Intelligent Routing for Software-Defined Media Networks

• Autor(a) principal: Simões, Diogo Miguel Gonçalves
• Data de Publicação: 2022
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10362/145233
• topic: Multimedia providers; Software-Defined Networking; Routing optimization; Machine learning; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática

The multimedia market is an industry with an ever-growing demand coupled with strict requirements. Be it in live streaming services or file content broadcast, multimedia providers need to deliver the best possible quality in order to meet their costumer’s requirements and gain or keep their trust. Multimedia traffic has a high impact on networks and, due to its nature, is sensitive to congestion or hardware failure. Thus, it is frequently that multimedia providers resort to third-party software to monitor quality parameters. Skyline Communications’ DataMiner® offers network monitoring, orchestrating and automation capabilities across a broad range of applications and environments. These features are enabled by the emergence of Software-Defined Networking (SDN) which provides a global view of networks and the ability to change network properties through software applications. This contrasts with traditional networks which are rigid, static and difficult to scale-up. An application that greatly benefits from the global network view of SDN is routing optimization. Through routing optimization, a network can effectively deliver more traffic by efficiently balancing load across the different links and paths between end points of a service, reaching an increased performance in data transport. This dissertation comes to light with the goal of optimizing DataMiner’s routing mechanism by exploring the routing optimization possibilities enabled by its SDN-like architecture. Both link cost optimization-based and Machine Learning (ML) approaches are evaluated as possible solutions to Skyline’s problem and several experiments were conducted to compare them and understand their impact on network performance while transporting multimedia streams.