The next generation of mobile telecommunication, 5G, will be specified in the near future. One of the proposed changes relative to the previous generation, 4G,is the inclusion of a new system control plane (SCP). The purpose of the SCP is to improve system scalability, forward compatibility, peak performance and to enable a higher degree of support for advanced antenna techniques. This is done by logically separating data transmitted explicitly from and to the user, the dynamic transmissions, from the broadcasted transmissions that remain constant regardless of user activity, the static transmissions, and by then redesigning the static part to make it more lean. This is expected to have several positive effects such as considerably more free resources, resulting in energy savings and potentially increased usage of MIMO. Knowing what effect the SCP has upon aspects such as random access is of importance when designing the solution that will go into the standard. Simulations show that there is potential in the inclusion of the new SCP. While the simulated 5G candidate systems that include the SCP have an increased delay when running the random access procedure, some aspects of the procedure have been improved. The main differences relative to the simulated 4G system are the performance of the first message in the procedure, which increased, and the performance of the second message in the procedure, which decreased. The deficiencies found in the handling of the second message, however, can be alleviated by using a more proper algorithm and channel design than what was used in this study.

Today more devices are being connected to the Internet via mobile networks. With more devices in mobile networks, the workload on radio base stations increases. Radio base stations must be energy efficient and cheap which makes high-performance central processing units (CPUs) a bad alternative to meet the increasing workload. An alternative could be a graphics processing unit (GPU) which have a different hardware architecture more suitable for data parallel problems. This thesis has investigated the parallelisation possibilities in the user-equipment handling part of radio base stations, and the aim was to use a GPU to take advantage of the parallelism. The investigation found a mixed pipeline and data parallelism in user-equipment handling. A parallelism suitable for a graphics processing unit (GPU) execution. The tasks which handle user-equipment were divided into smaller communication-free sub-tasks. Sub-task batches of user-equipment were collected and offloaded to a GPU. A peak throughput gain of 62.2 times over the single-threaded CPU was achieved, but with an impact on latency with more than a magnitude. The latency was for all workloads at least 1.24 higher for the GPU implementations compared to the CPU implementations. A radio base station with many more user-equipment than the once existing today was simulated. For this radio base station, a gain of 14.0 times the single-threaded CPU was achieved, while the latency increased by 2.4 times. To really make use of a GPU implementation the number of user-equipment, the load, must be higher than in existing radio base stations today.

Current SDN controllers have been designed based on a monolithic approach that integrates all of services and applications into one single, huge program. The monolithic design of SDN controllers restricts programmers who build management applications to specific programming interfaces and services that a given SDN controller provides, making application development dependent on the controller, and thereby restricting portability of management applications across controllers. Furthermore, the monolithic approach means an SDN controller must be recompiled whenever a change is made, and does not provide an easy way to add new functionality or scale to handle large networks. To overcome the weaknesses inherent in the monolithic approach, the next generation of SDN controllers must use a distributed, microservice architecture that disaggregates the control plane by dividing the monolithic controller into a set of cooperative microservices. Disaggregation allows a programmer to choose a programming language that is appropriate for each microservice. In this dissertation, we describe steps taken towards disaggregating the SDN control plane, consider potential ways to achieve the goal, and discuss the advantages and disadvantages of each. We propose a distributed architecture that disaggregates controller software into a small controller core and a set of cooperative microservices. In addition, we present a software defined network programming framework called Umbrella that provides a set of abstractions that programmers can use for writing of SDN management applications independent of NB APIs that SDN controllers provide. Finally, we present an intent-based network programming framework called OSDF to provide a high-level policy based API for programming of network devices using SDN. <br>

