Tesis sobre el tema "Multi-layer networks"

Thesis (Ph. D.)--George Mason University, 2007.
Title from PDF t.p. (viewed Jan. 18, 2008). Thesis director: Bijan Jabbari Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Electrical and Computer Engineering. Vita: p. 113. Includes bibliographical references (p. 109-112). Also available in print.

Durant les dernières années, on a pu constater le succès des réseaux sociaux d'Internet avec un nombre considérable d'utilisateurs. Friendster. Com, par exemple, a plus de 40 millions d'utilisateurs. Récemment, ces réseaux sociaux sont utilisés comme base pour des systèmes de recommandation. Surtout les techniques de recommandation qui utilisent des réseaux de confiance ont donné des résultats très précis. La contribution principale de cette thèse consiste dans une approche qui permet de produire des recommandations basées sur la confiance pour des objets qui sont associés par des (hyper-)liens. C’est pourquoi un deuxième type de réseau, un réseau de documents, est considéré à côté du réseau social. Ce réseau de documents est par exemple établi par les citations entre des publications scientifiques ou les liens entre les pages Web. Les deux réseaux, et des réseaux supplémentaires comme des réseaux d’organisation sont intégrés dans un réseau à plusieurs strates. Dans une telle architecture, les mesures de visibilité sur le réseau de documents qui considèrent un document comme très visible s'il est cité par beaucoup de documents visibles, peuvent être combinées avec les recommandations basées sur les relations de confiance. La thèse décrit différentes mesures de visibilité renforcées par la confiance. De plus, une approximation de ces mesures est développée afin de manier l’incertitude qui est introduit par les copies de documents. Les mesures sont évaluées en simulation. Le système de recommandation SPRec implémente l'architecture à deux strates et génère des recommandations personnalisées par une application Web
The huge interest in social networking applications – Friendster. Com, for example, has more than 40 million users – led to a considerable research interest in using this data for generating recommendations. Especially recommendation techniques that analyze trust networks were found to provide very accurate and highly personalized results. The main contribution of this thesis is to extend the approach to trust-based recommendations, which up to now have been made for unlinked items such as products or movies, to linked resources, in particular documents. Therefore, a second type of network, namely a document reference network, is considered apart from the trust network. This is, for example, the citation network of scientific publications or the hyperlink graph of webpages. Recommendations for documents are typically made by reference-based visibility measures which consider a document to be the more important, the more often it is referenced by important documents. These two networks, as well as further networks such as organization networks, are integrated in a multi-layer network. This architecture allows for combining classical measures for the visibility of a document with trust-based recommendations, giving trust-enhanced visibility measures. Moreover, an approximation approach is introduced which considers the uncertainty induced by duplicate documents. These measures are evaluated in simulation studies. The trust-based recommender system for scientific publications SPRec implements a two-layer architecture and provides personalized recommendations via a web interface

Wireless communication has seen a tremendous growth in the last decades. Continuing on this trend, wireless multi-hop networks  are nowadays used or planned for use in a multitude of contexts, spanning from Internet access at home to emergency situations. The Transmission Control Protocol (TCP) provides reliable and ordered delivery of a data and is used by major Internet applications such as web browsers, email clients and file transfer programs. TCP traffic is also the dominating traffic type on the Internet. However, TCP performs less than optimal in wireless multi-hop networks due to packet reordering, low link capacity, packet loss and variable delay. In this thesis, we develop novel proposals for enhancing the network and transport layer to improve TCP performance in wireless multi-hop networks. As initial studies, we experimentally evaluate the performance of different TCP variants, with and without mobile nodes. We further evaluate the impact of multi-path routing on TCP performance and propose packet aggregation combined with aggregation aware multi-path forwarding as a means to better utilize the available bandwidth. The last contribution is a novel extension to multi-path TCP to  enable single-homed hosts to fully utilize the network capacity.

Opponent changed. Prof. C. Lindeman from the University of Leipzig was substituted by Prof. Zhang.

MULTI -hop wireless networks are usually defined as a collection of nodes equipped with radio transmitters, which not only have the capability to communicate each other in a multi-hop fashion, but also to route each others’ data packets. The distributed nature of such networks makes them suitable for a variety of applications where there are no assumed reliable central entities, or controllers, and may significantly improve the scalability issues of conventional single-hop wireless networks. This Ph.D. dissertation mainly investigates two aspects of the research issues related to the efficient multi-hop wireless networks design, namely: (a) network protocols and (b) network management, both in cross-layer design paradigms to ensure the notion of service quality, such as quality of service (QoS) in wireless mesh networks (WMNs) for backhaul applications and quality of information (QoI) in wireless sensor networks (WSNs) for sensing tasks. Throughout the presentation of this Ph.D. dissertation, different network settings are used as illustrative examples, however the proposed algorithms, methodologies, protocols, and models are not restricted in the considered networks, but rather have wide applicability. First, this dissertation proposes a cross-layer design framework integrating a distributed proportional-fair scheduler and a QoS routing algorithm, while using WMNs as an illustrative example. The proposed approach has significant performance gain compared with other network protocols. Second, this dissertation proposes a generic admission control methodology for any packet network, wired and wireless, by modeling the network as a black box, and using a generic mathematical 0. Abstract 3 function and Taylor expansion to capture the admission impact. Third, this dissertation further enhances the previous designs by proposing a negotiation process, to bridge the applications’ service quality demands and the resource management, while using WSNs as an illustrative example. This approach allows the negotiation among different service classes and WSN resource allocations to reach the optimal operational status. Finally, the guarantees of the service quality are extended to the environment of multiple, disconnected, mobile subnetworks, where the question of how to maintain communications using dynamically controlled, unmanned data ferries is investigated.

Network theory analysis is applicable to many scientific disciplines (fields) such as biology, statistics and sociology. The social network analysis is one of the various branches of the broader network theory analysis, the social network analysis. It is of high interest among the researchers in social sciences. Social networks have had a significant impact on human civilizations for many centuries. During the last two decades, the main academic interest was addressed towards the research and analysis of a dynamically uprising sector of social networks, the on-line networks, primarily due to the domination of the Internet and technology over human attitudes and relations in modern societies. For many years, research was emphasized on the analysis of simple social networks, whilst during the last decade several researchers started working on the analysis of more complicated social networks, which consist by several smaller social networks. There are important differences between mono and multi-dimensional network analysis. Mono-dimensional analysis provides the research with relevant knowledge. On the other hand, multi-dimensional analysis is still at initial stage. As a result, several potential models related to the multi-networks analysis cannot always provide reliable and adequate outcomes. However, due to the fact that different social networks can be easily combined and form more extended and complicated networks, it is of high importance for the researchers to advance the multi-dimensional analysis and provide more adequate analytical models. The purpose of this thesis is to present the dynamic of the multi-dimensional analysis by consecutively applying both mono and multi-dimensional analysis on a social multi-network. The findings suggest that multi-dimensional analysis can add reliable knowledge on the social network analysis, but many problems that arose due the complexity of the multi-networks structures need to be addressed.

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2011.
Cataloged from PDF version of thesis.
Includes bibliographical references (p. 87-90).
A cross-layer design approach is proposed that can be used to optimize the cooperative use of multi-packet reception (MPR) and network coding. A simple and intuitive model is constructed for the behavior of an opportunistic network coding scheme called COPE proposed by Katti et. al., MPR, the 802.11 MAC, and their combination. The model is then applied to key small canonical topology components and their larger counterparts. The results obtained from this model match the available experimental results with fidelity. Using this model, fairness allocation by the 802.11 MAC is shown to significantly impede performance and cause non-monotonic saturation behaviors; hence, a new MAC approach is devised that not only substantially improves throughput by providing monotonic saturation but provides fairness to flows of information rather than to nodes. Using this improved MAC, it is shown that cooperation between network coding and MPR achieves super-additive gains of up to 6.3 times that of routing alone with the standard 802.11 MAC. Furthermore, the model is extended to analyze the improved MAC's asymptotic, delay, and throughput behaviors. Finally, it is shown that although network performance is reduced under substantial asymmetry or limited implementation of MPR to a central/bottleneck node, there are some important practical cases, even under these conditions, where MPR, network coding, and their combination provide significant gains.
by Jason M. Cloud.
S.M.

