Pendant les dernières années, le réseau cellulaire est devenu accessible presque partout. Pour compléter ce réseau cellulaire terrestre, plusieurs systèmes basés sur des satellites à basse orbite (LEO Low Earth Orbit) et moyenne orbite (MEO Medium Earth Orbit) ont été développés pour offrir une couverture globale. Les services multimédias sont largement demandés sur une échelle globale.<br />Notre but est de développer une couche d'accès MAC au canal satellitaire tout en respectant les contraintes de la constellation LEO. Un grand nombre d'utilisateurs de différents types vont essayer d'accéder au canal satellite (sens montant) afin d'envoyer leurs paquets. Il faut distribuer les ressources disponibles entre les différents utilisateurs de façon à respecter la QoS de chaque utilisateur et à minimiser le gaspillage. Nous avons traité dans cette thèse les ponts suivants :<br />·Nous avons adopté les techniques CDMA (Code Division Multiple Access) et TDMA (Time Division Multiple Access) pour définir les protocoles d'accès proposés en système LEOS. Ces techniques sont adaptées afin d'être utilisées d'une manière efficace dans le cas spécifique de satellites LEO (long RTD (Round Trip Delay), Handover, ).<br />·Nous avons utilisé un contrôle d'admission pour les différents types de services qui agit au niveau d'appel afin d'accepter un maximum d'utilisateurs de différents types tout en respectant la QoS de chacun. Ce contrôle utilise une fonction à capacité équivalente adaptée aux différents contextes de nos études.<br />·Nous avons défini des services au niveau MAC d'une façon générique. Les services sont supportés au niveau MAC par des MTCs (MAC Transfer Capabilities) afin de garantir la QoS demandée par chaque service. Nous avons proposé des MTCs qui supportent tous les services et applications possibles.<br />·Nous avons étudié le système pour des utilisateurs hétérogènes ce qui rend l'intégration de services nécessaire. La comparaison dépend de l'intégration et de la QoS de chaque utilisateur. Le but est toujours de maximiser la capacité totale du système.<br />·Nous avons étudié le multiplexage de services et l'allocation de ressources pour les différents protocoles d'accès. Des techniques d'ordonnancement et de partage de ressources sont proposées afin d'atteindre pour un protocole d'accès donné la performance maximale.<br />L'intégration de satellite avec l'UMTS et l'utilisation de l'ATM au sein du réseau de cur sont deux facteurs essentiels qui dirigent nos propositions sur la couche MAC. Les services vus par cette couche sont appelés les MTCs (MAC Transfer Capabilities) et doivent se coordonner avec les services de l'UMTS et l'ATM. Le choix est basé aussi sur les sources de trafics utilisées dans les différents chapitres.

