Current telecommunication infrastructure is undergoing significant changes. Such changes involve the type of traffic traveling through the network as well as the requirements imposed by the new traffic mix (e.g. strict delay control and low end-to-end delay). In this new networking scenario, the current infrastructure, which remained almost unchanged for the last several decades, is struggling to adapt, and its limitations in terms of power consumption, scalability, and economical viability have become more evident. In this dissertation we explore the potential advantages of using periodic data structures to handle efficiently bandwidth-intensive transactions, which constitute a significant portion of today's network traffic. We start by implementing an approach that can work as a standalone system aiming to provide the same advantages promised by all-optical approaches such as OBS and OFS. We show that our approach is able to provide similar advantages (e.g. energy efficiency, link utilization, and low computational load for the network hardware) while avoiding the drawbacks (e.g. use of optical buffers, inefficient resource utilization, and costly deployment), using commercially available hardware. Aware of the issues of large scale hardware redeployment, we adapt our approach to work within the current transport network architecture, reusing most of the hardware and protocols that are already in place, offering a more gradual evolutionary path, while retaining the advantages of our standalone system. We then apply our approach to Data Center Networks (DCNs), showing its ability to achieve significant improvements in terms of network performance stability, predictability, performance isolation, agility, and goodput with respect to popular DCN approaches. We also show our approach is able to work in concert with many proposed and deployed DCN architectures, providing DCNs with a simple, efficient, and versatile protocol to handle bandwidth-intensive applications within the DCs. / Graduate

La ràpida evolució d’Internet i l’àmplia gamma de noves aplicacions (per exemple, multimèdia, videoconferència, jocs en línia, etc.) ha fomentat canvis revolucionaris en la manera com ens comuniquem. A més, algunes d’aquestes aplicacions demanden grans quantitats de recursos d’ample de banda amb diversos requeriments de qualitat de servei (QoS). El desenvolupament de la multiplexació per divisió de longitud d’ona (WDM) en els anys noranta va fer molt rendible la disponibilitat d’ample de banda. Avui dia, les tecnologies de commutació òptica de circuits són predominants en el nucli de la xarxa, les quals permeten la configuració de canals (lightpaths) a través de la xarxa. No obstant això, la granularitat d’aquests canals ocupa tota la longitud d’ona, el que fa que siguin ineficients per a proveir canals de menor ample de banda (sub-longitud d’ona). Segons la comunitat científica, és necessari augmentar la transparència dels protocols, així com millorar l’aprovisionament d’ample de banda de forma dinàmica. Per tal de fer això realitat, és necessari desenvolupar noves arquitectures. La commutació òptica de ràfegues i de paquets (OBS/OPS), són dues de les tecnologies proposades. Aquesta tesi contribueix amb tres arquitectures de xarxa destinades a millorar el transport de dades sub-longitud d’ona. En primer lloc, aprofundim en la naturalesa sense connexió en OBS. En aquest cas, la xarxa incrementa el seu dinamisme a causa de les transmissions a ràfega. A més, les col·lisions entre ràfegues degraden el rendiment de la xarxa fins i tot a càrregues molt baixes. Per fer front a aquestes col·lisions, es proposa un esquema de resolució de col·lisions pro actiu basat en un algorisme d’encaminament i assignació de longitud d’ona (RWA) que balanceja de forma automàtica i distribuïda la càrrega en la xarxa. En aquest protocol, el RWA i la transmissió de ràfegues es basen en l’explotació i exploració de regles de commutació que incorporen informació sobre contencions i encaminament. Per donar suport a aquesta arquitectura, s’utilitzen dos tipus de paquets de control per a l’encaminament de les ràfegues i l’actualització de les regles de commutació, respectivament. Per analitzar els beneficis del nou algorisme, s’utilitzen quatre topologies de xarxa diferents. Els resultats indiquen que el mètode proposat millora en diferents marges la resta d’algorismes RWA en funció de la topologia i sense penalitzar altres paràmetres com el retard extrem a extrem. La segona contribució proposa una arquitectura híbrida sense i orientada a connexió sobre la base d’un protocol de control d’accés al medi (MAC) per a xarxes OBS (DAOBS). El MAC ofereix dos mètodes d’accés: arbitratge de cua (QA) per a la transmissió de ràfegues sense connexió, i pre-arbitratge (PA) per serveis TDM orientats a connexió. Aquesta arquitectura permet una àmplia gamma d’aplicacions sensibles al retard i al bloqueig. Els resultats avaluats a través de simulacions mostren que en l’accés QA, les ràfegues de més