of SGSN, or when implementation or upgrades have been done on routing application on SGSN, design and test engineers at Ericsson need to emulate the IP core network. This must be done with use of many routers to generate huge amounts of data that can simulate the real world IP core network. The major goal of this thesis was to analyze and verifying the use of a suitable and economical solution to emulating IP Core Network of the GPRS system for testing of different functionality of the routing application running in SGSN , instead of building up a physical Core Network with different infrastructure and many routers. The method chosen for emulating the IP core network with many routers, and investigated in the thesis, is based on a Cisco simulator called “Dynamips”, which runs many actual Cisco Internetwork Operating Systems (IOS) with many different models of Cisco products in a virtual environment on Windows or Linux platforms. With this simulator, engineers at Ericsson will be able to use this simulator to emulate IP core network easily and efficiently to accomplish system test cases. A conclusion of this work is that Dynamips could be used to emulate many complicated IP core network scenarios, with many routers to generate huge amounts of data to simulate the real world IP core network. The emulated system fulfils its purpose for testing of the routing application of SGSN regarding different functionality and characteristics. This is done to ensure and verify that SGSN routing application meets its functional and technical requirements, and also helps to find undiscovered errors as well as helps to ensure that the individual components of routing application on SGSN are working correctly.