El gran crecimiento y expansión de Internet en los últimos años, con el consecuente incremento de usuarios y tráfico, ha hecho que aumente la necesidad de ancho de banda en las redes de telecomunicación actuales. El desafío de la futura generación de redes de telecomunicación punta a pasar de la simple transmisión de señales ópticas de gran capacidad a efectivamente conmutar y gestionar esta cantidad de datos en el domino óptico. Estas funcionalidades, actualmente realizadas por componentes eléctricos, son las que actualmente causan un cuello de botella en la escalabilidad y crecimiento de Internet. Por otro lado, se espera que las futuras redes transporten servicios heterogéneos que incluyen tanto transferencia de datos como transmisión de aplicaciones multimedia e interactivas. Cada servicio por lo tanto necesita un requerimiento y tratamiento particular (por ejemplo garantir un limite en el retraso extremo-extremo o en el ancho de banda). En este entorno, proporcionar calidad de servicio (Quality of Service, QoS) resulta ser un factor obligatorio. En este ámbito el objetivo de esta tesis es el desarrollo de nuevos mecanismos para proporcionar QoS en redes de conmutación de paquetes ópticos (Optical Packet Switching, OPS) tanto en entorno metropolitano como de área extendida. OPS pretende ser una solución a largo plazo en cuanto requiere conmutadores de alta velocidad y componentes ópticos avanzados como los conversores sintonizables de longitud de onda y regeneradores completamente ópticos. Por lo que concierne el entorno metropolitano, las redes son generalmente sin memoria, en el sentido que una vez transmitida la información a la red, esa se queda en el dominio óptico sin encontrar ninguna cola en el camino hasta alcanzar su destino. Para evitar contenciones, protocolos de acceso al medio compartido (Medium Access Control, MAC) son necesarios y pueden integrase con mecanismos para proporcionar QoS. De este concepto se pueden diseñar varias arquitecturas distintas. En esta tesis nos concentramos en dos arquitecturas basadas en topologías compuestas, llamadas respectivamente redes multi-PON y redes multi-anillos. Ambas han sido desarrolladas en el proyecto de investigación DAVID financiado por la Unión Europea dentro del quinto programa marco. Nuestras contribuciones abarcan varios aspectos. Antes de todos se han identificado los requerimientos de las futuras redes metropolitanas basadas en OPS, con particular atención en determinar los servicios necesarios de acuerdo con lo que se ha diseñado en la anterior generación eléctrica de redes metropolitanas. Luego para ambas redes se ha seguido el mismo procedimiento. La primera etapa ha sido la evaluación de prestaciones a través de simulaciones con el objetivo de identificar los puntos débiles. Se ha luego pasado a la fase de optimización tanto de la arquitectura de las redes como de los mecanismos que gobiernan su funcionamiento y se han verificado las mejoras. Finalmente se han propuesto mecanismos para proporcionar QoS según los requerimientos definidos anteriormente y se ha hecho un estudio de coste/prestaciones comparando las dos arquitecturas con otras actualmente en comercio como SDH, RPR y Ethernet.Por lo que concierne el entorno de área extendida, se ha considerado una red de conmutación de paquetes ópticos orientado a la conexión donde los nodos tienen limitadas capacidad de encolamiento. En este contexto, se han tratado dos problemáticas: el establecimiento de las conexiones virtuales ópticas (Optical Virtual Circuit, OVC) configurando propiamente las tablas de expedición (forwarding table) en los nodos y la provisión de QoS.Para el primer punto, a la llegada de una petición de establecimiento de una nueva OVC, cada nodo debe asignar un puerto y una longitud de onda de salida a esta OVC. Mientras la elección del puerto de salida depende de los algoritmos de routing, la elección de la longitud de onda se puede decidir localmente en cada nodo según diferente políticas llamadas OVC-to-wavelength setup assignment (OWSA). En esta parte de la tesis se ha estudiado en detalle este problema y se han propuesto diferentes estrategias. En particular se ha demostrado que una buena política de asignación de longitud de onda incrementa notablemente la prestación de un conmutador OPS. Se ha usado la idea de agrupar, siempre que se pueda, los flujos de tráfico que entran en el conmutador por el mismo puerto y misma longitud de onda de manera de disminuir lo máximo posible la probabilidad de convención entre paquetes. Por lo que concierne proporcionar QoS, el estado del arte indica que hasta el momento se ha siempre seguido la misma técnica basada en: 1) diseñar un algoritmo de resolución de contenciones que minimice la probabilidad de pérdidas de paquetes (Packet Loss Rate, PLR) y luego 2) aplicar un mecanismo de reserva de recursos capaz de diferenciar la PLR entre dos o mas clases de tráfico. Considerando que el entorno de estudio es orientado a la conexión, se ha propuesto un enfoque diferente basado en el esquema aplicado en redes ATM donde se definen diferentes categorías de servicio, cada una con su propio tratamiento dentro de la red. En particular se han definido 3 categorías de servicio para entorno OPS y se han desarrollado 3 algoritmos de resolución de contenciones, cada uno pensado para proporcionar el servicio requerido. Con esta técnica, además de controlar la PLR, también se pueden considerar el retraso y la complejidad computacional como métricas QoS.

