| Summary: | The projection of 5G networks have caused a paradigm shift in the technologies that are used to support the preceding cellular networks. The fifth generation of mobile networks has been strongly driven by technological developments in the area of software-based networks, such as Virtualized Network Functions (NFV) and Software Defined Networks (SDN), which are drivers in changing the management of telecom infrastructures. In the face of demanding Quality of Service (QoS) and Quality of Experience (QoE) use cases, it is necessary to adapt the infrastructure, decentralizing and placing it at the edge. Allied to Cloud Computing, the possibility of having computing power at the edge, is called Multi-Access Edge Computing (MEC). Using the combined power of these technologies to create a new generation of mobile networks, it is possible to develop systems adapted to a growing demand for multimedia content consumption on the network, such as Virtual Content Delivery Networks (vCDN). The vCDNs are geographically distributed networks with the goal of having the access servers closer to the end user, in order to provide high availability of content with reduced latency. They are also differentiated by their scalability and flexibility, unattainable in previous systems. However, these advances introduce new challenges, namely in network automation, management and orchestration. To mitigate and solve these problems, several specialized platforms are constantly being developed and used, such as the Open Network Automation Platform (ONAP), an open source project that provides management and orchestration support for end-to-end services and infrastructures. This dissertation aims to address the QoS problems of a vCDN service, namely, hardware and application level data collection, so that, using ONAP, it is possible to create policies based on the data, in order to scale the service according to the network needs and form a closed automation loop.
|
|---|