Análise de uma metodologia MCDA baseada em modelos para fatiamento de redes 5G.

Abstract

Several challenges arise when planning the 5G deployment, such as the need to meet consolidated and emerging vertical markets and their associated requirements, which demand new solutions to enable the evolution of the network infrastructure through a service-oriented vision and provide greater customization and flexibility in its implementation and execution. To address these issues, the concept of Network Slicing arises, which is based on Software Defined Networks (SDN), Network Function Virtualization (NFV), and Multiple Access Edge Computing (MEC) to allow the creation of multiple end-to-end (E2E) virtual networks over the same infrastructure of fully or partially shared resources. These virtual networks can use different resource allocation models, which raises several perspectives and questions that need to be analyzed in the process of deploying the network slices since there is a wide variety of possible mappings between the resources of the infrastructure provider’s infrastructure. network and the service characteristics and requirements of a customer. This variety of mappings can cause an information overload that the network provider needs to handle to propose appropriate network slice alternatives for a certain service or business case. In this work, we aim to analyze and extend an existing methodology that uses multiple criteria decision-making (MCDA) techniques to address the problem of information overload in creating network slice instances in a shared resources infrastructure and we also seek to analyze the suitability of network slice instances created following this methodology to the main sets of use cases defined for 5G.