The use of 802.11-based multi-hop wireless mesh networks for Internet access is extensive and growing. The primary advantages of this approach are ease of deployment and lower cost. However, these networks are designed for web and e-mail applications. Highly interactive applications, such as multiplayer online games and VoIP, with their requirements for low delay, present significant challenges to these networks. In particular, the interaction between real-time traffic and TCP traffic tends to result in either a failure of the real-time traffic getting its needed QoS or the TCP traffic unnecessarily experiencing very poor throughput. To solve this problem we place real-time and TCP traffic into separate queues. We then rate-limit TCP traffic based on the average queue size of the local or remote real-time queues. Thus, TCP traffic is permitted to use excess bandwidth as long as it does not interfere with real-time traffic guarantees. We therefore call our scheme Real-time Queue-based Rate and Admission Control, RtQ-RAC. Extensive simulations using the network simulator, ns-2, demonstrate that our approach is effective in providing soft real-time support, while allowing efficient use of the remaining bandwidth for TCP traffic.

The use of 802.11-based multi-hop wireless mesh networks for Internet access is extensive and growing. The primary advantages of this approach are ease of deployment and lower cost. However, these networks are designed for web and e-mail applications. Highly interactive applications, such as multiplayer online games and VoIP, with their requirements for low delay, present significant challenges to these networks. In particular, the interaction between real-time traffic and TCP traffic tends to result in either a failure of the real-time traffic getting its needed QoS or the TCP traffic unnecessarily experiencing very poor throughput. To solve this problem we place real-time and TCP traffic into separate queues. We then rate-limit TCP traffic based on the average queue size of the local or remote real-time queues. Thus, TCP traffic is permitted to use excess bandwidth as long as it does not interfere with real-time traffic guarantees. We therefore call our scheme Real-time Queue-based Rate and Admission Control, RtQ-RAC. Extensive simulations using the network simulator, ns-2, demonstrate that our approach is effective in providing soft real-time support, while allowing efficient use of the remaining bandwidth for TCP traffic.

In the last decade, the world has witnessed rapid increasing applications of wireless networks. However, with the fixed spectrum allocation policy that has been used since the beginning of the spectrum regulation to assign different spectrum bands to different wireless applications, it has been observed that most of the allocated spectrum bands are underutilized. Therefore, if these bands can be opportunistically used by new emerging wireless networks, the spectrum scarcity can be resolved. Cognitive Radio (CR) is a revolutionary and promising technology that can identify and then exploit the spectrum opportunities. In Cognitive Radio Networks (CRNs), the spectrum can be utilized by two kinds of users: Primary Users (PUs) having exclusive licenses to use certain spectrum bands for specific wireless applications, and Secondary Users (SUs) having no spectrum licenses but seeking for any spectrum opportunities. The SUs can make use of the licensed unused spectrum if they do not make any harmful interference to the PUs. However, the variation of the spectrum availability over the time and locations, due to the coexistence with the PUs, and the spread of the spectrum opportunities over wide spectrum bands create a unique trait of the CRNs. This key trait poses great challenges in different aspects of the radio resource management in CRNs such as the spectrum sensing, spectrum access, admission control, channel allocation, Quality-of-Service (QoS) provisioning, etc. In this thesis, we study the resource management of both single-hop and multi-hop CRNs. Since most of the new challenges in CRNs can be tackled by designing an efficient Medium Access Control (MAC) framework, where the solutions of these challenges can be integrated for efficient resource management, we firstly propose a novel MAC framework that integrates a kind of cooperative spectrum sensing method at the physical layer into a cooperative MAC protocol considering the requirements of both the SUs and PUs. For spectrum identification, a computationally simple but efficient sensing algorithm is developed, based on an innovative deterministic sensing policy, to assist each sensing user for identifying the optimum number of channels to sense and the optimum sensing duration. We then develop an admission control scheme and channel allocation policy that can be integrated in the proposed MAC framework to regulate the number of sensing users and number of access users; therefore, the spectrum identification and exploitation can be efficiently balanced. Moreover, we propose a QoS-based spectrum allocation framework that jointly considers the QoS provisioning for heterogeneous secondary Real-Time (RT) and Non-Real Time (NRT) users with the spectrum sensing, spectrum access decision, and call admission control. We analyze the proposed QoS-based spectrum allocation framework and find the optimum numbers of the RT and NRT users that the network can support. Finally, we introduce an innovative user clustering scheme to efficiently manage the spectrum identification and exploitation in multi-hop ad hoc CRNs. We group the SUs into clusters based on their geographical locations and occurring times and use spread spectrum techniques to facilitate using one frequency for the Common Control Channels (CCCs) of the whole secondary network and to reduce the co-channel interference between adjacent clusters by assigning different spreading codes for different clusters. The research results presented in this thesis contribute to realize the concept of the CRNs by developing a practical MAC framework, spectrum sensing, spectrum allocation, user admission control, and QoS provisioning for efficient resource management in these promising networks.

