The Impact of Data Engineering on Service Quality in 5G-Enabled Cable and Media Networks

Abstract

The introduction of 5G technology significantly boosts the available transmission capacities, also in media and cable networks that offer fiber and HFC infrastructures. However, the high SLAs needed in this context, which request near-zero latency and real-time actions in case of failure triggering and recovery, make the use of open and heterogeneous four-layer architectures with a service abstraction at all layers become key factors. Several technological solutions, such as the use of ML and AI models at the service layer, the support of Intents, the YANG data models, and the present and coming efforts towards the standardized definition of zero-touch architectures, could be then used to overcome the service quality issues affecting user services in the layers below.

However, whenever a service is provided through the use of four-layer hierarchical architectures, this makes it needed to also define and introduce Impact Models (IMs). It is key not only to monitor the performance of the four-layered architecture but also to predict the impact of any changes in the configurations of the lower layers on the user service layer quality. A disruptive methodology for the definition and the introduction of a framework of IMs, which could improve service quality in the newly defined hierarchical software- and data-driven architectures, is Data Engineering. It plays a central role in data creation, control, and access, with the introduction of procedures for the automatic generation of all the needed Machine learning (ML) models. Its impact has been shown, in a convincing way, in the Digital Production Operation sector. It explains the worthy proposition of pursuing such a journey in the area of end-to-end service quality provision for media and cable networks.