Modem network control plane that supports versatile communication services (e.g. performance differentiation, access control, virtualization, etc.) is highly complex. Different control components such as routing protocols, security policy enforcers, resource allocation planners, quality of service modules, and more, are interacting with each other in the control plane to realize complicated control objectives. These different control components need to coordinate their actions, and sometimes they could even have conflicting goals which require careful handling. Furthermore, a lot of these existing components are distributed protocols running on large number of network devices. Because protocol state is distributed in the network, it is very difficult to tightly coordinate the actions of these distributed control components, thus inconsistent control actions could create serious problems in the network. As a result, such complexity makes it really difficult to ensure the optimality and consistency among all different components. Trying to address the complexity problem in the network control plane, researchers have proposed different approaches, and among these the centralized control plane architecture has become widely accepted as a key to solve the problem. By centralizing the control functionality into a single management station, we can minimize the state distributed in the network, thus have better control over the consistency of such state. However, the centralized architecture has fundamental limitations. First, the centralized architecture is more difficult to scale up to large network size or high requests rate. In addition, it is equally important to fairly service requests and maintain low request-handling latency, while at the same time having highly scalable throughput. Second, the centralized routing control is neither as responsive nor as robust to failures as distributed routing protocols. In order to enhance the responsiveness and robustness, one approach is to achieve the coordination between the centralized control plane and distributed routing protocols. In this thesis, we develop a centralized network control system, called Maestro, to solve the fundamental limitations of centralized network control plane. First we use Maestro as the central controller for a flow-based routing network, in which large number of requests are being sent to the controller at very high rate for processing. Such a network requires the central controller to be extremely scalable. Using Maestro, we systematically explore and study multiple design choices to optimally utilize modern multi-core processors, to fairly distribute computation resource, and to efficiently amortize unavoidable overhead. We show a Maestro design based on the abstraction that each individual thread services switches in a round-robin manner, can achieve excellent throughput scalability while maintaining far superior and near optimal max-min fairness. At the same time, low latency even at high throughput is achieved by Maestro's workload-adaptive request batching. Second, we use Maestro to achieve the coordination between centralized controls and distributed routing protocols in a network, to realize a hybrid control plane framework which is more responsive and robust than a pure centralized control plane, and more globally optimized and consistent than a pure distributed control plane. Effectively we get the advantages of both the centralized and the distributed solutions. Through experimental evaluations, we show that such coordination between the centralized controls and distributed routing protocols can improve the SLA compliance of the entire network.

Redes Definidas por Software ajudam a simplificar a programabilidade da rede ao desacoplar o plano de controle dos dispositivos de encaminhamento, e implementá-lo em um controlador logicamente centralizado. Apesar de permitir uma separação de conceitos mais clara, essa característica cria também uma relação de dependência entre controlador e dispositivos. Falhas no plano de controle prejudicam a visibilidade do estado da rede no controlador e podem tornar a rede inutilizável caso os dispositivos de encaminhamento sejam isolados. A relevância deste problema motivou uma série de propostas, incluindo a distribuição física de instâncias de controle e a delegação de tarefas aos dispositivos de encaminhamento. Esta dissertação contém a proposta e a avaliação de uma arquitetura que usa redes celulares de dados (4G) como enlaces reservas para o plano de controle. Nenhum trabalho anterior explorou esta ideia, apesar da pesquisa recente envolvendo Redes Definidas por Software e redes sem fio. A avaliação experimental permite uma melhor compreensão ao responder três perguntas: (i) Como o comportamento do tráfego do plano de controle é afetado pelas características de enlaces celulares, (ii) quão rapidamente o plano de controle é migrado para o enlace reserva quando uma falha ocorre e (iii) como funções de rede que dependem do estado da rede em um instante se comportam em tal arquitetura. Apesar da já esperada maior latência dos enlaces celulares, esta arquitetura mantém o funcionamento parcial de tarefas que dependem de visão global da rede quando falhas ocorrem nos enlaces primários, de maneira simples e com custo acessível. O grau de manutenção de tais tarefas é diretamente relacionado com sua dependência da rapidez de reação do plano de controle a eventos de rede. O principal benefício de prevenir a partição do plano de controle é a manutenção de uma visão global consistente da rede. / Software Defined Networks simplify network programmability by detaching the control plane from forwarding devices and deploying it into a logically centralized controller. While this allows a clearer separation of concerns, it also creates a dependency between them. Failures in the control plane break the controller view of the network state and could render the network unusable if forwarding devices cannot be reached. The relevance of this problem has led to a range of proposals, including physical distribution of controller instances and delegation of concerns to forwarding devices. This dissertation features the proposal and evaluation of an architecture that leverages cellular data networks (4G) as control plane backup links. No previous work has explored this idea, despite the recent research intersecting SDN and wireless networks. The experimental evaluation provides insights towards answering three research questions: (i) How is the behavior of control plane traffic affected by the characteristics of cellular links, (ii) how quickly is the control plane handed over to the backup link when a failure occurs and (iii) how well do network functions that rely on a snapshot of the network state behave on such an architecture. Despite the expected higher latency of cellular links, this architecture maintains partial functionality of tasks that depend on global network awareness when failures occur in primary control links in a simple, affordable fashion. The degree to which the functionality of these tasks is maintained is directly related to its dependency on the timeliness of control plane reaction to network events. The main benefit of preventing control plane partition is to maintain a consistent global view of the network.