The MultiProtocol Label Switching (MPLS) framework is used in many networks to provide efficient load balancing which distributes the traffic for efficient Quality of Service (QoS) provisioning in the network. If the MPLS framework is combined with Differentiated Services (DiffServ) architecture, together they can provide aggregate-based service differentiation and QoS. The combined use of DiffServ and MPLS in a network is called DiffServ-aware Traffic Engineering (DS-TE). Such DiffServ-based MPLS networks demand development of efficient methods for QoS provisioning. In this thesis, an automated manager for management of these DiffServ-based MPLS networks is proposed. This manager, called Traffic Engineering Automated Manager (TEAM), is a centralized authority for adaptively managing a DiffServ/MPLS domain and it is responsible for dynamic bandwidth and route management. TEAM is designed to provide a novel and unique architecture capable of managing large scale MPLS/DiffServ domains without any human interference. TEAM constantly monitors the network state and reconfigures the network for efficient handling of network events. Under the umbrella of TEAM, new schemes for Label Switched Path (LSP) setup/teardown, traffic routing, and network measurement are proposed and evaluated through simulations. Also, extensions to include Generalized MPLS (GMPLS) networks and inter-domain management are proposed.

Today, telecommunication is improving rapidly. People are online anywhere anytime. Due to increasing demand in communication, wireless technologies are progressing quickly trying to provide more services in a wide range. In order to address mobility and connectivity requirements of users in wide areas, Worldwide Interoperability for Microwave Access (Wimax) has been introduced as a forth generation telecommunication technology. Wimax, which is also called Metropolitan Area Network (MAN), is based on IEEE 802.16 standard where a Base Station (BS) provides last mile broadband wireless access to the end users known as Mobile Stations (MS). However, in places where high constructions exist, the signal rate between MS and BS decreases or even the signal can be lost completely due to shadow fading. As a response to this issue, recently an intermediate node specification, namely Relay Station, has been defined in IEEE 802.16j standard for relaying, which provides both throughput enhancement and coverage extension. However, this update has introduced a new problem / call admission control in non-transparent relay networks that support coverage extension. In this thesis, a Quality of Service (QoS) aware dynamic admission control algorithm for IEEE 802.16j non-transparent relay networks is introduced. Our objectives are admitting more service flows, utilizing the bandwidth, giving individual control to each relay station (RS) on call acceptance and rejection, and finally not affecting ongoing service flow quality in an RS due to the dense population of service flows in other RSs. The simulation results show that the proposed algorithm outperforms the other existing call admission control algorithms. Moreover, this algorithm can be interpreted as pioneer call admission control algorithm in IEEE 802.16j non-transparent networks.