alta prioritat tenen garantides zero pèrdues i latències d’accés molt baixes. Pel que fa a l’accés PA, es reporta que la duplicació de la càrrega TDM augmenta en més d’un ordre la probabilitat de bloqueig, però sense afectar en la mateixa mesura les ràfegues sense connexió. En aquest capítol també es tracten dos dels problemes relacionats amb l’arquitectura DAOBS i el seu funcionament. En primer lloc, es proposa un model matemàtic per aproximar el retard d’accés inferior i superior com a conseqüència de l’accés QA. En segon lloc, es formula matemàticament la generació i optimització de les topologies virtuals que suporten el protocol per a l’escenari amb tràfic estàtic. Finalment, l’última contribució explora els beneficis d’una arquitectura de xarxa òptica per temps compartit (TSON) basada en elements de càlcul de camins (PCE) centralitzats per tal d’evitar col·lisions en la xarxa. Aquesta arquitectura permet garantir l’aprovisionament orientat a connexió de canals sub-longitud d’ona. En aquest capítol proposem i simulem tres arquitectures GMPLS/PCE/TSON. A causa del enfocament centralitzat, el rendiment de la xarxa depèn en gran mesura de l’assignació i aprovisionament de les connexions. Amb aquesta finalitat, es proposen diferents algorismes d’assignació de ranures temporals i es comparen amb les corresponents formulacions de programació lineal (ILP) per al cas estàtic. Per al cas de tràfic dinàmic, proposem i avaluem mitjançant simulació diferents heurístiques. Els resultats mostren els beneficis de proporcionar flexibilitat en els dominis temporal i freqüencial a l’hora d’assignar les ranures temporals. / The rapid evolving Internet and the broad range of new data applications (e.g., multimedia, video-conference, online gaming, etc.) is fostering revolutionary changes in the way we communicate. In addition, some of these applications demand for unprecedented amounts of bandwidth resources with diverse quality of service (QoS). The development of wavelength division multiplexing (WDM) in the 90's made very cost-effective the availability of bandwidth. Nowadays, optical circuit switching technologies are predominant in the core enabling the set up of lightpaths across the network. However, full-wavelength lightpath granularity is too coarse, which results to be inefficient for provisioning sub-wavelength channels. As remarked by the research community, an open issue in optical networking is increasing the protocol transparency as well as provisioning true dynamic bandwidth allocation at the network level. To this end, new architectures are required. Optical burst/packet switching (OBS/OPS) are two such proposed technologies under investigation. This thesis contributes with three network architectures which aim at improving the sub-wavelength data transport from different perspectives. First, we gain insight into the connectionless nature of OBS. Here, the network dynamics are increased due to the short-lived burst transmissions. Moreover, burst contentions degrade the performance even at very low loads. To cope with them, we propose a proactive resolution scheme by means of a distributed auto load-balancing routing and wavelength assignment (RWA) algorithm for wavelength-continuity constraint networks. In this protocol, the RWA and burst forwarding is based on the exploitation and exploration of switching rule concentration values that incorporate contention and forwarding desirability information. To support such architecture, forward and backward control packets are used in the burst forwarding and updating rules, respectively. In order to analyze the benefits of the new algorithm, four different network topologies are used. Results indicate that the proposed method outperforms the rest of tested RWA algorithms at various margins depending on the topology without penalizing other parameters such as end-to-end delay. The second contribution proposes a hybrid connectionless and connection-oriented architecture based on a medium access control (MAC) protocol for OBS networks (DAOBS). The MAC provides two main access mechanisms: queue arbitrated (QA) for connectionless bursts and pre-arbitrated (PA) for TDM connection-oriented services. Such an architecture allows for a broad range of delay-sensitive applications or guaranteed services. Results evaluated through simulations show that in the QA access mode highest priority bursts are guaranteed zero losses and very low access latencies. Regarding the PA mode, we report that doubling the offered TDM traffic load increases in more than one order their connection blocking, slightly affecting the blocking of other connectionless bursts. In this chapter, we also tackle two of the issues related with the DAOBS architecture and its operation. Firstly, we model mathematically the lower and upper approximations of the access delay as a consequence of the connectionless queue arbitrated access. Secondly, we formulate the generation of the virtual light-tree overlay topology for the static traffic case.

