This paper discusses the problems Distributed Caching can be used to solve and evaluates a couple of Distributed Caching Platforms targeting the .NET Framework. Basic concepts and functionality that is general for all distributed caching platforms is covered in chapter 2. We discuss how Distributed Caching can resolve synchronization problems when using multiple local caches, how a caching tier can relieve the database and improve the scalability of the system, and also how memory consumption can be reduced by storing data distributed. A couple of .NET-based caching platforms are evaluated and tested, these are Microsoft AppFabric Caching, ScaleOut StateServer and Alachisoft NCache. For a quick overview see the feature comparison-table in chapter 3 and for the main advantages and disadvantages of each platform see section 6.1. The benchmark results shows the difference in read performance, between local caching and distributed caching as well as distributed caching with a coherent local cache, for each evaluated Caching Platform. Local caching frameworks and database read times are included for comparison. These benchmark results are in chapter 5.

In recent years, adaptive HTTP streaming protocols have become the de facto standard in the industry for the distribution of live and video-on-demand content over the Internet. In this thesis, we solve the problem of distributing adaptive HTTP live video streams to a large number of viewers using peer-to-peer (P2P) overlays. We do so by assuming that our solution must deliver a level of quality of user experience which is the same as a CDN while trying to minimize the load on the content provider’s infrastructure. Besides that, in the design of our solution, we take into consideration the realities of the HTTP streaming protocols, such as the pull-based approach and adaptive bitrate switching. The result of this work is a system which we call SmoothCache that provides CDN-quality adaptive HTTP live streaming utilizing P2P algorithms. Our experiments on a real network of thousands of consumer machines show that, besides meeting the the CDN-quality constraints, SmoothCache is able to consistently deliver up to 96% savings towards the source of the stream in a single bitrate scenario and 94% in a multi-bitrate scenario. In addition, we have conducted a number of pilot deployments in the setting of large enterprises with the same system, albeit tailored to private networks. Results with thousands of real viewers show that our platform provides an average offloading of bottlenecks in the private network of 91.5%. These achievements were made possible by advancements in multiple research areas that are also presented in this thesis. Each one of the contributions is novel with respect to the state of the art and can be applied outside of the context of our application. However, in our system they serve the purposes described below. We built a component-based event-driven framework to facilitate the development of our live streaming application. The framework allows for running the same code both in simulation and in real deployment. In order to obtain scalability of simulations and accuracy, we designed a novel flow-based bandwidth emulation model. In order to deploy our application on real networks, we have developed a network library which has the novel feature of providing on-the-fly prioritization of transfers. The library is layered over the UDP protocol and supports NAT Traversal techniques. As part of this thesis, we have also improved on the state of the art of NAT Traversal techniques resulting in higher probability of direct connectivity between peers on the Internet. Because of the presence of NATs on the Internet, discovery of new peers and collection of statistics on the overlay through peer sampling is problematic. Therefore, we created a peer sampling service which is NAT-aware and provides one order of magnitude fresher samples than existing peer sampling protocols. Finally, we designed SmoothCache as a peer-assisted live streaming system based on a distributed caching abstraction. In SmoothCache, peers retrieve video fragments from the P2P overlay as quickly as possible or fall back to the source of the stream to keep the timeliness of the delivery. In order to produce savings, the caching system strives to fill up the local cache of the peers ahead of playback by prefetching content. Fragments are efficiently distributed by a self-organizing overlay network that takes into account many factors such as upload bandwidth capacity, connectivity constraints, performance history and the currently being watched bitrate. / <p>QC 20131122</p>

Cette thèse consiste à développer des mécanismes distribués pour la gestion de ressources dans les réseaux cellulaires futurs. Dans la première partie de cette thèse, les défis techniques et économiques pour la mise en œuvre des politiques de stockage distribuées dans les réseaux à petites cellules sont traités. En particulier, un mécanisme de stockage proactif est proposé permettant aux stations de base d'exploiter les informations extraites des réseaux sociaux afin d'estimer la popularité locale des fichiers avant de le stocker. Une autre approche de stockage optimisée est proposée pour les réseaux cellulaires ultra-denses tout en prenant en compte les variations instantanées de l'état des unités de stockage. Pour faciliter le déploiement de ces solutions de stockage, de nouveaux mécanismes économiques sont développés pour motiver les fournisseurs de contenu à coopérer avec les opérateurs réseaux et stocker leurs fichiers au sein des petites stations de base. Dans la deuxième partie de cette thèse, le problème de gestion du spectre est étudié dans des réseaux contenant des stations de base munies de capacités de stockage ainsi que dans les systèmes LTE-U. En particulier, une approche de gestion de backhaul distribuée est proposée pour des réseaux cellulaire ayant des capacités de stockage et des liens de backhaul hétérogènes. D'autre part, un modèle multi-jeux est proposé comme un nouvel outil de la théorie des jeux pour faire face aux nouveaux problèmes d'allocation de ressources qui émergent avec l'introduction de la technologie LTE-U dans les réseaux sans fil. A cet égard, un multi-jeux composé de deux sous-jeux de types différents est formulé pour optimiser la coexistence des stations de base LTE-U et utilisateurs WiFi sur les bandes non-licenciés, tout en empêchant les stations de base LTE-U de dégrader la performance du réseau WiFi. / This thesis consists in developing distributed mechanisms for resource allocation in next-generation cellular networks. In the first part of this thesis, the technical and economic challenges for the implementation of distributed storage policies in small cell networks are addressed. In particular, a proactive storage approach is proposed enabling the small base stations to exploit the information extracted from online social networks to estimate the local popularity of the files. Another optimized storage approach is proposed for ultra-dense cellular networks while accounting for the instantaneous variations of the state of the storage units. To facilitate the deployment of these storage solutions, new economic mechanisms are developed to motivate content providers to cooperate with network operators and store their files within the operators' small base stations. In the second part of this thesis, the problem of spectrum management is studied in cache-enabled small cell networks as well as LTE-U systems. In particular, a distributed backhaul management approach is proposed for cellular networks with heterogeneous backhaul links. On the other hand, a multi-game framework is proposed as a new game theoretic tool to cope with the new resource allocation problems that emerge with the introduction of LTEU technology in wireless networks. In this regard, a multi-game composed of two subgames of different types is formulated to optimize the coexistence of LTE-U base stations and WiFi users over unlicensed bands, while preventing LTE-U base stations from jeopardizing the WiFi users.