Wireless broadband technologies provide ubiquitous broadband access to wireless users, enabling services that were available only to wireline users. At the same time, with the steady growth of real-time streaming applications such as video on demand (VoD), voice over IP (VoIP), massive online gaming, and so forth, the IEEE 802.16 standard (commonly known as WiMAX) has emerged as one of the strongest contenders to provide such kind of broadband wireless access services. WiMAX has specified some advanced lard features at physical (PHY) layer techniques and media access control (MAC) layer protocols, which adopted many Quality of Service (QoS) scheduling algorithms, resource allocation, Hybrid ARQ and so on. Moreover, forward error correction (FEC) and automatic repeat request (ARQ) techniques have already specified in the standard and they are used to support real-time streaming services in all kind of channel conditions. Inside this thesis, we propose an efficient fast feedback algorithm and buffer congestion control improvement scheme for data streaming over WiMAX networks. Two reserved bits in Generic MAC header (GMH) of each MPDU in WiMAX systems are utilized. The first reserved bit is used as a trigger in fast feedback strategy to add more robust coding and choose better feedback channel when the feedback message does not arrive properly within its cycle time trip (CTT) or after timeout. The second reserved bit is used to inform the base station about the serving subscriber stations¡¦ buffers states when their buffers are above the predefine-threshold value. Increasing number of retrieval of the feedback message, which means we can increase restore-bit-error probabilities within each frame, and then the throughput performance shall increase as well. In addition, by having each subscriber stations or service station¡¦s buffer states knowledge, the BS will be able to provide extra bandwidth allocation to the SSs more efficiency and accuracy. Keywords: WiMAX, PHY, MAC, Hybrid ARQ, QoS, Generic MAC header, CTT, Fast Feedback, Buffer Congestion Control Improvement, Real-Time streaming.

The Metro Ethernet network (MEN) expands the advantages of Ethernet to cover areas wider than LAN. MENs running Ethernet Services as specified by the Metro Ethernet Forum (MEF) are known as Carrier Ethernet Networks (CENs). CENs can cover not only metro areas, but it can expand to cover global areas by connecting multiple MENs. Next-generation CENs are expected to support 100 GbE. With arising technologies for Ultra Long-haul (ULH) networks the bandwidth bottleneck of CENs is shifting to other areas like the transport layer protocol (such as the Transport Control Protocol or TCP) and the chip-to-chip channel capacity found at the network edge, which in general has an electrical backplane. Traditional TCP is well known to have difficulties reaching the full available bandwidth, due to its inefficient AIMD mechanisms under a high-delay-bandwidth-product environment. At the network edge, network equipment with electrical backplanes poses many problems including inductive-capacitive effects that limit its bandwidth. These are the two main issues addressed in this work. To resolve the transport layer issue, this work proposes a transport protocol that fully utilizes the available bandwidth while preserving TCP-friendliness and providing QoS support that is compatible with Ethernet Services. It can guarantee throughputs above the Committed Information Rate (CIR), which is specified in the Service Level Agreement (SLA). To resolve the physical layer limitations, a novel optical coupling technique is examined to encourage the use of optical backplanes for network-edge and core technology. The proposed technique consists of aligning the normal of the laser emission plane, waveguide plane and the normal of the photodetector active region plane with the purpose of reducing optical power loss caused by common methods of light manipulation. By addressing the shortcomings of both Traditional TCP and electrical backplane technology the overall throughput can be significantly increased.

La qualité de service (QoS), liée au partage des ressources, prend tout son sens dans le cadre des réseaux multimédias. L'intégration de celle-ci dans les protocoles de routage, nécessite la prise en compte des phénomènes de congestion. Cela a favorisé l'apparition du protocole MPLS (Multi Protocol Label Switching). Cette nouvelle technologie, grâce à son routage par LSP (Label Swithed Path), permet une gestion plus fine des ressources disponibles dans le réseau. Nous traitons en première partie de ce travail le problème du routage des LSPs dans les réseaux IP/MPLS. Nous en formulons une modélisation originale qui tient compte de la QoS. Nous proposons aussi une heuristique de résolution (ILSP-OLS-ACO) qui gère un grand nombre de contraintes opérationnelles, tels que la bande passante, les contraintes d'affinités ou de sécurité sur les LSPs. Celle-ci fournit des solutions quasi-optimales tout en permettant le passage à l'échelle (grands réseaux, milliers de LSPs). La deuxième partie de notre travail concerne la conception optimale de topologie d'accès. L'originalité de l'approche réside dans le fait de prendre en compte le trafic générés par les clients ainsi que les coûts des équipements. Nous élaborons une modélisation basée sur la programmation linéaire en nombres entiers. Nous proposons pour la résoudre une méthode exacte basée sur des techniques de « Branch and Cut ». Nous proposons aussi une heuristique combinant une technique de « clustering » et une technique de recherche locale, qui permet d'obtenir très rapidement des solutions quasi-optimales.