The Internet has exponentially grown and now it is part of our everyday life. Internet services and applications rely on back-end servers that are deployed on local servers and data centers. With the growing use of data centers and cloud computing, the locations of these servers have been externalized and centralized, taking advantage of economies of scale. However, some applications need to define complex network topologies and require more than simple connectivity to the remote sites. Therefore, the network part of cloud computing, what is called cloud networking, needs to be improved and simplified. This thesis argues that network virtualization permits to fill the missing gap and we propose a network virtualization abstraction layer to ease the use of cloud networking for the end users. We implement a software prototype of our ideas using OpenFlow. We also evaluate our prototype with state of the art controllers that has similar functionalities for network virtualization. A second part of this thesis focuses on developing a tool for performance testing. We have improved the widely used tool pktgen with receiver functionalities. We use pktgen to generate traffic for our experiments with network virtualization. / <p>QC 20160428</p>

In the era of ever-increasing electrification, power-electronic converters play the crucial role of transforming electrical energy from one form to another. However, converters today face multiple challenges in meeting ever-growing demands for higher power density, broader operation ranges, and lower costs. The cost-benefits of economies of scale further emphasize the need for modular and scalable converters. While no single converter for high-power applications satisfies all criteria, the modular multilevel converter (MMC) emerges as the clear frontrunner. MMC is extremely modular, being developed using multiple smaller units or building blocks called power electronic building blocks (PEBBs) or submodules (SMs). The SMs are themselves developed using fast-switching low-voltage (LV) semiconductors meaningfully combined with energy storage components like capacitors or batteries. MMCs are highly modular and scalable and have a very broad operation range, making them a key solution already used today for a wide range of high-voltage applications like high-voltage direct-current (HVDC) transmission. However, the use of voluminous and heavy capacitors in SMs also makes MMCs much lower in power density compared to other similar voltage source converters (VSCs). Employing at least twice the number of devices compared to a conventional two-level VSC for the same ratings also increases the converter costs. These challenges have hindered MMC applications in medium-voltage (MV) and more power-dense HVDC systems. This research aims to overcome these limitations by enhancing MMCs in terms of power density, efficiency, and cost-effectiveness. These modifications would expand MMC's applications to much broader HV and MV markets. Three fundamental aspects are targeted to achieve such improvements: Topology, Components, and Controls. The first modification focuses on changing the topology by replacing some fast-switching LV-switch-based SMs with fewer low-frequency HV/MV switches. This greatly reduces the total number of components and, when combined with intelligent control, decreases costs and losses. The second modification focuses on components, proposing the replacement of fully controlled MV switches with more efficient and cost-effective but partially controlled ones like thyristors. Despite thyristors' historical controllability challenges, incorporating SMs can help resolve control challenges, creating a modular, scalable converter with a wide operation range, high power density, and lower costs. The third avenue explores advanced control strategies while maintaining the traditional MMC topology. By accelerating and precisely controlling the capacitor current, the SM capacitor energy, SM capacitor size can be significantly reduced. Although these control methods are complex, they offer potential improvements across all five criteria: modularity, scalability, power density, cost, and operational range. These innovations extend MMCs' applicability to emerging fields such as energy storage systems, electric vehicle charging stations, motor drives, and data centers. Moreover, these modifications enhance MMCs for traditional high-voltage direct-current transmission applications. The research emphasizes the advantages and addresses each modification's limitations, paving the way for a more efficient and versatile power electronics technology. / Doctor of Philosophy / In our electrically powered world, the unsung workhorse is the power(-electronic) converter. Power converters play the crucial role of transforming electrical energy from one form to another using switches that can turn on and off to accurately control the flow of electrical energy. Power converters are critical to integrating systems at different voltage, current, and power ratings. For instance, power converters enable low-power systems like cellphone chargers and high-power industrial drives to be integrated into the same interconnected power grid. However, these converters face challenges in adapting to the evolving demands of our modern world. The expectations from power converters are high – they need to be affordable, lightweight, and capable of processing large amounts of power in a compact size. Additionally, modularity and scalability are desired qualities to enable economies of scale and bring the total cost down. Yet, finding a converter that fulfills all these criteria remains a challenge. The modular multilevel converter (MMC) is a promising power converter developed to address most of these considerations. Currently employed in high-power, high-voltage applications such as transmitting energy over vast distances or linking power grids between countries, the MMC is constructed using smaller power units or building blocks called power electronic building blocks (PEBBs) or submodules (SMs). These SMs utilize fast-switching low-voltage switches along with energy storage components like capacitors or batteries. Despite its versatility, the MMC faces many limitations. The main challenge for MMCs is the inability to process more power in lower volume, commonly referred to as power density. The MMC power density is low due to the use of large capacitors or batteries. Additionally, it utilizes twice the number of switches compared to traditional non-modular power converters for the same rating, leading to higher costs. These challenges restrict its application in medium-voltage and power-dense high-voltage high-power systems. This research aims to address these challenges, focusing on enhancing the power density and cost-effectiveness of MMCs. Three key areas of MMC are targeted for improvement: topology, components, and controls. Firstly, MMC's structure is reimagined, replacing many low-voltage switches with fewer medium- or high-voltage, fully-controlled switches. Such a system is referred to as a hybrid MMC, and this reduces the converter volume and costs. This adjustment has the added benefit of making the converter more efficient. Secondly, the focus is also on the components used to develop hybrid MMCs. Instead of fully controlled medium- or high-voltage switches, partially controlled switches like thyristors provide advantages like lower losses and higher power ratings. However, these partially controlled switches have traditionally been very difficult to control. Despite historical controllability challenges, incorporating these partially controlled switches in conjunction with smart control of SMs addresses control issues, creating a modular, scalable converter with high power density and lower costs. The third enhancement involves fundamental improvements to MMC controls. By managing the energy flow to the capacitor at a much faster rate and precision than conventional methods, the size of a critical component can be significantly reduced, opening avenues for overall improvements. Furthermore, such fast control introduces additional challenges like active control in the face of non-idealities and higher losses. This dissertation further meaningfully addresses these challenges to develop a much more power-dense MMC. These improvements transform the MMC and its variants into a versatile power converter family that can extend much beyond traditional MMC applications of high-voltage transmission applications. With these modifications, the MMC can be further positioned as an excellent candidate to contribute towards energy storage systems, electric vehicle charging stations, industrial-level motor drives, dc microgrids, and data centers, meeting the diverse needs of our equally diverse and ever-more electrified world.