Τα σύγχρονα τηλεπικοινωνιακά δίκτυα οπτικών ινών χρησιμοποιούν την τεχνολογία WDM Wavelength Division Multiplexing) η οποία έχει κάνει εφικτή την αξιοποίηση – ως ένα βαθμό– του τεράστιου εύρους ζώνης της οπτικής ίνας. Στα πλαίσια της παρούσας διπλωματικής εργασίας θα παρουσιαστούν οι τρεις βασικές οπτικές αρχιτεκτονικές μεταγωγής (οπτική μεταγωγή κυκλώματος –OCS–, οπτική μεταγωγή πακέτου –OPS–, οπτική μεταγωγή ριπής–OBS–) οι οποίες μετατρέπουν τη διαθέσιμη χωρητικότητα σε ωφέλιμο throughput. Ιδιαίτερη έμφαση θα δοθεί στην αρχιτεκτονική OBS, η οποία έχει τραβήξει το ερευνητικό ενδιαφέρον τα τελευταία χρόνια, σαν μια ενδιαφέρουσα εναλλακτική της (ώριμης πλέον) αρχιτεκτονικής OCS. Συγκεκριμένα, θα διερευνηθεί η μετάδοση του TCP πρωτοκόλλου πάνω από OBS δίκτυα μέσα από λεπτομερείς προσομοιώσεις, και θα προταθούν κατάλληλες βελτιώσεις της αρχιτεκτονικής OBS. Ακόμα, θα περιγραφεί μια πρωτότυπη υβριδική αρχιτεκτονική οπτικής μεταγωγής ριπής. / Internet traffic has faced an exploding growth in recent years. The ever-growing demand for multimedia web services, as well as the advent of P2P technology, are driving core networks to their limits. This calls for the design of high capacity core networks, being able to serve the user’s high bandwidth requests. Optical networks have become a key part of the solution, mainly due to the vast capacity of optical fibers. Specifically, the advent of WDM technology has resulted in transmission capacities that have increased manifold in recent years. It is the router/switch throughput, however, that really transforms the raw bit rates into effective bandwidth. In this diploma thesis, we study the three basic optical architectures, that is Optical Circuit Switching (OCS), Optical Packet Switching (OPS) and Optical Burst Switching (OPS). Emphasis is given on OBS architecture, which has drawn research interest in recent year, as a possible replacement for the well-established OCS architecture. Specifically, we will study the transmission of TCP traffic over OBS networks through simulation, and propose modifications for the OBS architecture. Finally, a novel hybrid switch architecture will be proposed, combining the merits of OBS and OCS.

Wavelength division multiplexed (WDM) optical networks are rapidly becoming the technology of choice in network infrastructure and next-generation Internet architectures. WDM networks have the potential to provide unprecedented bandwidth, reduce processing cost, achieve protocol transparency, and enable efficient failure handling. This dissertation addresses the important issues of improving the performance and enhancing the reliability of WDM networks as well as modeling and evaluating the performance of these networks. Optical wavelength conversion is one of the emerging WDM enabling technologies that can significantly improve bandwidth utilization in optical networks. A new approach for the sparse placement of full wavelength converters based on the concept of the k-Dominating Set (k-DS) of a graph is presented. The k-DS approach is also extended to the case of limited conversion capability using three scalable and cost-effective switch designs: flexible node-sharing, strict node-sharing and static mapping. Compared to full search algorithms previously proposed in the literature, the K-DS approach has better blocking performance, has better time complexity and avoids the local minimum problem. The performance benefit of the K-DS approach is demonstrated by extensive simulation. Fiber delay line (FDL) is another emerging WDM technology that can be used to obtain limited optical buffering capability. A placement algorithm, k-WDS, for the sparse placement of FDLs at a set of selected nodes in Optical Burst Switching (OBS) networks is proposed. The algorithm can handle both uniform and non-uniform traffic patterns. Extensive performance tests have shown that k-WDS provides more efficient placement of optical fiber delay lines than the well-known approach of placing the resources at nodes with the highest experienced burst loss. Performance results that compare the benefit of using FDLs versus using optical wavelength converters (OWCs) are presented. A new algorithm, A-WDS, for the placement of an arbitrary numbers of FDLs and OWCs is introduced and is evaluated under different non-uniform traffic loads. This dissertation also introduces a new cost-effective optical switch design using FDL and a QoS-enhanced JET (just enough time) protocol suitable for optical burst switched WDM networks. The enhanced JET protocol allows classes of traffic to benefit from FDLs and OWCs while minimizing the end-to-end delay for high priority bursts. Performance evaluation models of WDM networks represent an important research area that has received increased attention. A new analytical model that captures link dependencies in all-optical WDM networks under uniform traffic is presented. The model enables the estimation of connection blocking probabilities more accurately than previously possible. The basic formula of the dependency between two links in this model reflects their degree of adjacency, the degree of connectivity of the nodes composing them and their carried traffic. The usefulness of the model is illustrated by applying it to the sparse wavelength converters placement problem in WDM networks. A lightpath containing converters is divided into smaller sub-paths such that each sub-path is a wavelength continuous path and the nodes shared between these sub-paths are full wavelength conversion capable. The blocking probability of the entire path is obtained by computing the blocking probabilities of the individual sub-paths. The analytical-based sparse placement algorithm is validated by comparing it with its simulation-based counterpart using a number of network topologies. Rapid recovery from failure and high levels of reliability are extremely important in WDM networks. A new Fault Tolerant Path Protection scheme, FTPP, for WDM mesh networks based on the alarming state of network nodes and links is introduced. The results of extensive simulation tests show that FTPP outperforms known path protection schemes in terms of loss of service ratio and network throughput. The simulation tests used a wide range of values for the load intensity, the failure arrival rate and the failure holding time. The FTPP scheme is next extended to the differentiated services model and its connection blocking performance is evaluated. Finally, a QoS-enhanced FTPP (QEFTPP) routing and path protection scheme in WDM networks is presented. QEFTPP uses preemption to minimize the connection blocking percentage for high priority traffic. Extensive simulation results have shown that QEFTPP achieves a clear QoS differentiation among the traffic classes and provides a good overall network performance.

No description available.