Redes Definidas por Software ajudam a simplificar a programabilidade da rede ao desacoplar o plano de controle dos dispositivos de encaminhamento, e implementá-lo em um controlador logicamente centralizado. Apesar de permitir uma separação de conceitos mais clara, essa característica cria também uma relação de dependência entre controlador e dispositivos. Falhas no plano de controle prejudicam a visibilidade do estado da rede no controlador e podem tornar a rede inutilizável caso os dispositivos de encaminhamento sejam isolados. A relevância deste problema motivou uma série de propostas, incluindo a distribuição física de instâncias de controle e a delegação de tarefas aos dispositivos de encaminhamento. Esta dissertação contém a proposta e a avaliação de uma arquitetura que usa redes celulares de dados (4G) como enlaces reservas para o plano de controle. Nenhum trabalho anterior explorou esta ideia, apesar da pesquisa recente envolvendo Redes Definidas por Software e redes sem fio. A avaliação experimental permite uma melhor compreensão ao responder três perguntas: (i) Como o comportamento do tráfego do plano de controle é afetado pelas características de enlaces celulares, (ii) quão rapidamente o plano de controle é migrado para o enlace reserva quando uma falha ocorre e (iii) como funções de rede que dependem do estado da rede em um instante se comportam em tal arquitetura. Apesar da já esperada maior latência dos enlaces celulares, esta arquitetura mantém o funcionamento parcial de tarefas que dependem de visão global da rede quando falhas ocorrem nos enlaces primários, de maneira simples e com custo acessível. O grau de manutenção de tais tarefas é diretamente relacionado com sua dependência da rapidez de reação do plano de controle a eventos de rede. O principal benefício de prevenir a partição do plano de controle é a manutenção de uma visão global consistente da rede. / Software Defined Networks simplify network programmability by detaching the control plane from forwarding devices and deploying it into a logically centralized controller. While this allows a clearer separation of concerns, it also creates a dependency between them. Failures in the control plane break the controller view of the network state and could render the network unusable if forwarding devices cannot be reached. The relevance of this problem has led to a range of proposals, including physical distribution of controller instances and delegation of concerns to forwarding devices. This dissertation features the proposal and evaluation of an architecture that leverages cellular data networks (4G) as control plane backup links. No previous work has explored this idea, despite the recent research intersecting SDN and wireless networks. The experimental evaluation provides insights towards answering three research questions: (i) How is the behavior of control plane traffic affected by the characteristics of cellular links, (ii) how quickly is the control plane handed over to the backup link when a failure occurs and (iii) how well do network functions that rely on a snapshot of the network state behave on such an architecture. Despite the expected higher latency of cellular links, this architecture maintains partial functionality of tasks that depend on global network awareness when failures occur in primary control links in a simple, affordable fashion. The degree to which the functionality of these tasks is maintained is directly related to its dependency on the timeliness of control plane reaction to network events. The main benefit of preventing control plane partition is to maintain a consistent global view of the network.