Telecom operators networks' management and control remains partitioned by technology, equipment supplier and networking layer. In some segments, the network operations are highly costly due to the need of the individual, and even manual, configuration of the network equipment by highly specialized personnel. In multi-vendor networks, expensive and never ending integration processes between Network Management Systems (NMSs) and the rest of systems (OSSs, BSSs) is a common situation, due to lack of adoption of standard interfaces in the management systems of the different equipment suppliers. Moreover, the increasing impact of the new traffic flows introduced by the deployment of massive Data Centers (DCs) is also imposing new challenges that traditional networking is not ready to overcome. The Fifth Generation of Mobile Technology (5G) is also introducing stringent network requirements such as the need of connecting to the network billions of new devices in IoT paradigm, new ultra-low latency applications (i.e., remote surgery) and vehicular communications. All these new services, together with enhanced broadband network access, are supposed to be delivered over the same network infrastructure. In this PhD Thesis, an holistic view of Network and Cloud Computing resources, based on the recent innovations introduced by Software Defined Networking (SDN), is proposed as the solution for designing an end-to-end multi-layer, multi-technology and multi-domain cloud and transport network management architecture, capable to offer end-to-end services from the DC networks to customers access networks and the virtualization of network resources, allowing new ways of slicing the network resources for the forthcoming 5G deployments. The first contribution of this PhD Thesis deals with the design and validation of SDN based network orchestration architectures capable to improve the current solutions for the management and control of multi-layer, multi-domain backbone transport networks. These problems have been assessed and progressively solved by different control and management architectures which has been designed and evaluated in real evaluation environments. One of the major findings of this work has been the need of developed a common information model for transport network's management, capable to describe the resources and services of multilayer networks. In this line, the Control Orchestration Protocol (COP) has been proposed as a first contriution towards an standard management interface based on the main principles driven by SDN. Furthermore, this PhD Thesis introduces a novel architecture capable to coordinate the management of IT computing resources together with inter- and intra-DC networks. The provisioning and migration of virtual machines together with the dynamic reconfiguration of the network has been successfully demonstrated in a feasible timescale. Moreover, a resource optimization engine is introduced in the architecture to introduce optimization algorithms capable to solve allocation problems such the optimal deployment of Virtual Machine Graphs over different DCs locations minimizing the inter-DC network resources allocation. A baseline blocking probability results over different network loads are also presented. The third major contribution is the result of the previous two. With a converged cloud and network infrastructure controlled and operated jointly, the holistic view of the network allows the on-demand provisioning of network slices consisting of dedicated network and cloud resources over a distributed DC infrastructure interconnected by an optical transport network. The last chapters of this thesis discuss the management and orchestration of 5G slices based over the control and management components designed in the previous chapters. The design of one of the first network slicing architectures and the deployment of a 5G network slice in a real Testbed, is one of the major contributions of this PhD Thesis.
La gestión y el control de las redes de los operadores de red (Telcos), todavía hoy, está segmentado por tecnología, por proveedor de equipamiento y por capa de red. En algunos segmentos (por ejemplo en IP) la operación de la red es tremendamente costosa, ya que en muchos casos aún se requiere con guración individual, e incluso manual, de los equipos por parte de personal altamente especializado. En redes con múltiples proveedores, los procesos de integración entre los sistemas de gestión de red (NMS) y el resto de sistemas (p. ej., OSS/BSS) son habitualmente largos y extremadamente costosos debido a la falta de adopción de interfaces estándar por parte de los diferentes proveedores de red. Además, el impacto creciente en las redes de transporte de los nuevos flujos de tráfico introducidos por el despliegue masivo de Data Centers (DC), introduce nuevos desafíos que las arquitecturas de gestión y control de las redes tradicionales no están preparadas para afrontar. La quinta generación de tecnología móvil (5G) introduce nuevos requisitos de red, como la necesidad de conectar a la red billones de dispositivos nuevos (Internet de las cosas - IoT), aplicaciones de ultra baja latencia (p. ej., cirugía a distancia) y las comunicaciones vehiculares. Todos estos servicios, junto con un acceso mejorado a la red de banda ancha, deberán ser proporcionados a través de la misma infraestructura de red. Esta tesis doctoral propone una visión holística de los recursos de red y cloud, basada en los principios introducidos por Software Defined Networking (SDN), como la solución para el diseño de una arquitectura de gestión extremo a extremo (E2E) para escenarios de red multi-capa y multi-dominio, capaz de ofrecer servicios de E2E, desde las redes intra-DC hasta las redes de acceso, y ofrecer ademas virtualización de los recursos de la red, permitiendo nuevas formas de segmentación en las redes de transporte y la infrastructura de cloud, para los próximos despliegues de 5G. La primera contribución de esta tesis consiste en la validación de arquitecturas de orquestración de red, basadas en SDN, para la gestión y control de redes de transporte troncales multi-dominio y multi-capa. Estos problemas (gestion de redes multi-capa y multi-dominio), han sido evaluados de manera incremental, mediante el diseño y la evaluación experimental, en entornos de pruebas reales, de diferentes arquitecturas de control y gestión. Uno de los principales hallazgos de este trabajo ha sido la necesidad de un modelo de información común para las interfaces de gestión entre entidades de control SDN. En esta línea, el Protocolo de Control Orchestration (COP) ha sido propuesto como interfaz de gestión de red estándar para redes SDN de transporte multi-capa. Además, en esta tesis presentamos una arquitectura capaz de coordinar la gestión de los recursos IT y red. La provisión y la migración de máquinas virtuales junto con la reconfiguración dinámica de la red, han sido demostradas con éxito en una escala de tiempo factible. Además, la arquitectura incorpora una plataforma para la ejecución de algoritmos de optimización de recursos capaces de resolver diferentes problemas de asignación, como el despliegue óptimo de Grafos de Máquinas Virtuales (VMG) en diferentes DCs que minimizan la asignación de recursos de red. Esta tesis propone una solución para este problema, que ha sido evaluada en terminos de probabilidad de bloqueo para diferentes cargas de red. La tercera contribución es el resultado de las dos anteriores. La arquitectura integrada de red y cloud presentada permite la creación bajo demanda de "network slices", que consisten en sub-conjuntos de recursos de red y cloud dedicados para diferentes clientes sobre una infraestructura común. El diseño de una de las primeras arquitecturas de "network slicing" y el despliegue de un "slice" de red 5G totalmente operativo en un Testbed real, es una de las principales contribuciones de esta tesis.
La gestió i el control de les xarxes dels operadors de telecomunicacions (Telcos), encara avui, està segmentat per tecnologia, per proveïdors d’equipament i per capes de xarxa. En alguns segments (Per exemple en IP) l’operació de la xarxa és tremendament costosa, ja que en molts casos encara es requereix de configuració individual, i fins i tot manual, dels equips per part de personal altament especialitzat. En xarxes amb múltiples proveïdors, els processos d’integració entre els Sistemes de gestió de xarxa (NMS) i la resta de sistemes (per exemple, Sistemes de suport d’operacions - OSS i Sistemes de suport de negocis - BSS) són habitualment interminables i extremadament costosos a causa de la falta d’adopció d’interfícies estàndard per part dels diferents proveïdors de xarxa. A més, l’impacte creixent en les xarxes de transport dels nous fluxos de trànsit introduïts pel desplegament massius de Data Centers (DC), introdueix nous desafiaments que les arquitectures de gestió i control de les xarxes tradicionals que no estan llestes per afrontar. Per acabar de descriure el context, la cinquena generació de tecnologia mòbil (5G) també presenta nous requisits de xarxa altament exigents, com la necessitat de connectar a la xarxa milers de milions de dispositius nous, dins el context de l’Internet de les coses (IOT), o les noves aplicacions d’ultra baixa latència (com ara la cirurgia a distància) i les comunicacions vehiculars. Se suposa que tots aquests nous serveis, juntament amb l’accés millorat a la xarxa de banda ampla, es lliuraran a través de la mateixa infraestructura de xarxa. Aquesta tesi doctoral proposa una visió holística dels recursos de xarxa i cloud, basada en els principis introduïts per Software Defined Networking (SDN), com la solució per al disseny de una arquitectura de gestió extrem a extrem per a escenaris de xarxa multi-capa, multi-domini i consistents en múltiples tecnologies de transport. Aquesta arquitectura de gestió i control de xarxes transport i recursos IT, ha de ser capaç d’oferir serveis d’extrem a extrem, des de les xarxes intra-DC fins a les xarxes d’accés dels clients i oferir a més virtualització dels recursos de la xarxa, obrint la porta a noves formes de segmentació a les xarxes de transport i la infrastructura de cloud, pels propers desplegaments de 5G. La primera contribució d’aquesta tesi doctoral consisteix en la validació de diferents arquitectures d’orquestració de xarxa basades en SDN capaces de millorar les solucions existents per a la gestió i control de xarxes de transport troncals multi-domini i multicapa. Aquests problemes (gestió de xarxes multicapa i multi-domini), han estat avaluats de manera incremental, mitjançant el disseny i l’avaluació experimental, en entorns de proves reals, de diferents arquitectures de control i gestió. Un dels principals troballes d’aquest treball ha estat la necessitat de dissenyar un model d’informació comú per a les interfícies de gestió de xarxes, capaç de descriure els recursos i serveis de la xarxes transport multicapa. En aquesta línia, el Protocol de Control Orchestration (COP, en les seves sigles en anglès) ha estat proposat en aquesta Tesi, com una primera contribució cap a una interfície de gestió de xarxa estàndard basada en els principis bàsics de SDN. A més, en aquesta tesi presentem una arquitectura innovadora capaç de coordinar la gestió de els recursos IT juntament amb les xarxes inter i intra-DC. L’aprovisionament i la migració de màquines virtuals juntament amb la reconfiguració dinàmica de la xarxa, ha estat demostrat amb èxit en una escala de temps factible. A més, l’arquitectura incorpora una plataforma per a l’execució d’algorismes d’optimització de recursos, capaços de resoldre diferents problemes d’assignació, com el desplegament òptim de Grafs de Màquines Virtuals (VMG) en diferents ubicacions de DC que minimitzen la assignació de recursos de xarxa entre DC. També es presenta una solució bàsica per a aquest problema, així com els resultats de probabilitat de bloqueig per a diferents càrregues de xarxa. La tercera contribució principal és el resultat dels dos anteriors. Amb una infraestructura de xarxa i cloud convergent, controlada i operada de manera conjunta, la visió holística de la xarxa permet l’aprovisionament sota demanda de "network slices" que consisteixen en subconjunts de recursos d’xarxa i cloud, dedicats per a diferents clients, sobre una infraestructura de Data Centers distribuïda i interconnectada per una xarxa de transport òptica. Els últims capítols d’aquesta tesi tracten sobre la gestió i organització de "network slices" per a xarxes 5G en funció dels components de control i administració dissenyats i desenvolupats en els capítols anteriors. El disseny d’una de les primeres arquitectures de "network slicing" i el desplegament d’un "slice" de xarxa 5G totalment operatiu en un Testbed real, és una de les principals contribucions d’aquesta tesi.