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

Les workflows sont des systèmes typiques traitant le big data. Ces systèmes sont déployés sur des sites géo-distribués pour exploiter des infrastructures cloud existantes et réaliser des expériences à grande échelle. Les données générées par de telles expériences sont considérables et stockées à plusieurs endroits pour être réutilisées. En effet, les systèmes workflow sont composés de tâches collaboratives, présentant de nouveaux besoins en terme de dépendance et d'échange de données intermédiaires pour leur traitement. Cela entraîne de nouveaux problèmes lors de la sélection de données distribuées et de ressources de stockage, de sorte que l'exécution des tâches ou du job s'effectue à temps et que l'utilisation des ressources soit rentable. Par conséquent, cette thèse aborde le problème de gestion des données hébergées dans des centres de données cloud en considérant les exigences des systèmes workflow qui les génèrent. Pour ce faire, le premier problème abordé dans cette thèse traite le comportement d'accès aux données intermédiaires des tâches qui sont exécutées dans un cluster MapReduce-Hadoop. Cette approche développe et explore le modèle de Markov qui utilise la localisation spatiale des blocs et analyse la séquentialité des fichiers spill à travers un modèle de prédiction. Deuxièmement, cette thèse traite le problème de placement de données intermédiaire dans un stockage cloud fédéré en minimisant le coût de stockage. A travers les mécanismes de fédération, nous proposons un algorithme exacte ILP afin d’assister plusieurs centres de données cloud hébergeant les données de dépendances en considérant chaque paire de fichiers. Enfin, un problème plus générique est abordé impliquant deux variantes du problème de placement lié aux dépendances divisibles et entières. L'objectif principal est de minimiser le coût opérationnel en fonction des besoins de dépendances inter et intra-job / The typical cloud big data systems are the workflow-based including MapReduce which has emerged as the paradigm of choice for developing large scale data intensive applications. Data generated by such systems are huge, valuable and stored at multiple geographical locations for reuse. Indeed, workflow systems, composed of jobs using collaborative task-based models, present new dependency and intermediate data exchange needs. This gives rise to new issues when selecting distributed data and storage resources so that the execution of tasks or job is on time, and resource usage-cost-efficient. Furthermore, the performance of the tasks processing is governed by the efficiency of the intermediate data management. In this thesis we tackle the problem of intermediate data management in cloud multi-datacenters by considering the requirements of the workflow applications generating them. For this aim, we design and develop models and algorithms for big data placement problem in the underlying geo-distributed cloud infrastructure so that the data management cost of these applications is minimized. The first addressed problem is the study of the intermediate data access behavior of tasks running in MapReduce-Hadoop cluster. Our approach develops and explores Markov model that uses spatial locality of intermediate data blocks and analyzes spill file sequentiality through a prediction algorithm. Secondly, this thesis deals with storage cost minimization of intermediate data placement in federated cloud storage. Through a federation mechanism, we propose an exact ILP algorithm to assist multiple cloud datacenters hosting the generated intermediate data dependencies of pair of files. The proposed algorithm takes into account scientific user requirements, data dependency and data size. Finally, a more generic problem is addressed in this thesis that involve two variants of the placement problem: splittable and unsplittable intermediate data dependencies. The main goal is to minimize the operational data cost according to inter and intra-job dependencies