[pt] A difusão seletiva e os recentes avanços na tecnologia de transmissão óptica, mais especificamente na multiplexação por comprimento de onda (Wavelength Division Multiplexing- WDM), aliados à consolidação do IP como protocolo dominante das redes convergentes, vêm oferecendo novas perspectivas para as futuras gerações de inter-redes. Este trabalho faz uso da evolução dessas tecnologias para propor um conjunto de adaptações à difusão seletiva, em especial ao IP Multicast, denominado MIRROR (Multicast IP para Redes baseadas em Rajadas Ópticas Rotuladas). A proposta MIRROR sugere modificações e adequações para tornar o IP Multicast menos complexo, mais escalável em relação ao número de grupos ativos simultaneamente e mais adequado às redes baseadas em comutação óptica. Basicamente, MIRROR revê a necessidade de todos os roteadores ao longo da árvore de distribuição multiponto manterem informações de estado relacionadas a esta, bem como sugere adequações na forma como os caminhos multiponto são estabelecidos quando se emprega comutação baseada em rótulos na difusão seletiva. Para avaliar a proposta MIRROR, investiu-se em duas frentes distintas, uma baseada na análise comparativa entre a MIRROR e algumas alternativas ao IP Multicast apresentadas na literatura, e outra baseada no desenvolvimento de um protótipo da proposta no simulador NS (Network Simulator), com o intuito de referendar os resultados da análise comparativa. Na análise comparativa, confronta-se parâmetros como: requisitos de informações de estado, custo com informações de controle, custo de encaminhamento dos pacotes e custo da árvore de multiponto. O desenvolvimento do protótipo envolveu a criação de uma nova estrutura de nó e a alteração de módulos já existentes no NS, para tornar possível a simulação de redes comutadas por rajadas ópticas rotuladas no contexto da difusão seletiva. / [en] Multicast communication and recent advances in optical technology, most specifically in Wavelength Division Multiplexing (WDM), allied with the consolidation of IP as the dominant protocol of convergent networks, offer new perspectives for the next generation Internet. This thesis utilises these technologies to propose a set of adaptations, called MIRROR, to multicast communication, specifically IP Multicast, in labelled burst-switched optical networks. MIRROR proposes modifications to traditional IP Multicast in order to improve its scalability as a function of the number of simultaneously active groups, as well as making it more appropriate for use in optically switched networks. Basically, MIRROR includes new proposals for handling state information about the multicast distribution tree, as well as for the establishment of label-based multicast paths. In order to evaluate this proposal, two approaches are followed, one based on a comparative analysis between MIRROR and a number of other alternatives to IP Multicast proposed in the literature, and the other based on the implementation of a prototype in the simulation environment provided by NS (Network Simulator). The comparative analysis evaluates such parameters as: state requirement information, control overhead, packet processing efficiency and tree cost. The prototype implementation implements a new node structure and alters existing NS modules (OBS e MPLS), to make possible the simulation of labelled burst-switched optical networks in the multicast context.

Innovations in optical communication are expected to transform the landscape of global communications, internet and datacenter networks. This dissertation investigates several important issues in optical communication such as fairness, throughput, blocking probability and differentiated quality of service (QoS). Novel algorithms and new approaches have been presented to improve the performance of optical circuit switching (OCS) and optical burst switching (OBS) for long haul, and datacenter networks. Extensive simulations tests have been conducted to evaluate the effectiveness of the proposed algorithms. These simulation tests were performed over a number of network topologies such as ring, mesh and U.S. Long-Haul, some high processing computing (HPC) topologies such as 2D and 6D mesh torus topologies and modern datacenter topologies such as FatTree and BCube. Two new schemes are proposed for long haul networks to improve throughput and hop count fairness in OBS networks. The idea is motivated by the observation that providing a slightly more priority to longer bursts over short bursts can significantly improve the throughput of the OBS networks without adversely affecting hop-count fairness. The results of extensive performance tests have shown that proposed schemes improve the throughput of optical OBS networks and enhance the hop-count fairness. Another contribution of this dissertation is the research work on developing routing and wavelength assignment schemes in multimode fiber networks. Two additional schemes for long haul networks are presented and evaluated over multimode fiber networks. First for alleviating the fairness problem in OBS networks using wavelength-division multiplexing as well as mode-division multiplexing while the second scheme for achieving higher throughput without sacrificing hop count fairness. We have also shown the significant benefits of using both mode division multiplexing and wavelength division multiplexing in real-life short-distance optical networks such as the optical circuit switching networks used in the hybrid electronic-optical switching architectures for datacenters. We evaluated four mode and wavelength assignment heuristics and compared their throughput performance. We also included preliminary results of impact of the cascaded mode conversion constraint on network throughput. Datacenter and high performance computing networks share a number of common performance goals. Another highly efficient adaptive mode wavelength- routing algorithm is presented over OBS networks to improve throughput of these networks. The effectiveness of the proposed model has been validated by extensive simulation results. In order to optimize bandwidth and maximize throughput of datacenters, an extension of TCP called multipath-TCP (MPTCP) has been evaluated over an OBS network using dense interconnect datacenter topologies. We have proposed a service differentiation scheme using MPTCP over OBS for datacenter traffic. The scheme is evaluated over mixed workload traffic model of datacenters and is shown to provide tangible service differentiation between flows of different priority levels. An adaptive QoS differentiation architecture is proposed for software defined optical datacenter networks using MPTCP over OBS. This scheme prioritizes flows based on current network state.

