Service Modelling and End-to-End Orchestration in 5G Networks

Abstract

5G represents a new vision over the networks and the services provided, aiming at realizing not only an evolution of mobile broadband networks but also at delivering it in a comprehensive infrastructure. Designed to attend the higher throughput, connection density, and latency demanded by myriad areas such as industrial automation, wireless health services, delivery of content everywhere, significant improvements in automation and dynamicity are required. To allow for the necessary transformations, Network Function Virtualization (NFV) and Software Defined Networking (SDN) emerged as the major forwarding technologies towards flexible and programmable software networks. Supported by a Telco Cloud vision of operators' infrastructures and software-based assets, End-to-End flexibility, is one of the key features of 5G networks. Although the upgrade by Telecom Operators to NFV- and SDN-based networks is complex and costly, it is nonetheless recognized as necessary to allow for CapEx and OpEx reduction, lifecycle improvements and for successful 5G services deployments. However, this transition introduces considerable challenges due to initial resistance to the design, deployment and operations over infrastructures and services, and due to the hybrid and multi-domain networks and services, which include both NFV aligned systems and legacy systems. The concept of orchestration comprehends the main logic to automate, enforce and execute processes and actions over the services and infrastructure, and it can operate both an ETSI NFV layer and also an End-to-End layer. The latter, in particular, provides the control and vision which allows the hybrid infrastructures and services to be made flexible and automated. Platform solutions exist to tackle orchestration needs. While some limit orchestration to the ETSI aligned NFV Management and Orchestration (MANO) level, such as Open Source MANO, the Open Network Automation Platform, ONAP, realizes End-to-End orchestration. It is the platform which operates above the infrastructure layer, providing network automation capabilities. This dissertation aims at leveraging the ONAP platform by analyzing, validating and extending it's Service Orchestration capabilities in its support for 5G networks. It focuses on the development and implementation of algorithms defined by execution flows, following its initial deployment and validation as a network and services Orchestrator.