Search

This paper presents a new approach to service assignment in Data Centers (DC), relating it to a classical combinatory problem called Bin Packing Problem and adding the possibility of delay and collaboration with users and energy providers. This possibility proves to reduce in much the energy consumption of the DC as well as the CO2 emissions.

Network virtualization is recognized as an enabling technology for the Future Internet that overcomes network ossification. However, it introduces a set of challenges. In any network virtualization environment, the problem of optimally mapping virtual demands to physical resources, known as virtual network embedding (VNE), is a crucial challenge. This paper analyses the behaviour of the main algorithms proposed to solve VNE by means of the ALEVIN framework. The VNE algorithms are evaluated with regard to appropriate metrics such as: cost, revenue, and virtual network acceptance ratio. We also analyse the impact of the recently introduced hidden hop demand concept in the performance of the VNE algorithms.

This paper introduces a framework to enhance network virtualization allocation in optical networks. A new control plane approach for the new generation of electrical and optical networks technologies is presented. To increase resources utilization efficiency, a new layer above virtualized elements is introduced in order to take into account the demands in order to control the resources and infrastructures below. For that layer, a strategy based on a resource squatting model is presented.

Descripció del recurs: 2 de maig de 2012 Vivimos completamente rodeados por una miríada de dispositivos electrónicos interconectados, que a su vez originan nuevos servicios y nuevas formas de entender y establecer las relaciones sociales. Nuevas palabras y conceptos acompañan inexorablemente dichos avances. Un progreso que, en muchas ocasiones, implica profundas reflexiones y ciertas dosis de incomprensión: ¿Cómo funcionan tales dispositivos? ¿Cómo es posible la conexión con las antípodas de forma tan simple? Este libro pretende dar respuestas a tales cuestiones, utilizando conceptos básicos de las telecomunicaciones y desafiando al lector a afrontar los retos tecnológicos de las comunicaciones.

The pervasive distribution of digital devices (e.g., laptops, mobile phone, digital camera, music players, RFID, sensors, smart cards) will create a large distributed computational and networking environment that will foster the evolution towards ecosystems of resource, services and data. Dynamicity and ubiquity of said future environments pose new challenges and requirements that current infrastructures cannot meet. Several limitations have to be overcomed, for instance: low flexibility and scalability, missing optimized cross-layer/cognitive (and cross-domain) resources allocation, limited potentialities of fostering new business models and above all a brittle integration of IT and network solutions/technologies. This paper proposes a systemic design to develop a future network architecture overcoming said limitations. Architecture is based on three main decentralised planes: the Knowledge Plane (KP), the Network Control Plane (NCP) and the Resource Control Plane (RCP). The three planes are implemented a distributed way and interact with each other (through a signalling overlay) for the allocation and management of virtual (communication, storage and processing) resources in overlay networks: novelty stands in coupling some enabling technologies with the creation and use of a cross-layer network knowledge. Paper also presents, as an example, some simulation results showing the optimal allocation of resources for the proposed architecture. Eventually, recommendations and future work conclude the paper.

Internet is becoming unable to overcome the challenges required by new services due to the lack of coordination among Internet service providers. This situation, frequently called network ossification, makes increasingly hard the deployment and the testing new network technologies. Network Virtualization has emerged as one solution of this problem. One network (substrate network), composed of physical nodes (e.g. routers) interconnected by means of links, may be used to create several isolated virtual networks, formed by virtual nodes and links, sharing the physical resources of the substrate network. One of the known problems of network virtualization is the allocation of physical resources. An uncontrolled allocation of bandwidth might entail bottlenecked virtual networks. The main goal of this paper is to analyze the problem of bandwidth allocation among virtual networks and to show mechanisms to offer a fair bandwidth distribution.

Bandwidth allocation is one of the main problems in network virtualization. Mechanisms to allocate bandwidth may avoid bottlenecked virtual links. This paper proposes a model based on optimization theory, to distribute the bandwidth among virtual links looking for the minimization of the spare bandwidth in the substrate network.

Bibliografia Actualmente existe una necesidad cada vez mayor de ofrecer servicios de red avanzados, que proporcionen un alto valor añadido al cliente. Eso es ya posible gracias a las nuevas capacidades de computación añadidas a los nodos de red, que proporcionan inteligencia y facilitan la gestión de red a los operadores y proveedores de servicios. En este texto, se presentan las nuevas redes inteligentes basadas en señalización número 7, los nuevos servicios sobre Internet (voz sobre IP, videoconferencia, etc.) y la aplicación de nuevas técnicas, como los agentes inteligentes y las redes activas.

Análisis de redes y sistemas de comunicaciones es una obra orientada a estudiantes de segundo ciclo de ingeniería de telecomunicaciones o telemática que buscan una guía de referencia y consulta en el campo de las redes de comunicaciones. Con esta obra se pretende llenar un vacío en esta área, mediante la introducción, de forma clara, de los conceptos fundamentales en redes de comunicaciones y en las técnicas para su análisis. El contenido se basa en la experiencia acumulada en las asignaturas de grado impartidas por sus autores en la ETSETB y describe el funcionamiento de las redes modernas, desde las de área local hasta las de banda ancha, y hace especial hincapié en el empleo de herramientas matemáticas en su estudio. De esta forma, el lector irá adquiriendo progresivamente los criterios que le facilitarán el análisis y el dimensionado de las redes y los dispositivos de comunicaciones. Los primeros capítulos presentan algunos conceptos generales sobre las redes, incluyendo modelos arquitectónicos, mecanismos de conmutación y cableados. A continuación se introducen las herramientas matemáticas necesarias para abordar el problema del modelado y análisis matemático, tras lo cual se exponen de forma clara y concisa las modernas tecnologías de red que permiten conocer, mediante el análisis, el comportamiento de las redes de área local. La obra también aborda las tecnologías vinculadas a las redes corporativas y concluye con las redes de banda ancha basadas en tecnología ATM.

This paper introduces a new strategy to provide QoS in IP/OBS networks, using Routing with Prioritization Based on Statistics (RPBS). This proposal uses the feedback scheme in optical networks to provide statistical knowledge with the objective of finding a suitable route for reach each destination from a specific source node, with more chance of success. This yields a twofold outcome. First, the losses can be reduced due to the statistics. Second, the delays are also reduced compare with other methods based on feedback scheme. These two improvements allow better QoS provision, supporting class differentiation and more efficient resources utilization. The benefits of this proposal are compared against existent alternatives by simulation. Keywords: OBS, Statistical Routing, WDM, QoS Provision, JET, RPBS