Redes Definidas por Software ajudam a simplificar a programabilidade da rede ao desacoplar o plano de controle dos dispositivos de encaminhamento, e implementá-lo em um controlador logicamente centralizado. Apesar de permitir uma separação de conceitos mais clara, essa característica cria também uma relação de dependência entre controlador e dispositivos. Falhas no plano de controle prejudicam a visibilidade do estado da rede no controlador e podem tornar a rede inutilizável caso os dispositivos de encaminhamento sejam isolados. A relevância deste problema motivou uma série de propostas, incluindo a distribuição física de instâncias de controle e a delegação de tarefas aos dispositivos de encaminhamento. Esta dissertação contém a proposta e a avaliação de uma arquitetura que usa redes celulares de dados (4G) como enlaces reservas para o plano de controle. Nenhum trabalho anterior explorou esta ideia, apesar da pesquisa recente envolvendo Redes Definidas por Software e redes sem fio. A avaliação experimental permite uma melhor compreensão ao responder três perguntas: (i) Como o comportamento do tráfego do plano de controle é afetado pelas características de enlaces celulares, (ii) quão rapidamente o plano de controle é migrado para o enlace reserva quando uma falha ocorre e (iii) como funções de rede que dependem do estado da rede em um instante se comportam em tal arquitetura. Apesar da já esperada maior latência dos enlaces celulares, esta arquitetura mantém o funcionamento parcial de tarefas que dependem de visão global da rede quando falhas ocorrem nos enlaces primários, de maneira simples e com custo acessível. O grau de manutenção de tais tarefas é diretamente relacionado com sua dependência da rapidez de reação do plano de controle a eventos de rede. O principal benefício de prevenir a partição do plano de controle é a manutenção de uma visão global consistente da rede. / Software Defined Networks simplify network programmability by detaching the control plane from forwarding devices and deploying it into a logically centralized controller. While this allows a clearer separation of concerns, it also creates a dependency between them. Failures in the control plane break the controller view of the network state and could render the network unusable if forwarding devices cannot be reached. The relevance of this problem has led to a range of proposals, including physical distribution of controller instances and delegation of concerns to forwarding devices. This dissertation features the proposal and evaluation of an architecture that leverages cellular data networks (4G) as control plane backup links. No previous work has explored this idea, despite the recent research intersecting SDN and wireless networks. The experimental evaluation provides insights towards answering three research questions: (i) How is the behavior of control plane traffic affected by the characteristics of cellular links, (ii) how quickly is the control plane handed over to the backup link when a failure occurs and (iii) how well do network functions that rely on a snapshot of the network state behave on such an architecture. Despite the expected higher latency of cellular links, this architecture maintains partial functionality of tasks that depend on global network awareness when failures occur in primary control links in a simple, affordable fashion. The degree to which the functionality of these tasks is maintained is directly related to its dependency on the timeliness of control plane reaction to network events. The main benefit of preventing control plane partition is to maintain a consistent global view of the network.

Orientador: Helio Waldman / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-07T07:45:28Z (GMT). No. of bitstreams: 1 Pavani_GustavoSousa_D.pdf: 2684581 bytes, checksum: 17646b6816963a02bde2ea189d451458 (MD5) Previous issue date: 2006 / Resumo: Este trabalho aborda o uso de otimização por colônia de formigas (ACO) em redes ápticas. Como essa técnica exibe naturalmente capacidade de balanceamento de carga e auto-organização, além de prover controle distribuído na rede, apresentamos um algoritmo de roteamento baseado em formigas artificiais integrado a um plano de controle GMPLS, propondo seu uso nas seguintes aplicações: engenharia de tráfego e restauração em redes comutadas por caminhos ápticos e por pacotes ápticos, roteamento na presença de degenerações na camada física e gerenciamento integrado de caminhos ápticos e de recursos de processamento em arquiteturas de grid. O algoritmo proposto é comparado com técnicas tradicionais de roteamento baseado em topologia, como o roteamento por caminho mais curto. Apresentamos as vantagens e as limitações do algoritmo proposto em relação a essas técnicas para diferentes cenários, sendo que os resultados indicam que o algoritmo proposto neste trabalho é um excelente candidato para o controle de redes ápticas em substituição aos métodos tradicionais / Abstract: This work discusses the usage of Ant Colony Optimization (ACO) in optical networking. Since this technique exhibits naturally the capacity of network load-balancing and self-organization, besides it provides a distributed control of the network, we present a routing algorith:Q1 based on artificial ants, which is integrated with a GMPLS control plane. We propose the following applications for it: traffic engineering and restoration for wavelength-routed and optical packet switched networks, routing aware of physical-Iayer impairments, and integrated management of processing resources and lightpaths in grid architectures. The proposed algorithm is compared to conventional routing techniques based on topology, such as shortest-path routing. We present the advantages and limitations of the proposed algorithm when compared to those techniques considering different scenarios. The results indicate that the proposed algorithm is an excellent candidate to manage and control optical networks in lieu of the conventional methods / Doutorado / Telecomunicações e Telemática / Doutor em Engenharia Elétrica e de Computação

