Performance management in virtualized and programmable network environments (NFV and SDN)

• Autor(a) principal: Pedro, Rui Filipe
• Data de Publicação: 2017
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10773/23472
• topic: Sistemas de comunicação móveis; Redes de telecomunicações; Gestão de redes

More than tendencies and exploratory knowledge domains, there is a strong conviction in the industry that network function virtualization (NFV) and software-de ned network (SDN) paradigms came here to stay in the telecommunication services world. So that operators can surf the waves of change, they will have to signi cantly change their network architecture, their management mechanisms, and, simultaneously , their business model. This Master Thesis intends to contribute to the operators operation management mechanisms evolution, namely in the supervision/monitoring domains. More concretely, the work to be developed in this Master Thesis, will have as its main objective the evolution of the performance management platform (Altaia) from Altice Labs to the new networking paradigm based on the NFV and SDN concepts. It is also important to outline that the activities to be developed in the scope of this work, will be aligned in an international R&D project, nanced by the European Commission, covered by the H2020 5G-PPP program, that is SELFNET (A Framework for Self-Organized Network Management in Virtualized and Software De ned Networks). Altice Labs is one the participants in the SELFNET consortium, together with 10 more international partners. The project was a vast scope and intends to address Self-Organizing Network (SON) scenarios, and prediction over potential network and services prediction. It is in this context that the evolved version of Altaia will be introduced to the SELFNET project. This thesis proposes to develop the necessary tools to model, persist and perform real-time processing over network infrastructure data. More speci cally, this thesis developed: rstly the Raw and Aggregation Data Model, that uni ed both raw and aggregated data under a single model, secondly the Raw Data Loader, a component that receives network sensed data and transforms it so it can be persisted, lastly the Complex Event Processing Framework, a real-time processing framework for processing data using a dynamic rule-based approach. Out of what it was proposed no objective was left behind, all components were successful developed. The Raw and Aggregation Data Model, together with the Raw Data Loader, enforced SELFNET sensors to follow this model, thus unifying all sensed data under a known model. Moreover, the Complex Event Processing Framework was put into to place, with aggregation rules relative to the SELFNET Self- Protection use case, and is able to provide, in real-time, information about detected botnets around the underlying network. From the development of this framework a new component emerged, a Con guration Manager that manages data to be shared using distributed coordination services, used to apply dynamic rules over the Complex Event Processing Framework. This component is not only a contribution for SELFNET but as well as for Altice Labs. In a more high-level point-of-view, this task brought a new understanding about the role of network management tools for network operators and next-generation networks.