[ES] La distribución de contenidos multimedia, tanto a través de Internet como de redes gestionadas dedicadas exclusivamente a este fin, está teniendo una mayor cuota del tráfico total y crece cada día, especialmente el servicio de vídeo streaming. Por ello, ha surgido el interés de investigar para mantener y mejorar la Calidad de Servicio (QoS) en redes de acceso y backbone, y ofrecer Calidad de Experiencia (QoE) a los usuarios finales; mediante el ajuste de parámetros, configuraciones, protocolos y algoritmos a los requerimientos de la red en cuanto a la alta demanda de tráfico de vídeo. Lo anterior aprovechando al máximo los recursos de la red, tanto en el núcleo y las redes de acceso, como el ancho de banda y la capacidad total instalada, y al mismo tiempo, proponiendo alternativas de transporte eficientes para el tráfico de vídeo agregado sin incrementar los costos de los servicios; y alternativas para reducir los costos operativos y gastos de mantenimiento de manera eficiente, como por ejemplo la reducción del consumo de energía. En primer lugar, esta tesis aborda el estudio de propuestas de conmutación óptica en redes troncales y la distribución de contenidos multimedia, particularmente la propuesta de Red de Conmutación Óptica de Ráfagas (OBS) desde el punto de vista de la Ingeniería Telemática. Por lo tanto, se estudia la arquitectura, protocolos y algoritmos de estas redes para hacer aportes a la evaluación del tráfico de vídeo y alternativas para ofrecer QoS y QoE para el tráfico multimedia. En este sentido, este trabajo contribuye con propuestas para la evaluación de la calidad mediante simulación utilizando trazas de transmisiones de vídeo reales, y propuestas de algoritmos de scheduling y ensamblaje de ráfagas ópticas para ofrecer QoS y QoE al tráfico de vídeo en las redes OBS. En segundo lugar, se estudian los patrones de tráfico de vídeo a nivel de aplicación y de red para caracterizarlos y realizar estudios de eficiencia energética y ahorro de consumo energético en las tarjetas de red utilizadas por servidores de vídeo en streaming y clientes que implementan el estándar IEEE 802.3az Energy Efficient Ethernet (EEE), tanto en redes gestionadas, como IPTV que implementa transmisión de vídeo RTP, como en redes no gestionadas, como Internet, que implementa el estándar Dynamic Adaptive Streaming over HTTP (DASH) para servicios Over The Top (OTT). Para alcanzar este objetivo, se implementan testbeds y se llevan a cabo experimentos para concluir cómo el tráfico de vídeo podría beneficiarse de la eficiencia energética y lograr una reducción a gran escala en el consumo de energía si el estándar EEE se implementa a nivel mundial en las tarjetas Ethernet de los dispositivos involucrados en la distribución de vídeo. El tema integrador que se aborda en esta tesis es la distribución de vídeo digital. Por lo tanto, este trabajo también presenta un estudio sobre la codificación y adaptación de transmisión de vídeo en redes IP, presentando aspectos relacionados con la evaluación de la calidad de vídeo y la evaluación de la calidad de transmisión de la red. Por ello, se proponen e implementan metodologías y bancos de pruebas que involucran la programación y ejecución de simuladores, scripts, algoritmos, software y, en general, frameworks para realizar estudios de transmisiones de vídeo en redes de telecomunicaciones, cubriendo el amplio espectro de las Comunicaciones Multimedia. Este estudio podría ser de utilidad para emprender otros estudios necesarios ante la rápida evolución tanto de la demanda de servicios multimedia como de las propuestas de nuevas técnicas de distribución de contenidos. / [CAT] La distribució de continguts multimèdia, a través d'Internet i a través de xarxes gestionades dedicades exclusivament a aquesta finalitat, està tenint una part més gran del trànsit total i augmenta cada dia, especialment el servei de streaming de vídeo. Per tant, ha sorgit l'interés de la recerca per mantenir i millorar la Qualitat de Servei (QoS) en les xarxes d'accés i troncals, i per oferir Qualitat d'Experiència (QoE) als usuaris finals, ajustant paràmetres, configuracions, protocols i algoritmes als requeriments de la xarxa pel que fa a l'alta demanda de trànsit de vídeo. Aquest últim, aprofitant al màxim els recursos de xarxa al nucli i a les xarxes d'accés, com ara l'amplada de banda i la capacitat instal·lada total, i alhora, proposant alternatives de transport eficients per al trànsit agregat de vídeo sense augmentar els costos dels serveis i alternatives per reduir les despeses operatives i de manteniment d'una manera eficient, com ara reduir el consum d'energia. En primer lloc, aquesta tesi tracta l'estudi de les propostes de commutació òptica en xarxes troncals pel que fa a la distribució de continguts multimèdia, especialment l'estudi de la proposta de xarxes de commutació de ràfegues òptiques (OBS) des del punt de vista de l'Enginyeria Telemàtica. Per tant, s'estudia l'arquitectura, els protocols i els algoritmes d'aquestes xarxes per fer contribucions a l'avaluació del trànsit de vídeo i alternatives per oferir QoS i QoE per al trànsit multimèdia. En aquest sentit, aquest treball contribueix amb propostes per a l'avaluació de la qualitat mitjançant la simulació utilitzant traces de transmissions de vídeo reals, i propostes d'algoritmes de programació i muntatge de ràfegues òptiques per oferir QoS i QoE al trànsit de vídeo a les xarxes OBS. En segon lloc, s'estudien els patrons de trànsit de vídeo a nivell d'aplicació i xarxa per caracteritzar-los i