The interworking of multi-hop wireless networks can conveniently provide ubiquitous seamless service and continual access for network users. Ubiquity can be achieved through multi-technology mobile terminals (nodes), which enables network users to access any available network. In addition, such nodes extend network coverage by relaying traffic for each other. However, to exploit the benefits of the internetworking, it is desirable to have appropriate traffic engineering (TE) mechanisms. Taking into account the co-existence of different physical and Medium Access Control (MAC) layer standards in the individual networks, the multi-hop communication capability and the impairment-prone wireless medium, TE becomes challenging.

En les xarxes IP/MPLS sobre WDM on es transporta gran quantitat d'informacio, la capacitat de garantir que el trafic arriba al node de desti ha esdevingut un problema important, ja que la fallada d'un element de la xarxa pot resultar en una gran quantitat d'informacio perduda. Per garantir que el trafic afectat per una fallada arribi al node desti, s'han definit nous algoritmes d'encaminament que incorporen el coneixement de la proteccio en els dues capes: l'optica (WDM) i la basada en paquets (IP/MPLS). D'aquesta manera s'evita reservar recursos per protegir el trafic a les dues capes. Els nous algoritmes resulten en millor us dels recursos de la xarxa, ofereixen rapid temps de recuperacio, eviten la duplicacio de recursos i disminueixen el numero de conversions del trafic de senyal optica a electrica.
The use of optical technology in core networks combined with IP/Multi-Protocol Label Switching (MPLS) solution has been presented as a suitable choice for the next generation Internet architecture. The integration of both layers is facilitated by the development of Generalized MPLS (GMPLS). In this network architecture, a single fibre failure can result in potentially huge data losses as the effects propagate up and through the network causing disruptions in the service of many applications. This research provides and evaluates new QoSP routing schemes that consider both IP/MPLS and optical network layers to compute the paths and backup paths subject to the QoS requirements of the traffic. Although effort has been devoted in developing multi-layer routing algorithms that consider all switching layers, protection is not considered amongst them. This is considered in this thesis. Where electrical to optical signal conversions have been reduded as well as the avoidance of traffic duplications resulting in better use of the network resources.

Unlike traditional wireless networks, characterized by the presence of last-mile, static and reliable infrastructures, Mobile ad Hoc Networks (MANETs) are dynamically formed by collections of mobile and static terminals that exchange data by enabling each other's communication. Supporting multi-hop communication in a MANET is a challenging research area because it requires cooperation between different protocol layers (MAC, routing, transport). In particular, MAC and routing protocols could be considered mutually cooperative protocol layers. When a route is established, the exposed and hidden terminal problems at MAC layer may decrease the end-to-end performance proportionally with the length of each route. Conversely, the contention at MAC layer may cause a routing protocol to respond by initiating new routes queries and routing table updates. Multi-hop communication may also benefit the presence of pseudo-centralized virtual infrastructures obtained by grouping nodes into clusters. Clustering structures may facilitate the spatial reuse of resources by increasing the system capacity: at the same time, the clustering hierarchy may be used to coordinate transmissions events inside the network and to support intra-cluster routing schemes. Again, MAC and clustering protocols could be considered mutually cooperative protocol layers: the clustering scheme could support MAC layer coordination among nodes, by shifting the distributed MAC paradigm towards a pseudo-centralized MAC paradigm. On the other hand, the system benefits of the clustering scheme could be emphasized by the pseudo-centralized MAC layer with the support for differentiated access priorities and controlled contention. In this thesis, we propose cross-layer solutions involving joint design of MAC, clustering and routing protocols in MANETs. As main contribution, we study and analyze the integration of MAC and clustering schemes to support multi-hop communication in large-scale ad hoc networks. A novel clustering protocol, named Availability Clustering (AC), is defined under general nodes' heterogeneity assumptions in terms of connectivity, available energy and relative mobility. On this basis, we design and analyze a distributed and adaptive MAC protocol, named Differentiated Distributed Coordination Function (DDCF), whose focus is to implement adaptive access differentiation based on the node roles, which have been assigned by the upper-layer's clustering scheme. We extensively simulate the proposed clustering scheme by showing its effectiveness in dominating the network dynamics, under some stressing mobility models and different mobility rates. Based on these results, we propose a possible application of the cross-layer MAC+Clustering scheme to support the fast propagation of alert messages in a vehicular environment. At the same time, we investigate the integration of MAC and routing protocols in large scale multi-hop ad-hoc networks. A novel multipath routing scheme is proposed, by extending the AOMDV protocol with a novel load-balancing approach to concurrently distribute the traffic among the multiple paths. We also study the composition effect of a IEEE 802.11-based enhanced MAC forwarding mechanism called Fast Forward (FF), used to reduce the effects of self-contention among frames at the MAC layer. The protocol framework is modelled and extensively simulated for a large set of metrics and scenarios. For both the schemes, the simulation results reveal the benefits of the cross-layer MAC+routing and MAC+clustering approaches over single-layer solutions.

L'augment de volum de tràfic IP provocat per l'increment de serveis multimèdia com HDTV o vídeo conferència planteja nous reptes als operadors de xarxa per tal de proveir transmissió de dades eficient. Tot i que les xarxes mallades amb multiplexació per divisió de longitud d'ona (DWDM) suporten connexions òptiques de gran velocitat, aquestes xarxes manquen de flexibilitat per suportar tràfic d’inferior granularitat, fet que provoca un pobre ús d'ample de banda. Per fer front al transport d'aquest tràfic heterogeni, les xarxes multicapa representen la millor solució. Les xarxes òptiques multicapa permeten optimitzar la capacitat mitjançant l'empaquetament de connexions de baixa velocitat dins de connexions òptiques de gran velocitat. Durant aquesta operació, es crea i modifica constantment una topologia virtual dinàmica gràcies al pla de control responsable d’aquestes operacions. Donada aquesta dinamicitat, un ús sub-òptim de recursos pot existir a la xarxa en un moment donat. En aquest context, una re-optimizació periòdica dels recursos utilitzats pot ser aplicada, millorant així l'ús de recursos. Aquesta tesi està dedicada a la caracterització, planificació, i re-optimització de xarxes òptiques multicapa de nova generació des d’un punt de vista unificat incloent optimització als nivells de capa física, capa òptica, capa virtual i pla de control. Concretament s'han desenvolupat models estadístics i de programació matemàtica i meta-heurístiques. Aquest objectiu principal s'ha assolit mitjançant cinc objectius concrets cobrint diversos temes oberts de recerca. En primer lloc, proposem una metodologia estadística per millorar el càlcul del factor Q en problemes d'assignació de ruta i longitud d'ona considerant interaccions físiques (IA-RWA). Amb aquest objectiu, proposem dos models estadístics per computar l'efecte XPM (el coll d'ampolla en termes de computació i complexitat) per problemes IA-RWA, demostrant la precisió d’ambdós models en el càlcul del factor Q en escenaris reals de tràfic. En segon lloc i fixant-nos a la capa òptica, presentem un nou particionament del conjunt de longituds d'ona que permet maximitzar, respecte el cas habitual, la quantitat de tràfic extra proveït en entorns de protecció compartida. Concretament, definim diversos models estadístics per estimar la quantitat de tràfic donat un grau de servei objectiu, i diferents models de planificació de xarxa amb l'objectiu de maximitzar els ingressos previstos i el valor actual net de la xarxa. Després de resoldre aquests problemes per xarxes reals, concloem que la nostra proposta maximitza ambdós objectius. En tercer lloc, afrontem el disseny de xarxes multicapa robustes davant de fallida simple a la capa IP/MPLS i als enllaços de fibra. Per resoldre aquest problema eficientment, proposem un enfocament basat en sobre-dimensionar l'equipament de la capa IP/MPLS i recuperar la connectivitat i el comparem amb la solució convencional basada en duplicar la capa IP/MPLS. Després de comparar solucions mitjançant models ILP i heurístiques, concloem que la nostra solució permet obtenir un estalvi significatiu en termes de costos de desplegament. Com a quart objectiu, introduïm un mecanisme adaptatiu per reduir l'ús de ports opto-electrònics (O/E) en xarxes multicapa sota escenaris de tràfic dinàmic. Una formulació ILP i diverses heurístiques són desenvolupades per resoldre aquest problema, que permet reduir significativament l’ús de ports O/E en temps molt curts. Finalment, adrecem el problema de disseny resilient del pla de control GMPLS. Després de proposar un nou model analític per quantificar la resiliència en topologies mallades de pla de control, usem aquest model per proposar un problema de disseny de pla de control. Proposem un procediment iteratiu lineal i una heurística i els usem per resoldre instàncies reals, arribant a la conclusió que es pot reduir significativament la quantitat d'enllaços del pla de control sense afectar la qualitat de servei a la xarxa.
The explosion of IP traffic due to the increase of IP-based multimedia services such as HDTV or video conferencing poses new challenges to network operators to provide a cost-effective data transmission. Although Dense Wavelength Division Multiplexing (DWDM) meshed transport networks support high-speed optical connections, these networks lack the flexibility to support sub-wavelength traffic leading to poor bandwidth usage. To cope with the transport of that huge and heterogeneous amount of traffic, multilayer networks represent the most accepted architectural solution. Multilayer optical networks allow optimizing network capacity by means of packing several low-speed traffic streams into higher-speed optical connections (lightpaths). During this operation, a dynamic virtual topology is created and modified the whole time thanks to a control plane responsible for the establishment, maintenance, and release of connections. Because of this dynamicity, a suboptimal allocation of resources may exist at any time. In this context, a periodically resource reallocation could be deployed in the network, thus improving network resource utilization. This thesis is devoted to the characterization, planning, and re-optimization of next-generation multilayer networks from an integral perspective including physical layer, optical layer, virtual layer, and control plane optimization. To this aim, statistical models, mathematical programming models and meta-heuristics are developed. More specifically, this main objective has been attained by developing five goals covering different open issues. First, we provide a statistical methodology to improve the computation of the Q-factor for impairment-aware routing and wavelength assignment problems (IA-RWA). To this aim we propose two statistical models to compute the Cross-Phase Modulation variance (which represents the bottleneck in terms of computation time and complexity) in off-line and on-line IA-RWA problems, proving the accuracy of both models when computing Q-factor values in real traffic scenarios. Second and moving to the optical layer, we present a new wavelength partitioning scheme that allows maximizing the amount of extra traffic provided in shared path protected environments compared with current solutions. Specifically, we define several statistical models to estimate the traffic intensity given a target grade of service, and different network planning problems for maximizing the expected revenues and net present value. After solving these problems for real networks, we conclude that our proposed scheme maximizes both revenues and NPV. Third, we tackle the design of survivable multilayer networks against single failures at the IP/MPLS layer and WSON links. To efficiently solve this problem, we propose a new approach based on over-dimensioning IP/MPLS devices and lightpath connectivity and recovery and we compare it against the conventional solution based on duplicating backbone IP/MPLS nodes. After evaluating both approaches by means of ILP models and heuristic algorithms, we conclude that our proposed approach leads to significant CAPEX savings. Fourth, we introduce an adaptive mechanism to reduce the usage of opto-electronic (O/E) ports of IP/MPLS-over-WSON multilayer networks in dynamic scenarios. A ILP formulation and several heuristics are developed to solve this problem, which allows significantly reducing the usage of O/E ports in very short running times. Finally, we address the design of resilient control plane topologies in GMPLS-enabled transport networks. After proposing a novel analytical model to quantify the resilience in mesh control plane topologies, we use this model to propose a problem to design the control plane topology. An iterative model and a heuristic are proposed and used to solve real instances, concluding that a significant reduction in the number of control plane links can be performed without affecting the quality of service of the network.

