Intent-Based Network Management with User-Friendly Interfaces and Natural Language Processing

This paper investigates the scaling of business ordering, delivery and fulfillment of 5G/6G services especially applied to network slicing. We propose a system including architecture and methods to translate customer expectations expressed as intents into management services enabling network elements configuration. Our architecture uses a clear separation between engagement channels and the actual management of services and is compatible with the TM Forum Open Digital Architecture (ODA). Some components of the solution were successfully demonstrated such as a business intent resolver that interfaces with two engagement channels offering customers the choice of the user experience and to benefit from a Natural Language Processing based (NLP) interaction with a chatbot.

I. Introduction and Problem Statement

5G networks nowadays enable enterprises to benefit from private wireless networks either stand-alone or by means of Public-Network-Integrated Non-Public Networks (PNI-NPNs [1]) with opportunities for customization at a granularity unheard of before. In order to offer such solutions to a wide range of enterprises not necessarily knowledgeable about the technical configuration details of a mobile network, a simplified ordering is required as well as a vertical automation down to network deployment or configuration to meet business demands in a rapid timeline and to facilitate adaptation to expected demands raise and evolutions.