Avaliação da experiência do usuário com base na arbitragem de parâmetros de QoS em fluxo multimídia, em cenários com roteadores Cisco e placa NetFPGA

• Ano de defesa: 2016
• Autor(a) principal: Jacaúna, Rafael Sant'Ana
• Orientador(a): Moreno Ordonez, Edward David
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Sergipe
• Programa de Pós-Graduação: Pós-Graduação em Ciência da Computação
• País: Brasil
• Palavras-chave em Português: Computação; Tecnologia streaming (telecomunicação); Roteadores (redes de computadores); Rede de computador (protocolos); TCP/IP (protocolo de rede de computador); Fluxo de dados (computadores); Streaming; QoS; QoE; TCP; UDP; NetFPGA; iPerf
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: https://ri.ufs.br/handle/riufs/3383

A streaming video has features that differs from other streams: the occupation of large amount of bandwidth, and the possible variation of scenes with consequent increase (or decrease) in amount of bytes transmitted. Applications for video streaming transmission (YouTube, Vimeo, Netflix, Telecine Play, etc.) have aroused the interest of the scientific community regarding the behavior of networks. Our proposal is to measure, from a video stream, the user experience (QoE) based on arbitration QoS parameters in a controlled environment which use NetFPGA. The results of the experiments has shown the efficacy of TCP over UDP, which at the stage of choosing the most appropriate protocol for testing, showed to be ineffective to the Gigabit Ethernet rate. These experiments occurred in distinct physical topologies; only Cisco routers, with and without QoS, with and without injection of traffic generated by software iperf in the network, and using NetFPGA board set to router only with and without traffic via iperf, once the design developed to make the board work as a router does not allow QoS settings. in these scenarios, there was the need to start the client running at different times, setting their respective caches different networks so that did not coincide scenes, avoiding high bandwidth peaks in stretches with a lot of movement.