Made available in DSpace on 2018-08-02T00:01:59Z (GMT). No. of bitstreams: 1 tese_4741_Ontology-Based Provisioning for TI ML TN - Barcelos PPF20160506-140153.pdf: 4589336 bytes, checksum: 7ddab62f0b037be73ddeab2edf797c82 (MD5) Previous issue date: 2015-12-18
O aprovisionamento é uma atividade importante na configuração de redes. A Recomendação ITU-T M.3400 define aprovisionamento de redes como sendo os "procedimentos necessários para se colocar um equipamento em serviço, não incluindo a instalação". Os aprovisionamentos de recursos e de serviços são desafios recentes no planejamento de redes de comunicação, sendo atividades importantes nos paradigmas de redes futuras, como as redes orientadas a serviços, as redes em nuvem e a virtualização de redes. Considerando os problemas identificados na literatura, esta tese investiga o uso de tecnologias semânticas, especialmente ontologias, para resolver o problema da falta de interoperabilidade na área e o uso dessas tecnologias como base para uma solução computacional capaz de aprovisionar redes de transporte multicamadas independentes de tecnologia considerando os estados dos equipamentos da rede. Esta tese tem como objetivo desenvolver uma solução computacional para as redes de transporte, contribuindo assim com a área de aprovisionamento de redes, uma subárea da gerência de redes. Para atingir esse objetivo, (i) uma Ontologia de Referência para redes de transporte multicamadas independentes de tecnologia foi construída com base na Recomendação ITU-T G.800 utilizando-se de uma linguagem de ontologias bem fundamentada e expressiva para a definição de uma semântica precisa para a área. Essa Ontologia de Referência permite a comunicação, a aprendizagem e a interoperação na área de redes de transporte. Além disso, (ii) um modelo de rede semanticamente melhorado para o aprovisionamento de redes de transporte, aqui chamado Ontologia Computacional em OWL, foi gerado a partir da Ontologia de Referência através de uma rígida engenharia de ontologias; e (iii) foi implementado um sistema baseado em conhecimento para aprovisionamento de redes de transporte que usa a Ontologia Computacional em OWL como base de conhecimento. Os resultados de um teste em uma Rede Óptica de Transporte confirmam que o sistema desenvolvido é capaz de realizar o aprovisionamento de circuitos e o aprovisionamento de conexões em redes de transportes multicamadas considerando os estados dos equipamentos.

During the route discovery process, each node receiving the route request packet (RReq) will retransmit it exactly once. A distant neighbor may accidentally receive/loose the only RReq and use it to announce a new route, although that link is inferior/superior for route reply packets (RRep) or actual message routing. Overall, the constructed route may be far from the optimal. All existing route discovery schemes (including DSR/AODV) apply retransmission during route discovery exactly once (1R). Based on a realistic physical layer model, we propose two new route discovery schemes: n-retransmission (nR, retransmitting exactly n times) and n-retransmission c-reception (ncRR), retransmitting until we either reach a total of n own retransmissions or c copies from neighbors are heard. We compare our two new scheme with the traditional one, under otherwise identical conditions (same metric, same packet reception probability on each link) and the same choices about possibly retransmitting again upon discovering a better route (R+) or discarding it (R1), generating route reply packet for every received RRep (B*), or for first and better discovered routes only (B2), and retransmitting RRep exactly once (A1), up to a maximum of three times (A3), or optimally u times decided by link quality (Au). Experimental results show that the proposed ncRR scheme (for n=2 and c=3 or c=4) achieves the best tradeoff between quality of route, success rate and message overhead in the route discovery process, followed by the nR scheme, and both of them are superior to the existing traditional schemes.

Given any form of mobility management through wireless communication, one useful enhancement is improving the reliability and robustness of transport-layer connections in a heterogeneous mobile environment. This is particularly true in the case of mobile networks with multiple vertical handovers. In this thesis, issues and challenges in mobility management for mobile terminals in such a scenario are addressed, and a number of techniques to facilitate and improve efficiency and the QoS for such a handover are proposed and investigated. These are initially considered in an end-to-end context and all protocols and changes happened in the middleware of the connection where the network is involved with handover issues and end user transparency is satisfied. This thesis begins by investigating mobility management solutions particularly the transport layer models, also making significant observation pertinent to multi-homing for moving networks in general. A new scheme for transport layer tunnelling based on SCTP is proposed. Consequently a novel protocol to handle seamless network mobility in heterogeneous mobile networks, named nSCTP, is proposed. Efficiency of this protocol in relation to QoS for handover parameters in an end-to-end connection while wired and wireless networks are available is considered. Analytically and experimentally it has been proved that this new scheme can significantly increase the throughput, particularly when the mobile networks roam frequently. The detailed plan for the future improvements and expansion is also provided.

The ever growing Internet Protocol (IP) traffic driven by both consumer and business applications pose a fundamental challenge for network operators. Optical networks offering terabyte capacity and advanced switching capability is at the forefront of transport technology. As optical networks and data traffic proliferate there is a growing need to scale cost effectively and allow for multiple switching technologies. Addressing this, the G.709 Optical Transport Network (OTN) combined with the intelligence of the Generalised Multi Protocol Label Switching (GMPLS) control plane is identified to reduce transport costs and deliver enhanced network and performance management functions. Connections are established using GMPLS suite of protocols namely routing, signaling and the Path Computation Element (PCE). The conventional wavelength resource reservation mechanism in the Optical Channel (OCh) layer is stud- ied and shortcomings identified. Two distributed wavelength resource reservation algorithms that improve end-to-end wavelength allocation are proposed and offer significant network blocking performance improvement. Further, the Resource Reservation Protocol with Traffic Engineering (RSVP- TE) / PCE wavelength reservation is extended to the newly defined Optical channel Data Unit (ODU) digital layer to provision at sub-wavelength granularity. Network blocking performance, rough estimates of unit cost and wavelength utilisation in comparison with OCh layer are analysed showing improved network blocking performance and wavelength utilisation at the expense of increased cost unit. Also at higher transmission bit rates of 40Gbps and beyond, it is inefficient to provision light paths without taking into consideration the optical performance requirements of the client signal. To demonstrate Impairment-Aware (lA) wavelength resource reservation, the conventional RSVP- TE and best performing proposed algorithm are adapted for 10/ 40 Gbps client signals and network blocking performance is observed. Further, hardware-accelerated control plane functions to overcome challenges posed by overhead processing for short-lived connections in dynamic high-speed networks is discussed.