realitzar estudis sobre eficiència energètica i estalvi energètic en les targetes de xarxa utilitzades pels servidors i clients de vídeo en streaming que implementen l'IEEE 802.3az Energy Efficient Ethernet (EEE) estàndard, tant en xarxes gestionades com IPTV, que implementa la transmissió de vídeo RTP, com en xarxes no gestionades com Internet, que implementa l'estàndard Dynamic Adaptive Streaming sobre HTTP (DASH) per als serveis Over The Top (OTT). Per assolir aquest objectiu, s'implementen bancs de proves i es realitzen experiments per concloure com el trànsit de vídeo podria beneficiar l'eficiència energètica i aconseguir una reducció a gran escala del consum d'energia si s'implementa l'estàndard EEE a les targetes Ethernet dels dispositius implicats en la distribución de vídeo. El tema integrador que s'aborda en aquesta tesi és la distribució de vídeo digital, per tant, aquest treball també presenta un estudi sobre la codificació i adaptació de la transmissió de vídeo a les xarxes IP, presentant aspectes relacionats amb l'avaluació de la qualitat del vídeo i l'avaluació de la qualitat de la transmissió de la xarxa. Per això, es proposen i s'implementen metodologies i bancs de proves que impliquen la programació i execució de simuladors, scripts, algoritmes, programari i, en general, frameworks per a la realització d'estudis de transmissions de vídeo en xarxes de telecomunicacions, abastant l'ampli espectre de les Comunicacions Multimèdia. Aquest estudi podria ser útil per emprendre altres estudis necessaris en la ràpida evolució tant de la demanda de serveis multimèdia com de les noves propostes de tècniques de distribució de continguts. / [EN] Multimedia content distribution, both over the Internet and over managed networks dedicated exclusively for this purpose, is having a greater share of total traffic and it is raising every day, especially the video streaming service. Therefore, research interest has arisen to maintain and improve the Quality of Service (QoS) in access and backbone networks, and to offer Quality of Experience (QoE) to end users, by adjusting parameters, configurations, protocols and algorithms to the network requirements regarding the video traffic high demand. The latter is carried out taking full advantage of the network resources at core and access networks such as bandwidth and total installed capacity, and at the same time, proposing efficient transportation alternatives for the aggregated video traffic without increasing services costs and alternatives to reduce operational and maintenance expenditures in an efficient manner, such as reducing the energy consumption. In the first place, this thesis deals with the study of optical switching proposals in backbone networks regarding multimedia content distribution, particularly studying the Optical Bursts Switching (OBS) Networks proposal from the Telematics Engineering point of view. Therefore, architecture, protocols and algorithms of these networks are studied to make contributions to the video traffic evaluation and alternatives to offer QoS and QoE for multimedia traffic. In this sense, this work contributes with proposals for the evaluation of quality through simulation using traces of real video transmissions, and optical burst assembly and scheduling algorithms proposals to offer QoS and QoE to video traffic in the OBS networks. Secondly, video traffic patterns at the application and network level are studied to characterize them and carry out studies on energy efficiency and energy consumption savings in the network cards used by streaming video servers and clients that implement the IEEE 802.3az Energy Efficient Ethernet (EEE) standard, both in managed networks, such as IPTV that implements RTP video streaming, and unmanaged networks, such as Internet which implements Dynamic Adaptive Streaming over HTTP (DASH) standard for Over The Top (OTT) services. To reach this goal, testbeds are implemented, and experiments are carried out to conclude how video traffic could benefit from the energy efficiency and achieve a large-scale reduction in energy consumption if the EEE standard is implemented word wide in the Ethernet cards of devices involved in video distribution. The integrating subject that is addressed in this thesis is the distribution of digital video. Therefore, this work also presents a study on the IP networks video transmission encoding and adaptation, presenting aspects regard the video quality evaluation and the network transmission quality assessment. Therefore, methodologies and testbeds are proposed and implemented that involves programming and execution of simulators, scripts, algorithms, software and, in general, frameworks to carry out studies for video transmissions in telecommunications networks, covering the broad spectrum of Multimedia Communications. This study could be useful to undertake other necessary studies in the rapid evolution of both, multimedia services demand and new content distribution techniques proposals. / Vargas Hernández, TR. (2023). A Contribution to Video Streaming Quality and Energy Efficiency in Optical and Ethernet Networks [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/192760

Les réseaux optiques à commutation de rafales (OBS) sont des candidats pour jouer un rôle important dans le cadre des réseaux optiques de nouvelle génération. Dans cette thèse, nous nous intéressons au routage adaptatif et au provisionnement de la qualité de service dans ce type de réseaux. Dans une première partie de la thèse, nous nous intéressons à la capacité du routage multi-chemins et du routage alternatif (par déflection) à améliorer les performances des réseaux OBS, pro-activement pour le premier et ré-activement pour le second. Dans ce contexte, nous proposons une approche basée sur l’apprentissage par renforcement où des agents placés dans tous les nœuds du réseau coopèrent pour apprendre, continuellement, les chemins du routage et les chemins alternatifs optimaux selon l’état actuel du réseau. Les résultats numériques montrent que cette approche améliore les performances des réseaux OBS comparativement aux solutions proposées dans la littérature. Dans la deuxième partie de cette thèse, nous nous intéressons au provisionnement absolu de la qualité de service où les performances pire-cas des classes de trafic de priorité élevée sont garanties quantitativement. Plus spécifiquement, notre objectif est de garantir la transmission sans pertes des rafales de priorité élevée à l’intérieur du réseau OBS tout en préservant le multiplexage statistique et l’utilisation efficace des ressources qui caractérisent les réseaux OBS. Aussi, nous considérons l’amélioration des performances du trafic best effort. Ainsi, nous proposons deux approches : une approche basée sur les nœuds et une approche basée sur les chemins. Dans l’approche basée sur les nœuds, un ensemble de longueurs d’onde est assigné à chaque nœud du bord du réseau OBS pour qu’il puisse envoyer son trafic garanti. Cette assignation prend en considération les distances physiques entre les nœuds du bord. En outre, nous proposons un algorithme de sélection des longueurs d’onde pour améliorer les performances des rafales best effort. Dans l’approche basée sur les chemins, le provisionnement absolu de la qualité de service est fourni au niveau des chemins entre les nœuds du bord du réseau OBS. À cette fin, nous proposons une approche de routage et d’assignation des longueurs d’onde qui a pour but la réduction du nombre requis de longueurs d’onde pour établir des chemins sans contentions. Néanmoins, si cet objectif ne peut pas être atteint à cause du nombre limité de longueurs d’onde, nous proposons de synchroniser les chemins en conflit sans le besoin pour des équipements additionnels. Là aussi, nous proposons un algorithme de sélection des longueurs d’onde pour les rafales best effort. Les résultats numériques montrent que l’approche basée sur les nœuds et l’approche basée sur les chemins fournissent le provisionnement absolu de la qualité de service pour le trafic garanti et améliorent les performances du trafic best effort. En outre, quand le nombre de longueurs d’ondes est suffisant, l’approche basée sur les chemins peut accommoder plus de trafic garanti et améliorer les performances du trafic best effort par rapport à l’approche basée sur les nœuds. / Optical Burst Switching (OBS) networks are candidates to play an important role in the context of next generation optical networks. In this thesis, we are interested in adaptive routing and quality of service provisioning for these networks. In the first part of the thesis, we study the capability of multi-path routing and alternative routing (deflection routing) to improve the performance of the OBS network proactively for the former and reactively for the latter. In this context, we propose a reinforcement learning-based approach where learning agents, placed in each OBS node, cooperate to learn, continuously, optimal routing paths and alternative paths according to the current state of the network. Numerical results show that the proposed approach improves the performance of the OBS network compared to existing solutions in the literature. In the second part of the thesis, we consider the problem of absolute quality of service provisioning for OBS networks where worst-case performance of high priority traffic is guaranteed quantitatively. Particularly, we are interested in the loss-free transmission, inside the OBS network, of high priority bursts, while preserving statistical multiplexing gain and high resources utilization of the OBS network. Also, we aim to improve the performance of best effort traffic. Hence, we propose two approaches: (a) the node-based approach; and (b) the path-based approach. In the node-based approach, we propose to assign a set of wavelengths to each OBS edge node that it can use to send its guaranteed traffic. This assignment takes into consideration physical distances between edge nodes. Furthermore, we propose a wavelength selection algorithm to improve the performance of best effort bursts. In the path-based approach, absolute quality of service provisioning is offered at end-to-end path level. To do this, we propose a routing and wavelength assignment approach which aims to reduce the number of wavelengths required to establish contention free paths. Nevertheless, if this objective cannot be reached because of the limited number of wavelengths in each fiber link, we propose an approach to synchronize overlapping paths without the need for additional equipments for synchronization. Here again, we propose a wavelength selection algorithm for best effort bursts. Numerical results show that both the node-based and the path-based approaches successfully provide absolute quality of service provisioning for guaranteed traffic and improve the performance of best effort traffic. Also, path-based approach could accommodate more guaranteed traffic and improve the performance of best effort traffic compared to node-based approach when the number of wavelengths is sufficient.

Les réseaux optiques à commutation de rafales (OBS) sont des candidats pour jouer un rôle important dans le cadre des réseaux optiques de nouvelle génération. Dans cette thèse, nous nous intéressons au routage adaptatif et au provisionnement de la qualité de service dans ce type de réseaux. Dans une première partie de la thèse, nous nous intéressons à la capacité du routage multi-chemins et du routage alternatif (par déflection) à améliorer les performances des réseaux OBS, pro-activement pour le premier et ré-activement pour le second. Dans ce contexte, nous proposons une approche basée sur l’apprentissage par renforcement où des agents placés dans tous les nœuds du réseau coopèrent pour apprendre, continuellement, les chemins du routage et les chemins alternatifs optimaux selon l’état actuel du réseau. Les résultats numériques montrent que cette approche améliore les performances des réseaux OBS comparativement aux solutions proposées dans la littérature. Dans la deuxième partie de cette thèse, nous nous intéressons au provisionnement absolu de la qualité de service où les performances pire-cas des classes de trafic de priorité élevée sont garanties quantitativement. Plus spécifiquement, notre objectif est de garantir la transmission sans pertes des rafales de priorité élevée à l’intérieur du réseau OBS tout en préservant le multiplexage statistique et l’utilisation efficace des ressources qui caractérisent les réseaux OBS. Aussi, nous considérons l’amélioration des performances du trafic best effort. Ainsi, nous proposons deux approches : une approche basée sur les nœuds et une approche basée sur les chemins. Dans l’approche basée sur les nœuds, un ensemble de longueurs d’onde est assigné à chaque nœud du bord du réseau OBS pour qu’il puisse envoyer son trafic garanti. Cette assignation prend en considération les distances physiques entre les nœuds du bord. En outre, nous proposons un algorithme de sélection des longueurs d’onde pour améliorer les performances des rafales best effort. Dans l’approche basée sur les chemins, le provisionnement absolu de la qualité de service est fourni au niveau des chemins entre les nœuds du bord du réseau OBS. À cette fin, nous proposons une approche de routage et d’assignation des longueurs d’onde qui a pour but la réduction du nombre requis de longueurs d’onde pour établir des chemins sans contentions. Néanmoins, si cet objectif ne peut pas être atteint à cause du nombre limité de longueurs d’onde, nous proposons de synchroniser les chemins en conflit sans le besoin pour des équipements additionnels. Là aussi, nous proposons un algorithme de sélection des longueurs d’onde pour les rafales best effort. Les résultats numériques montrent que l’approche basée sur les nœuds et l’approche basée sur les chemins fournissent le provisionnement absolu de la qualité de service pour le trafic garanti et améliorent les performances du trafic best effort. En outre, quand le nombre de longueurs d’ondes est suffisant, l’approche basée sur les chemins peut accommoder plus de trafic garanti et améliorer les performances du trafic best effort par rapport à l’approche basée sur les nœuds. / Optical Burst Switching (OBS) networks are candidates to play an important role in the context of next generation optical networks. In this thesis, we are interested in adaptive routing and quality of service provisioning for these networks. In the first part of the thesis, we study the capability of multi-path routing and alternative routing (deflection routing) to improve the performance of the OBS network proactively for the former and reactively for the latter. In this context, we propose a reinforcement learning-based approach where learning agents, placed in each OBS node, cooperate to learn, continuously, optimal routing paths and alternative paths according to the current state of the network. Numerical results show that the proposed approach improves the performance of the OBS network compared to existing solutions in the literature. In the second part of the thesis, we consider the problem of absolute quality of service provisioning for OBS networks where worst-case performance of high priority traffic is guaranteed quantitatively. Particularly, we are interested in the loss-free transmission, inside the OBS network, of high priority bursts, while preserving statistical multiplexing gain and high resources utilization of the OBS network. Also, we aim to improve the performance of best effort traffic. Hence, we propose two approaches: (a) the node-based approach; and (b) the path-based approach. In the node-based approach, we propose to assign a set of wavelengths to each OBS edge node that it can use to send its guaranteed traffic. This assignment takes into consideration physical distances between edge nodes. Furthermore, we propose a wavelength selection algorithm to improve the performance of best effort bursts. In the path-based approach, absolute quality of service provisioning is offered at end-to-end path level. To do this, we propose a routing and wavelength assignment approach which aims to reduce the number of wavelengths required to establish contention free paths. Nevertheless, if this objective cannot be reached because of the limited number of wavelengths in each fiber link, we propose an approach to synchronize overlapping paths without the need for additional equipments for synchronization. Here again, we propose a wavelength selection algorithm for best effort bursts. Numerical results show that both the node-based and the path-based approaches successfully provide absolute quality of service provisioning for guaranteed traffic and improve the performance of best effort traffic. Also, path-based approach could accommodate more guaranteed traffic and improve the performance of best effort traffic compared to node-based approach when the number of wavelengths is sufficient.