Data Centers (DCs) have become an intrinsic element of emerging technologies such as big data, artificial intelligence, cloud services; all of which entails interconnected and sophisticated computing and storage resources. Recent studies of conventional data center networks (DCNs) revealed two key challenges: a biased distribution of inter-rack traffic and unidentified flow classes: delay sensitive mice flows (MFs) and throughput-hungry elephant flows (EFs). Unfortunately, existing DCN topologies support only uniform distribution of capacities, provide limited bandwidth flexibilities and lacks of efficient flow classification mechanism. Fortunately, wireless DCs can leverage wireless communication emerging technologies, such as multi-terabit free-space optic (FSO), to provide flexible and reconfigurable DCN topologies. It is worth noting that indoor FSO links are less vulnerable to outdoor FSO channel impairments. Consequently, indoor FSO links are more robust and can offer high bandwidths with long stability, which can further be enhanced with wavelength division multiplexing (WDM) methods. In this thesis, we alleviate the bandwidth inefficiency by FSO links that have the desired agility by allocating the transmission powers to adapt link capacity for dynamically changing traffic conditions, and to reduce the maintenance costs and overhead. While routing the two classes along the same path causes unpleasant consequences, the DC researchers proposed traffic management solutions to treat them separately. However, the solutions either suffer from packet reordering and high queuing delay, or lack of accurate visibility and estimation on end-to-end path status. Alternatively, we leverage WDM to design elastic network topologies (i.e., part of the wavelengths are assigned to route MFs and the remaining for EFs). Since bandwidth demands can be lower than available capacity of WDM channels, we use traffic grooming to aggregate multiple flows into a larger flow and to enhance the link utilization. On the other hand, to reap the benefits of the proposed WDM isolated topology, an accurate and fast EF detection mechanism is necessary. Accordingly, we propose a scheme that uses TCP communication behavior and collect indicative packets for its flow classification algorithm, it demonstrates perfect flow classification accuracy, and is in order of magnitudes faster than existing solutions with low communication and computation overhead.

Orientadora: Profª. Drª. Christiane Marie Schweitzer / Dissertação (mestrado) - Universidade Federal do ABC, Programa de Pós-Graduação em Engenharia da Informação, 2014. / Os sistemas em nuvem representam uma nova solução para provedores de serviços de rede para empresas de pequeno e médio porte. Esses provedores possuem centrais de dados com grande quantidade de servidores, instalados sobre uma estrutura virtualizada e que permite a alocação dos espaços para diversas empresas que tem necessidade de hospedagem de serviços e dados e não possuem recursos financeiros para sua própria central de dados. Por meio das Redes Definidas por Software (Software Defined Networks - SDN) o plano de controle está separadodo plano de dados fazendo surgir uma nova solução de operação e gerenciamento das redes. Oplano de dados é composto por switches, roteadores e hosts, responsável pelo tráfego dainformação através de seus fluxos, enquanto o plano de controle é implementado emsoftwaree é responsável pelo roteamento de pacotes e arquitetura. Os controladores são responsáveispela manutenção dos fluxos. Atualmente os controladores disponíveis para utilização e decódigo aberto, são: NOX, POX, Maestro, Beacon e Floodlight. Assim, o uso de SDN emsistemas de computação em nuvem, permite uma flexibilização da operação e gerenciamentode centrais de dados. Neste sentido, este trabalho procurou avaliar a resistência de controladoresSDNs diante de ataques de negação de serviço, que possam interromper seu funcionamento,comprometendo a segurança da rede, por meio de testes constatou-se que o controlador NOXapresentou um comportamento bastante satisfatório, mantendo as conexões de rede ativas sobstress. Além disso, diante do problema que administradores de centrais de dados se defrontamquando precisam alocar ou aumentar as máquinas virtuais no conjunto de servidores da nuvem,nem sempre é possível alocar máquinas virtuais próximas umas das outras no mesmo servidor,interferindo na vazão da rede, foi investigada uma solução para permitir esta reorganização.Esta solução foi chamada de desfragmentador de redes e permite o deslocamento das máquinasvirtuais, fazendo com que sejam alocadas em um único conjunto físico de servidores. Estareorganização é feita sob demanda e em tempo real, e os resultados apresentados são bastante interessantes mostrando a viabilidade da solução. / Cloud systems represent a new solution for network service providers for small and mid-sized companies. These providers have datacenters with large amount of servers, installed on a virtualized structure and allowing the allocation of spaces for several companies that have needof hosting services and data and they do not have the financial resources to your own datacenter. In the Software Defined Networks (SDNs), the control plane is separated from the data plan making a new solution of networks operation and management. The data plan is composed of switches, routers and hosts, responsible for the traffic information through its streams, while the control plan is implemented in software and is responsible for the architecture and routing of packets. Controllers are responsible for the management of the streams. Currently, the available controllers, for use and open source, are: NOX, POX, Maestro, Beacon and Floodlight. Thus, the use of SDN, in cloud computing systems, allows a flexibility of operation and management of datacenters. In this sense, this work sought to evaluate the resistance of SDNs controllers in front of denial-of-service attacks that can disruptits operation, compromising network security, by testing it was found that the NOX controllerpresented a quite satisfying behavior, keeping active network connections under stress.Moreover, faced with the problem that datacenter administrators when they need to allocate orincrease the virtual machines in the cloud servers, it is not always possible to allocate virtualmachines near to each other on the same server, interfering with network flow, was investigateda solution to allow this reorganization. This solution was called the network defragmenter and allows the movement of virtual machines, which are allocated on a single set of physical servers. This reorganisation is made on demand and in real time, and the results are quite interesting, showing the viability of the solution.

The main aim of this work is to become familiar with the technology of software-defined networks and learn to use some of the tools to measure and simulate these networks.