In recent years there has been a proliferation of research on a number of wireless multi-hop networks that include mobile ad-hoc networks, wireless mesh networks, and wireless sensor networks (WSNs). Routing protocols in such networks are often required to meet design objectives that include a combination of factors such as throughput, delay, energy consumption, network lifetime etc. In addition, many modern wireless networks are equipped with multi-channel radios, where channel selection plays an important role in achieving the same design objectives. Consequently, addressing the routing problem together with cross-layer adaptations such as channel selection is an important issue in such networks. In this work, we study the joint routing and channel selection problem that spans two domains of wireless networks. The first is a cost-effective and scalable wireless-optical access networks which is a combination of high-capacity optical access and unethered wireless access. The joint routing and channel selection problem in this case is addressed under an anycasting paradigm. In addition, we address two other problems in the context of wireless-optical access networks. The first is on optimal gateway placement and network planning for serving a given set of users. And the second is the development of an analytical model to evaluate the performance of the IEEE 802.11 DCF in radio-over- fiber wireless LANs. The second domain involves resource constrained WSNs where we focus on route and channel selection for network lifetime maximization. Here, the problem is further exacerbated by distributed power control, that introduces additional design considerations. Both problems involve cross-layer adaptations that must be solved together with routing. Finally, we present an analytical model for lifetime calculation in multi-channel, asynchronous WSNs under optimal power control.

A fundamental problem in multi-hop wireless networks is to estimate their throughout capacity. The problem can be informally stated as follows: given a multi-hop wireless network and a set of source destination pairs, determine the maximum rate r at which data can be transmitted between each source destination pair. Estimating the capacity of a multi-hop wireless network is practically useful --- it yields insights into the fundamental performance limits of the wireless network and at the same time aids the development of protocols that can utilize the network close to this limit. A goal of this dissertation is to develop rigorous mathematical foundations to compute the capacity of any given multi-hop wireless network with known source-destination pairs. An important factor that affects the capacity of multi-hop wireless networks is radio interference. As a result, researchers have proposed increasingly realistic interference models that aim to capture the physical characteristics of radio signals. Some of the commonly used simple models that capture radio interference are based on geometric disk-graphs. The simplicity of these models facilitate the development of provable and often conceptually simple methods for estimating the capacity of wireless networks. A potential weakness of this class of models is that they oversimplify the physical process by assuming that the signal ends abruptly at the boundary of a geometric region (a disk for omni-directional antennas). A more sophisticated interference model is the physical interference model, also known as the Signal to Interference Plus Noise Ratio (SINR) model. This model is more realistic than disk-graph models as it captures the effects of signal fading and ambient noise. This work considers both disk-graph and SINR interference models. In addition to radio interference, the throughput capacity of a multi-hop wireless network also depends on other factors, including the specific paths selected to route the packets between the source destination pairs (routing), the time at which packets are transmitted (scheduling), the power with which nodes transmit (power control) and the rate at which packets are injected (rate control). In this dissertation, we consider three different problems related to estimating network capacity. We propose an algorithmic approach for solving these problems. We first consider the problem of maximizing throughput with the SINR interference model by jointly considering the effects of routing and scheduling constraints. Second, we consider the problem of maximizing throughput by performing adaptive power control, scheduling and routing for disk-graph interference models. Finally, we examine the problem of minimizing end-to-end latency by performing joint routing, scheduling and power control using the SINR interference model. Recent results have shown that traditional layered networking principles lead to inefficient utilization of resources in multi-hop wireless networks. Motivated by these observations, recent papers have begun investigating cross-layer design approaches. Although our work does not develop new cross-layered protocols, it yields new insights that could contribute to the development of such protocols in the future. Our approach for solving these multi-objective optimization problems is based on combining mathematical programming with randomized rounding to obtain polynomial time approximation algorithms with provable worst case performance ratios. For the problems considered in this work, our results provide the best analytical performance guarantees currently known in the literature. We complement our rigorous theoretical and algorithmic analysis with simulation-based experimental analysis. Our experimental results help us understand the limitations of our approach and assist in identifying certain parameters for improving the performance of our techniques.
Ph. D.

Thesis (Ph. D.)--West Virginia University, 2005.
Title from document title page. Document formatted into pages; contains xii, 185 p. : ill. (some col.). Includes abstract. Includes bibliographical references (p. 140-146).

Thesis (M.S. in Electrical Engineering)--Naval Postgraduate School, March 2006.
"March 2006." Thesis Advisor(s): Monique P. Fargues, Ravi Vaidyanathan. Includes bibliographical references (p. 159-162). Also available online.

In my dissertation, we focus on the wireless network coexistence problem with advanced physical-layer techniques. For the first part, we study the problem of Wireless Body Area Networks (WBAN)s coexisting with cross-technology interference (CTI). WBANs face the RF cross-technology interference (CTI) from non-protocol-compliant wireless devices. Werst experimentally characterize the adverse effect on BAN caused by the CTI sources. Then we formulate a joint routing and power control (JRPC) problem, which aims at minimizing energy consumption while satisfying node reachability and delay constraints. We reformulate our problem into a mixed integer linear programing problem (MILP) and then derive the optimal results. A practical JRPC protocol is then proposed. For the second part, we study the coexistence of heterogeneous multi-hop networks with wireless MIMO. We propose a new paradigm, called cooperative interference mitigation (CIM), which makes it possible for disparate networks to cooperatively mitigate the interference to/from each other to enhance everyone's performance. We establish two tractable models to characterize the CIM behaviors of both networks by using full IC (FIC) and receiver-side IC (RIC) only. We propose two bi-criteria optimization problems aiming at maximizing both networks' throughput, while cooperatively canceling the interference between them based on our two models. In the third and fourth parts, we study the coexistence problem with MIMO from a different point of view: the incentive of cooperation. We propose a novel two-round game framework, based on which we derive two networks' equilibrium strategies and the corresponding closed-form utilities. We then extend our game-theoretical analysis to a general multi-hop case, specifically the coexistence problem between primary network and multi-hop secondary network in the cognitive radio networks domain. In the final part, we study the benefits brought by reconfigurable antennas (RA). We systematically exploit the pattern diversity and fast reconfigurability of RAs to enhance the throughput of MWNs. Werst propose a novel link-layer model that captures the dynamic relations between antenna pattern, link coverage and interference. Based on our model, a throughput optimization framework is proposed by jointly considering pattern selection and link scheduling, which is formulated as a mixed integer non-linear programming problem.

The emerging use of multi-homed wireless devices along with simultaneous multi-path data transfer offers tremendous potentials to improve the capacity of multi-hop wireless networks. Concurrent Multi-path Transfer (CMT) over Stream Control Transmission Protocol (SCTP) is a form of reliable multi-path transport layer protocol with unique features that resonate with multi-path nature of the multi-hop wireless networks. The present thesis identifies and addresses some of the challenges of CMT-SCTP over wireless multi-hop networks. One main challenge raised by the multi-hop wireless network for CMT-SCTP is the out-of-order packet arrival. SCTP uses packet sequence number to ensure delivery. As such, the out-of-order packet arrival triggers receive buffer blocking in CMT-SCTP that causes throughput degradation. Another challenge in using CMT-SCTP over multi-hop wireless networks is the unfair resource allocation towards flows coming from farther away hops. The first part of this thesis focuses on integrating machine learning and network coding in CMT-SCTP to resolve the receive buffer blocking problem. Our state-of-the-art scheme uses Q-learning, a form of Reinforcement Learning (RL), to enable the network coding module to adjust to network dynamics. We confirm the veracity of our proposal by a queuing theory based mathematical model. Moreover, the effectiveness of the proposed scheme is demonstrated through simulations and testbed experiments. In the second part of the thesis, we investigate the fairness behavior of CMT-SCTP towards other CMT or non-CMT flows coming from farther away hops on a multi-hop wireless network. We introduce a Q-learning distributed mechanism to enhance fairness in CMT-SCTP. The proposed method uses Q-learning to acquire knowledge about the dynamics of the network. Consequently, the acquired knowledge is used to choose the best action to improve the fairness index of the network. We evaluate our proposal against standard CMT-SCTP and Resource Pool CMT-SCTP (CMT/RP-SCTP). In the third part of this thesis, we apply our findings in the second part to TCP to demonstrate that the benefits of our fairness mechanism can be extended to other transport layer protocols. The findings of this thesis bring us closer to realization of the vast potential of multi-path data transfer over multi-hop wireless networks.
Applied Science, Faculty of
Electrical and Computer Engineering, Department of
Graduate

A new scheme for training Rectified Linear Unit (ReLU) based feedforward neural networks is examined in this thesis. The project starts with the row-by-row updating strategy designed for Single-hidden Layer Feedforward neural Networks (SLFNs). This strategy exploits the properties held by ReLUs and optimizes each row in the input weight matrix individually, under the common optimization scheme. Then the Direct Updating Strategy (DUS), which has two different versions: Vector-Based Method (VBM) and Matrix-Based Method (MBM), is proposed to optimize the input weight matrix as a whole. Finally DUS is extended to Multi-hidden Layer Feedforward neural Networks (MLFNs). Since the extension, for general ReLU-based MLFNs, faces an initialization dilemma, a special structure MLFN is presented. Verification experiments are conducted on six benchmark multi-class classification datasets. The results confirm that MBM algorithm for SLFNs improves the performance of neural networks, compared to its competitor, regularized extreme learning machine. For most datasets involved, MLFNs with the proposed special structure perform better when adding extra hidden layers.
Ett nytt schema för träning av rektifierad linjär enhet (ReLU)-baserade och framkopplade neurala nätverk undersöks i denna avhandling. Projektet börjar med en rad-för-rad-uppdateringsstrategi designad för framkopplade neurala nätverk med ett dolt lager (SLFNs). Denna strategi utnyttjar egenskaper i ReLUs och optimerar varje rad i inmatningsviktmatrisen individuellt, enligt en gemensam optimeringsmetod. Därefter föreslås den direkta uppdateringsstrategin (DUS), som har två olika versioner: vektorbaserad metod (VBM) respektive matrisbaserad metod (MBM), för att optimera ingångsviktmatrisen som helhet. Slutli- gen utvidgas DUS till framkopplade neurala nätverk med flera lager (MLFN). Eftersom utvidgningen för generella ReLU-baserade MLFN står inför ett initieringsdilemma presenteras därför en MLFN med en speciell struktur. Verifieringsexperiment utförs på sex datamängder för klassificering av flera klasser. Resultaten bekräftar att MBM-algoritmen för SLFN förbättrar prestanda hos neurala nätverk, jämfört med konkurrenten, den regulariserade extrema inlärningsmaskinen. För de flesta använda dataset, fungerar MLFNs med den föreslagna speciella strukturen bättre när man lägger till extra dolda lager.