GMPLS is a still developing protocol family which is indented to assume the role of a control plane in transport networks. GMPLS is designed to provide traffic engineering in transport networks composed of different network technologies such as wavelength switched optical networks, Ethernet networks, point-to-point microwave links, etc. Integrating the different network technologies while using label switched paths to provide traffic engineering poses a challenge. The purpose of integrating multiple technologies under a single GMPLS control plane is to enable rapid service provisioning and efficient traffic engineering. Traffic engineering in networks provides two primary advantages, network resource utilization optimization and the ability to provide Quality of Service. Utilizing network resources more efficiently translates to lower expenditures for the network provider. Quality of Service can be used to provide the customer with for example guaranteed minimum bandwidth packet services. Specifically this thesis focused on the problems of signaling and establishing Forward Adjacency Label Switched Paths (FA-LSPs), and on a experimental method of connecting different network technologies. A testbed integrating an Ethernet network and a wave length division multiplexing network was used to show that the proposed solutions can work in practice. / GMPLS består av en samling protokoll under utveckling, de är tänkta att anta rollen som kontrollplan i transportnätverk. GMPLS är designat för att tillhandahålla trafikplanering i transportnätverk bestående av flera olika nätverksteknologier såsom Ethernet, våglängds switchande nätverk m.fl. Integration av dessa olika nätverksteknologier under ett gemensamt kontrollplan och uppsättning av ”label switched paths” i dataplanet är en utmaning. Syftet med att integrera multipla teknologier under ett ensamt GMPLS kontroll plan är att snabbt kunna tillhandahålla tjänster över nätverket samt möjliggöra advancerad trafikplanering. Trafikplanering i nätverk ger två stora fördelar, optimering av utnyttjandet av nätverksresurser samt ökade möjligheter att erbjuda ”Quality of Service” till kunder. Bättre utnyttjande av nätverksresurser innebär lägre kostnader för nätverksleverantören medans ”Quality of Service” kan ge kunden t.ex. en garanterad bandbredd. Specifikt fokuserar denna avhandling på problemen med att signalera och etablera ”Forwarding Adjaceny Label Switched Paths” samt en experimentell metod som båda sammankopplar olika typer av nätverk. En testbed bestående av ett Ethernet nätverk samt ett optiskt våglängdsswitchande nätverk användes för att visa att lösningarna kan fungera i praktiken.

Chapter 4 of this thesis is a minor revision of the work published in "Daisuke Kotani and Yasuo Okabe, Fast Failure Detection of OpenFlow Channels, The 11th Asian Internet Engineering Conference (AINTEC 2015), pp.32-39, November 2015. http://dx.doi.org/10.1145/2837030.2837035" © ACM 2015. / 京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第19847号 / 情博第598号 / 新制||情||104(附属図書館) / 32883 / 京都大学大学院情報学研究科知能情報学専攻 / (主査)教授 岡部 寿男, 教授 美濃 導彦, 教授 喜多 一 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM