State Management of Knowledge Base in Intent Management Functions in 6G Networks

Abstract

As networks become increasingly complex with sophisticated underlying structures, automating configuration and management becomes more important than ever. Intent management addresses the network automation guided by requirements and constraints described to networks as intents submitted to an Intent Management Function (IMF). An IMF will then extract information from an intent and store it in a knowledge base. The knowledge base will be used by the IMF for inference and to fulfill requirements in the intent by setting goal(s) and trigger appropriate actions when needed. Since the knowledge base is a dynamic entity in an IMF, it is essential to keep track of changes and previous states and be able to apply changes on the knowledge base while keeping the main storage unchanged, e.g. for evaluation of different proposed action tracks, etc. In this paper we propose a solution for state management of a knowledge base. A prototype has been developed to prove the feasibility of the proposed solution and the evaluation results are presented.