Thesis (Ph. D.)--Electrical and Computer Engineering, Georgia Institute of Technology, 2008.
Committee Chair: Ingram, Mary Ann; Committee Member: Andrew, Alfred; Committee Member: Copeland, John; Committee Member: Owen, Henry; Committee Member: Sivakumar, Raghupathy.

Traditionally, achieving good performance for a multi-hop wireless network is known to be difficult. The main approach to control the operation of such a network relies on a distributed paradigm, assuming that a centralized approach is not feasible. Relying on a distributed paradigm could be justified at the time when the basic technical building blocks (e.g., node computational power, communication technology, positioning technology) were the bottlenecks. Recent advances and breakthroughs in these technical areas along with the emergence of programmable networks with softwarized control plane intelligence allow us to consider employing a centralized optimization paradigm to control and manage the operation of a multi-hop wireless network. The programmable control provides a platform on which the centralized global network optimization paradigm can be supported. The benefits of a centralized network optimization lie specially in that a network may be configured in such a way that offers optimal performance, which is hardly possible for a network relying on distributed operation. The objectives of this dissertation are to fully understand the potential benefits of a centralized control plane for a multi-hop wireless network, to identify any new challenges under this new paradigm, and to devise innovative solutions for optimal performance via a centralized control plane. Given that the performance of a wireless network heavily depends on its physical layer capabilities, we will consider a number of advanced wireless technologies, including MIMO, full duplex, and interference cancellation at the physical layer. The focus is on building tractable computational models for these wireless technologies that can be used for modeling, analysis and optimization in the centralized control plane. Problem formulation and efficient solution procedures are developed for various centralized optimization problems across multiple layers. End-to-end throughput maximization is a key objective among these optimization problems on the centralized control plane and is used to demonstrate the superior advantage of this paradigm. We study several problems: - Integration of SIC and MIMO DoF IC. We propose to integrate MIMO Degree-of-Freedom (DoF) interface cancellation (IC) and Successive Interference Cancellation (SIC) in MIMO multi-hop network under DoF protocol model. We show that DoF-based IC and SIC can be jointly integrated to combat the interference more effectively and improve the end-to-end throughput significantly. We develop the necessary mathematical models to realize the idea in a multi-hop wireless network. - Full-Duplex MIMO Wireless Networks Throughput. We investigate the performance of MIMO full-duplex (FD) in a multi-hop network. We show that if IC is exploited, MIMO FD can achieve significant throughput gain over MIMO HD in a multi-hop network, which is contrary to the recent literature suggesting an unexpected marginal gain. Our proposed model handles the additional network interference by joint efficient link scheduling and interference cancellation. - PCP in Tactical Wireless Networking. We propose the idea of the Programmable Control Plane (PCP) for the tactical wireless network under the protocol model. PCP decouples the control and data plane and allows the network control layer functionalities to be dynamically configured to adapt to specific wireless channel conditions, customized applications and/or certain tactical situations. The proposed PCP functionalities are cast into a centralized optimization problem, which can be updated as needed and provide a centralized intelligence to manage the operation of a wireless MIMO multi-hop network under the protocol model. - UPCP in Heterogeneous Wireless Networks. We propose the idea of the Unified Programmable Control Plane (UPCP) for tactical heterogeneous wireless networks with interference management capabilities under the SINR model. The UPCP abstracts the complexity of the underlying network comprised of heterogeneous wireless technologies and provides a centralized intelligence over the network resources. We develop necessary mathematical model to realize the UPCP.
Ph. D.

Les réseaux sans fil multi-saut (WMN : Wireless multi-hop Networks) sont passés du stade de simple curiosité pour revêtir aujourd'hui un intérêt certain aussi bien du point de vue de la communauté de recherche que des opérateurs de réseaux et services. En analysant les services et applications fournis au sein des réseaux WMNs, nous pouvons constater que certaines applications telles que la visioconférence, la VoIP, etc sont sensibles au délai et nécessitent une certaine qualité de service (QoS). D'autres applications telles que le transfert de fichier, le streaming vidéo, etc. sont gourmands en terme d'utilisation de bande passante. Par conséquent, les architectures de communication des réseaux WMNs doivent intégrer des mécanismes de routage efficaces et adaptés pour répondre aux besoins des services et applications envisagés. Dans cette thèse, Nous nous intéressons à la problématique du routage dans les réseaux WMNs. Notre objectif est de proposer une nouvelle approche de routage qui prend en compte différents métriques de coûts. Tout d'abord, nous avons montré que le routage sous contraintes multiples est un problème NP complet et que trois étapes sont nécessaires à la conception d'une nouvelle solution de routage: (i) modélisation de l'interférence, (ii) l'estimation de la de la bande passante restante, (iii) l'estimation du délai à un saut. Suivant cette vision, nous avons proposé deux variantes du protocole de routage OLSR (SP-OLSR, S2P-OLSR) se basant sur la métrique SINR. Les résultats des simulations ont montré l'intérêt de la proposition dans un contexte de communication vocale (VoIP). Ensuite, nous avons proposé un algorithme d'estimation d'interférence à 2 sauts (2-HEAR) afin d'estimer la bande passante disponible. Puis, et sur la base de cet algorithme, nous avons proposé une nouvelle métrique de routage pour les WMNs: Estimated Balanced Capacity (EBC) en vue de parvenir à l'équilibrage de charge entre des différents flux. La dernière question abordée dans cette thèse est celle de l'estimation du délai à un saut. La solution proposée donne une borne du délai en se basant sur un modèle de file d'attente de type G/G/1. Enfin, nous avons englobé toutes les précédentes contributions pour mettre en place une nouvelle approche de routage hybride sous contraintes multiples. Ce protocole comporte une partie proactive utilisant la nouvelle métrique de routage (EBC) et une partie réactive qui permet de prendre en compte le délai relative à une connexion donné
There is a growing interest in wireless multi-hop networks (WMNs) since there are promising in opening new business opportunity for network operators and service providers. This research field aims at providing wireless communication means to carry different types of applications (FTP, Web browsing, video streaming, in addition to VoIP). Such applications have different constraints and their specific requirements in terms of Quality of Service (QoS) or performance metrics (delay jitter, end-to-end delay). We examine, in this thesis, the problem of routing in WMNs. Our main goal is to propose a new multi-metrics routing capable to fit these particular needs. In this thesis, we make several contributions toward WMN multi-constrained routing. First, we show that the multi-constrained path finding problem is NP-Complete and inherently a cross-layer issue, and that three steps are necessary to design the multi-metric routing protocol: (i) modeling of the inferring signal, (ii) estimation of the remaining bandwidth, (iii) estimation of the one-hop delay. Second, moving in such direction, we propose two enhanced versions of the OLSR routing protocol. The suggested protocols consider the SINR as a routing metric to build a reliable topology graph. Performance evaluation shows that utilizing such routing metric helps to improve significantly the VoIP application quality in the context of ad hoc network while maintaining a reasonable overhead cost. Third, we have proposed a 2-Hop interference Estimation Algorithm (2-HEAR) in order to estimate the available bandwidth. Then, and based on such algorithm, we have proposed a novel routing metric for WMNs: Estimated Balanced Capacity (EBC) in order to achieve load-balancing among the different flows. The next issue tackled in this thesis is the one-hop delay estimation, the one-hop delay is estimated by means of an analytical model based on G/G/1 queue. Finally, we have encompassed all the previous contributions to address our main goal, i.e. the design of a multi-constrained routing protocol for WMNs. A hybrid routing protocol is then proposed. This protocol is a junction of two parts : a proactive part that makes use of the previously estimated constraint, and a reactive part, which is triggered ”on demand” when news applications are expressed

