Non-RT RAN Intelligent Controller Functional Split Architecture

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The challenge of facilitating AI/ML lifecycle management in a non-real-time RIC is exacerbated by its ML-specific nature, making it difficult to adapt to a wide range of future applications without a functional split.

Innovation Solution

A method enabling a functional split between a non-real-time RIC and an external AI/ML server through a service management and orchestration entity, involving data collection, transfer, training model input, and configuration phases, with the aid of a transceiver and processor to facilitate intelligent beam management.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If the non-RT RIC maintains its ML-specific nature without functional split, then it can manage AI/ML lifecycle effectively, but it becomes difficult to adapt to a wide range of future applications

Engineering Contradiction:
Improveadaptability to future applicationsVSAvoidfunctional architecture complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent segments the non-RT RIC into two distinct functional components: a domain-generic SMO entity that handles resource management and orchestration, and a domain-specific non-RT RIC that focuses on AI/ML lifecycle management. This segmentation allows the system to maintain ML-specific capabilities while gaining adaptability to diverse applications through the generic SMO layer.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The SMO entity acts as an intermediary between the external environment and the ML-specific non-RT RIC. It provides a standardized interface that enables different applications and services to interact with the AI/ML lifecycle management without requiring changes to the core ML-specific functionality, thus enhancing adaptability while preserving specialized capabilities.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Ease of operation

If the non-RT RIC is made ML-specific for effective AI/ML lifecycle management, then AI/ML operations are optimized, but flexibility for diverse applications is reduced

Engineering Contradiction:
ImproveAI/ML lifecycle management efficiencyVSAvoidapplicability to different services
Core Design Contradiction:
Ease of operationVSAdaptability or versatility

Solution Approach 1:

The patent extracts the domain-specific AI/ML lifecycle management functionality into a separate non-RT RIC component, while the SMO entity handles domain-generic operations. This extraction allows the ML-specific component to be optimized for AI/ML operations without being constrained by diverse application requirements, while the SMO provides the necessary flexibility through standardized interfaces.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The SMO entity is designed with universal, domain-generic functions that can serve multiple different applications and services. By providing a standardized interface and common resource management capabilities, it enables the ML-specific non-RT RIC to maintain ease of operation for AI/ML lifecycle management while the SMO handles the adaptability to diverse services.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP4122169B1Functional architecture and interface for non-real-time ran intelligent controller
Publication Date: 2025.10.01 SAMSUNG ELECTRONICS CO LTD
  • EP4122169B1 patent drawingFigure 1
  • EP4122169B1 patent drawingFigure 2
  • EP4122169B1 patent drawingFigure 3

AI summary

The present disclosure relates to a 5G communication system or a 6G communication system for supporting higher data rates beyond a 4G communication system such as long term evolution (LTE). A service management and orchestration (SMO) entity enabling a functional split between a non-real-time (RT) radio access network (RAN) intelligent controller (RIC) and an external artificial intelligence (AI)/machine learning (ML) server will, during a data collection phase, utilize the SMO entity and the non-RT RIC to collect and process RAN data and non-RAN data and, during a data transfer phase, transfer processed RAN and non-RAN data from the SMO entity to an external AI/ML server via an interface. During a training model input phase, the SMO entity receives a trained AI/ML model, metadata, and training results from the external AI/ML server via an interface and, during a configuration phase, the SMO entity uses the trained AI/ML model within the SMO entity and the non-RT RIC to transfer configuration parameters to a near-RT RIC.