In IEEE 802.11 WLANs, Link Adaptation mechanisms, which choose the transmission rate of each node, provoke unexpected and random variations on the effective channel capacity. When these changes are towards lower bitrates, inelastic flows, such as VoIP, can suffer from sudden congestion, which results on higher packet delays and losses. In this thesis, a VoIP codec adaptation algorithm is proposed as a solution, based on a cross-layer feedback from RTCP packets and the MAC layer, which can adapt the codecs of active calls to adjust them to the multirate scenario. A combination of this algorithm with a call admission control mechanism is also studied. The results show an important improvement in terms of the QoS of the already active flows as also in the total hotspot's capacity. Additionally, by defining a new Grade of Service related parameter, the Q-Factor, which captures the trade-off between dropping and blocking ratio and perceived speech quality, the codec adaptation algorithm can be tuned to achieve maximum capacity without severely penalizing any of those variables, and hence satisfying both technical and user quality requirements. Finally, a new QoS-enabled AP, which implements these enhancements is designed.
En las redes inalámbricas del estándar IEEE 802.11, los mecanismos de adaptación de enlace que eligen la tasa de transmisión de cada nodo, pueden provocar variaciones aleatorias e inesperadas en la capacidad efectiva del canal. Cuando estos cambios son hacia tasas de transmisión mas bajas, los flujos inelásticos, tales como los de VoIP, pueden de repente sufrir congestión, lo que se traduce en aumento de retrasos y pérdidas de paquetes. En esa tesis, se propone un algoritmo de adaptación de codificadores de voz como solución, basado en técnicas multinivel (cross-layer) que combinan el uso de información de diferentes capas, como los paquetes RTCP y la capa MAC, y que puede adaptar los codecs de las llamadas activas para ajustarlos al escenario "multi-rate". Adicionalmente, la combinación de este algoritmo con un mecanismo de control de admisión de llamadas (CAC) se ha estudiado. Los resultados muestran una importante mejora en términos de QoS de los flujos activos como también en la capacidad total del hotspot. Además, mediante la definición de un nuevo factor, el Q-Factor, que puede captar la compensación entre la tasa de corte y de bloqueo de llamadas y de la calidad percibida por esas, el algoritmo de adaptación de codecs se puede ajustar para lograr la máxima capacidad sin penalizar severamente ninguna de esas variables y así satisfacer los requisitos técnicos de calidad y los usuarios. Por último, un nuevo punto de acceso (AP) habilitado para ofrecer calidad de servicio, ha sido diseñado que lleva a cabo estas mejoras.

Thesis (Ph. D.)--West Virginia University, 2007.
Title from document title page. Document formatted into pages; contains xiii, 200 p. : ill. (some col.). Includes abstract. Includes bibliographical references (p. 143-151).

In recent years, Multi-hop Wireless Networks (MHWNs) have experienced an explosion of deployment due to the increasing demand for continuous connectivity regardless of the physical location. Internet predominant transport protocols, i.e. Transmission Control Protocol (TCP), face performance degradation in MHWNs because of the high loss and link failure rates. Several solutions have been proposed which are based on network state estimation or use information from MAC layer (called metrics) in a cross-layer manner to better comprehend the network state. The first part of this thesis provides a survey and comprehensive definition of common metrics from Physical, MAC, Network and Transport layers and thus provides a multi-criteria and hierarchical classification. After that, the effectiveness in reflecting network information of MAC metrics is also investigated in a systematic way by simulating various network situations and measuring the MAC metrics. Thus, the good MAC metric for congestion control which is coupled with the network contention level and the medium induced losses will be found out. From the results of the effectiveness study, new rate control schemes for transport protocols are proposed which adapt efficiently the source bit rate depending on the network condition provided by some MAC metrics. Through an extensive set of simulations, the performance of the proposed rate control schemes in MHWNs is investigated thoroughly with several network situations.

De nombreux problèmes réels relèvent d’une structure bi-relationnelle et peuvent être modélisés suivant des réseaux bipartis. Une telle modélisation permet l'utilisation de solutions standards pour la prédiction et/ou la recommandation de nouvelles relations entre objets de ces réseaux. La tâche de prédiction de liens est un problème largement étudié dans les réseaux simples, c’est-à-dire les réseaux avec un seul type d'interaction entre sommets. Cependant, pour les réseaux multicouche (i.e. réseaux avec plusieurs types d'arêtes entre sommets), ce problème n'est pas encore entièrement résolu.Cette thèse est motivée par l'importance d'une tâche réelle, à savoir la prédiction d'interaction entre un médicament et une cible thérapeutique. La recherche de candidats médicaments prometteurs pour une cible thérapeutique biologique donnée est une partie essentielle de la conception d’un médicament moderne. Dans cette thèse, nous modélisons ce problème comme une tâche de prédiction de lien dans un réseau multicouche biparti. Cette modélisation du problème permet de rassembler différentes sources d'information en une seule structure et ainsi d'améliorer la qualité de la prédiction d’un lien.Cette thèse se concentre sur le problème de la prédiction de liens dans les réseaux multicouches bipartis et apporte deux contributions principales à ce sujet. La première contribution est une solution pour résoudre la prédiction de liens sans limiter le nombre et le type de réseaux, ce qui est le principal défaut des méthodes de l'état de l'art. L'algorithme que nous avons développé modélise une marche aléatoire à la manière du PageRank et est capable de prédire de nouvelles interactions dans le réseau que nous construisons à partir de différentes sources d'information. La deuxième contribution, qui porte aussi sur ce problème, s’appuie sur les méthodes de détection de communautés. Cette solution, moins immédiate et plus dépendante du choix des valeurs des paramètres, donne de meilleurs résultats. Pour cela, nous adaptons des mesures utilisées pour la détection de communautés à la problématique de la prédiction de liens dans les réseaux multicouche bipartis et nous développons de nouvelles méthodes associant des communautés pour la prédiction de liens. Nous évaluons aussi nos méthodes sur des données autres que celles des interactions entre médicaments et cibles thérapeutiques montrant ainsi le caractère générique de notre approche.D’autre part, nous proposons un protocole expérimental de validation des interactions prédites reposant sur l’exploitation de ressources externes. Fondé sur une collection de concepts biomédicaux utilisés comme source de connaissances, ce protocole effectue une validation des paires de médicaments-cibles thérapeutiques qui sont prédites à partir de scores de confiance que nous avons définis. Une évaluation des interactions prédites sur des données tests montre l'efficacité de ce protocole.Enfin, nous nous intéressons au problème de l'identification et de la caractérisation de composés promiscues qui existe dans le processus de développement de médicaments. Nous modélisons ce problème comme une tâche de classification et le résolvons par l'apprentissage automatique. Notre contribution repose sur une approche d'exploration de graphes et d'échantillonnage. De plus, nous avons développé une interface graphique pour fournir un retour d'information aux experts sur les résultats
Many aspects from real life with bi-relational structure can be modeled as bipartite networks. This modeling allows the use of some standard solutions for prediction and/or recommendation of new relations between these objects in such networks. Known as the link prediction task, it is a widely studied problem in network science for single graphs, networks assuming one type of interaction between vertices. For multi-layer networks, allowing more than one type of edges between vertices, the problem is not yet fully solved.The motivation of this thesis comes from the importance of an application task, drug-target interaction prediction. Searching valid drug candidates for a given biological target is an essential part of modern drug development. In this thesis, the problem is modeled as link prediction in a bipartite multi-layer network. Modeling the problem in this setting helps to aggregate different sources of information into one single structure and as a result to improve the quality of link prediction.The thesis mostly focuses on the problem of link prediction in bipartite multi-layer networks and makes two main contributions on this topic.The first contribution provides a solution for solving link prediction in the given setting without limiting the number and type of networks, the main constrains of the state of the art methods. Modeling random walk in the fashion of PageRank, the algorithm that we developed is able to predict new interactions in the network constructed from different sources of information. The second contribution, which solves link prediction using community information, is less straight-forward and more dependent on fixing the parameters, but provides better results. Adopting existing community measures for link prediction to the case of bipartite multi-layer networks and proposing alternative ways for exploiting communities, the method offers better performance and efficiency. Additional evaluation on the data of a different origin than drug-target interactions demonstrate the genericness of proposed approach.In addition to the developed approaches, we propose a framework for validation of predicted interactions founded on an external resource. Based on a collection of biomedical concepts used as a knowledge source, the framework is able to perform validation of drug-target pairs using proposed confidence scores. An evaluation of predicted interactions performed on unseen data shows effectiveness of this framework.At the end, a problem of identification and characterization of promiscuous compounds existing in the drug development process is discussed. The problem is solved as a machine learning classification task. The contribution includes graph mining and sampling approaches. In addition, a graphical interface was developed to provide feedback of the result for experts

The hybridization of soft computing methods of Radial Basis Function (RBF) neural networks, Multi Layer Perceptron (MLP) neural networks with back-propagation learning, fuzzy sets and rough sets are studied in the scope of this thesis. Conventional MLP, conventional RBF, fuzzy MLP, fuzzy RBF, rough fuzzy MLP, and rough fuzzy RBF networks are compared. In the fuzzy neural networks implemented in this thesis, the input data and the desired outputs are given fuzzy membership values as the fuzzy properties &ldquo
low&rdquo
, &ldquo
medium&rdquo
and &ldquo
high&rdquo
. In the rough fuzzy MLP, initial weights and near optimal number of hidden nodes are estimated using rough dependency rules. A rough fuzzy RBF structure similar to the rough fuzzy MLP is proposed. The rough fuzzy RBF was inspected whether dependencies like the ones in rough fuzzy MLP can be concluded.

Virtual optical networking supports the dynamic provisioning of dedicated networks over the same network infrastructure, which has received a lot of attention by network providers. The stringent network requirements (e.g., Quality of Service -QoS-, Service Level Agreement -SLA-, dynamicity) of the emerging high bandwidth and dynamic applications such as high-definition video streaming (e.g., telepresence, television, remote surgery, etc.), and cloud computing (e.g., real-time data backup, remote desktop, etc.) can be supported by the deployment of dynamic infrastructure services to build ad-hoc Virtual Optical Networks (VON), which is known as Infrastructure as a Service (IaaS). Future Internet should support two separate entities: infrastructure providers (who manage the physical infrastructure) and service providers (who deploy network protocols and offer end-to-end services). Thus, network service providers shall request, on a per-need basis, a dedicated and application-specific VON and have full control over it. Optical network virtualization technologies allow the partitioning/composition of the network infrastructure (i.e., physical optical nodes and links) into independent virtual resources, adopting the same functionality as the physical resource. The composition of these virtual resources (i.e., virtual optical nodes and links) allows the deployment of multiple VONs. A VON must be composed of not only a virtual transport plane but also of a virtual control plane, with the purpose of providing the required independent and full control functionalities (i.e., automated connection provisioning and recovery (protection/restauration), traffic engineering (e.g., QoS, SLA), etc.). This PhD Thesis focuses on optical network virtualization, with three main objectives. The first objective consists on the design, implementation and evaluation of an architecture and the necessary protocols and interfaces for the virtualization of a Generalized Multi-Protocol Label Switching (GMPLS) controlled Wavelength Switched Optical Network (WSON) and the introduction of a resource broker for dynamic virtual GMPLS-controlled WSON infrastructure services, whose task is to dynamically deploy VONs from service provider requests. The introduction of a resource broker implies the need for virtual resource management and allocation algorithms for optimal usage of the shared physical infrastructure. Also, the deployment of independent virtual GMPLS control plane on top of each VON shall be performed by the resource broker. This objective also includes the introduction of optical network virtualization for Elastic Optical Networks (EON). The second objective is to design, implement and experimentally evaluate a system architecture for deploying virtual GMPLS-controlled Multi-Protocol Label Switching Transport Profile (MPLS-TP) networks over a shared WSON. With this purpose, this PhD Thesis also focuses on the design and development of MPLS-TP nodes which are deployed on the WSON of the ADRENALINE Testbed at CTTC premises. Finally, the third objective is the composition of multiple virtual optical networks with heterogeneous control domains (e.g., GMPLS, OpenFlow). A multi-domain resource broker has been designed, implemented and evaluated.
La gestió de xarxes òptiques virtuals permet la provisió dinàmica de xarxes dedicades a sobre la mateixa infraestructura de xarxa i ha cridat molt l’atenció als proveïdors de xarxes. Els requisits de xarxa (per exemple la qualitat de servei, els acords de nivell de servei o la dinamicitat) són cada cop més astringents per a les aplicacions emergents d'elevat ample de banda i dinàmiques, que inclouen per exemple la reproducció en temps real de vídeo d'alta definició (telepresència, televisió, telemedicina) i serveis d’informàtica en núvol (còpies de seguretat en temps real, escriptori remot). Aquests requisits poden ser assolits a través del desplegament de serveis de infraestructura dinàmics per construir xarxes òptiques virtuals (VON, en anglès), fet que és conegut com a infraestructura com a servei (IaaS). La internet del futur hauria de suportar dos entitats diferenciades: els proveïdors d'infraestructures (responsables de gestionar la infraestructura física), i els proveïdors de serveis (responsables dels protocols de xarxa i d'oferir els serveis finals). D'aquesta forma els proveïdors de serveis podrien sol•licitar i gestionar en funció de les necessitats xarxes òptiques virtuals dedicades i específiques per les aplicacions. Les tecnologies de virtualització de xarxes òptiques virtuals permeten la partició i composició de infraestructura de xarxa (nodes i enllaços òptics) en recursos virtuals independents que adopten les mateixes funcionalitats que els recursos físics. La composició d'aquests recursos virtuals (nodes i enllaços òptics virtuals) permet el desplegament de múltiples VONs. Una VON no sols està composada per un pla de transport virtual, sinó també per un pla de control virtual, amb l'objectiu d'incorporar les funcionalitats necessàries a la VON (provisió de connexions automàtiques i recuperació (protecció/restauració), enginyeria de tràfic, etc.). Aquesta tesis es centra en la virtualització de xarxes òptiques amb tres objectius principals. El primer objectiu consisteix en el disseny, implementació i avaluació de l'arquitectura i els protocols i interfícies necessaris per la virtualització de xarxes encaminades a través de la longitud d'ona i controlades per GMPLS. També inclou la introducció d'un gestor de recursos per desplegar xarxes òptiques virtuals de forma dinàmica. La introducció d'aquest gestor de recursos implica la necessitat d'una gestió dels recursos virtuals i d’algoritmes d’assignació de recursos per a la utilització òptima dels recursos físics. A més el gestor de recursos ha de ser capaç del desplegament dels recursos assignats, incloent un pla de control GMPLS virtual independent per a cada VON desplegada. Finalment, aquest objectiu inclou la introducció de mecanismes de virtualització per a xarxes elàstiques òptiques (EON, en anglès). El segon objectiu és el disseny, la implementació i l’avaluació experimental d'una arquitectura de sistema per oferir xarxes MPLS-TP virtuals controlades per GMPLS sobre una infraestructura i WSON compartida. Per això, aquesta tesis també es centra en el disseny i desenvolupament d'un node MPLS-TP que ha estat desplegat al demostrador ADRENALINE, al CTTC. Finalment, el tercer objectiu és la composició de múltiples xarxes òptiques virtuals en dominis de control heterogenis (GMPLS i OpenFlow). Un gestor de recursos multi-domini ha estat dissenyat, implementat i avaluat.
La gestión de redes ópticas virtuales permite la provisión dinámica de redes dedicadas encima la misma infraestructura de red y ha llamado mucho la atención a los proveedores de redes. Los requisitos de red (por ejemplo la calidad de servicio, los acuerdos de nivel de servicio o la dinamicidad) son cada vez más estringentes para las aplicaciones emergentes de elevado ancho de banda y dinámicas, que incluyen por ejemplo la reproducción en tiempo real de vídeo de alta definición (telepresencia, televisión, telemedicina) y servicios de computación en la nube (copias de seguridad en tiempo real, escritorio remoto). Estos requisitos pueden ser logrados a través del despliegue de servicios de infraestructura dinámicos para construir redes ópticas virtuales (VON, en inglés), hecho que es conocido como infraestructura como servicio (IaaS). La internet del futuro tendrá que soportar dos entidades diferenciadas: los proveedores de infraestructuras (responsables de gestionar la infraestructura física), y los proveedores de servicios (responsables de los protocolos de red y de ofrecer los servicios finales). De esta forma los proveedores de servicios podrán solicitar y gestionar en función de las necesitados redes ópticas virtuales dedicadas y específicas por las aplicaciones. Las tecnologías de virtualización de redes ópticas virtuales permiten la partición y composición de infraestructura de red (nodos y enlaces ópticos) en recursos virtuales independientes que adoptan las mismas funcionalidades que los recursos físicos. La composición de estos recursos virtuales (nodos y enlaces ópticos virtuales) permite el despliegue de múltiples VONs. Una VON no sólo está compuesta por un plan de transporte virtual, sino también por un plan de control virtual, con el objetivo de incorporar las funcionalidades necesarias a la VON (provisión de conexiones automáticas y recuperación (protección/restauración), ingeniería de tráfico, etc.). Esta tesis se centra en la virtualización de redes ópticas con tres objetivos principales. El primer objetivo consiste en el diseño, implementación y evaluación de la arquitectura y los protocolos e interfaces necesarios por la virtualización de redes encaminadas a través de la longitud de ola y controladas por GMPLS. También incluye la introducción de un gestor de recursos para desplegar redes ópticas virtuales de forma dinámica. La introducción de este gestor de recursos implica la necesidad de una gestión de los recursos virtuales y de algoritmos de asignación de recursos para la utilización óptima de los recursos físicos. Además el gestor de recursos tiene que ser capaz del despliegue de los recursos asignados, incluyendo un plan de control GMPLS virtual independiente para cada VON desplegada. Finalmente, este objetivo incluye la introducción de mecanismos de virtualización para redes elásticas ópticas (EON, en inglés). El segundo objetivo es el diseño, la implementación y la evaluación experimental de una arquitectura de sistema para ofrecer redes MPLS-TP virtuales controladas por GMPLS sobre una infraestructura WSON compartida. Por eso, esta tesis también se centra en el diseño y desarrollo de un nodo MPLS-TP que ha sido desplegado al demostrador ADRENALINE, en el CTTC. Finalmente, el tercer objetivo es la composición de múltiples redes ópticas virtuales en dominios de control heterogéneos (GMPLS y OpenFlow). Un gestor de recursos multi-dominio ha sido diseñado, implementado y evaluado.

Assessing the short-term socio-economic impacts of climate-led disasters on food trade networks requires new bottom-up models and vulnerability metrics rooted in complexity theory. Indeed, such shocks could generate cascading socio-economic losses across the networks layers where emerging agents¿ responses could trigger tipping points. We contribute to address this research gap by developing a multi-layer behavioral network methodology composed of multiple spatially-explicit layers populated by heterogeneous interacting agents. Then, by introducing a new multi-layer risk measure called vulnerability rank, or VRank, we quantify the stress in the aftermath of a shock. Our approach allows us to analyze both the supply- and the demand-side dimensions of the shock by quantifying short-term behavioral responses, the transmission channels across the layers, the conditions for reaching tipping points, and the feedback on macroeconomic indicators. By simulating a stylized two-layer supply-side production and demand-side household network model we find that, (i) socio-economic vulnerability to climate-led disasters is cyclical, (ii) the distribution of shocks depends critically on the network structure, and on the speed of supply-side and demand-side responses. Our results suggest that such a multi-layer framework could provide a comprehensive picture of how climate-led shocks cascade and how indirect losses can be measured. This is crucial to inform effective post-disaster policies aimed to build food trade network resilience to climate-led shocks, in particular in more agriculture-dependent bread-basket regions.
Series: Ecological Economic Papers