AI Model Coordination in RAN for Complex 5G Network Operations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional operation and maintenance methods for 5G networks struggle to adapt to the complex network structures, increased network traffic, and diversified dynamic services, hindering the efficiency and promotion of 5G applications.
Innovation Solution
A method involving training hosts and inference hosts in a radio access network that interact to identify and utilize primary and secondary processing models for various wireless network activities, such as handover decisions and load balancing, through data inference and model training across different network nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional operation and maintenance management methods are used, then the network structure remains simple and manageable, but the system cannot adapt to the complex network structure, surge in network traffic and diversified dynamic services of 5G
Solution Approach 1:
The patent replaces traditional mechanical operation and maintenance management methods with an artificial intelligence-based system. The AI system uses machine learning models to automatically analyze network data, predict failures, and optimize performance, thereby adapting to complex 5G network structures and diversified services without requiring proportional increases in manual management complexity.
2Productivity
If the network scale is expanded to support more services and traffic, then the network capacity increases, but the operation and maintenance management becomes increasingly difficult
Solution Approach 1:
The patent implements a self-service AI system that autonomously performs network monitoring, fault detection, and optimization tasks. The system uses trained machine learning models to automatically analyze network data, identify issues, and execute corrective actions without human intervention, thereby maintaining ease of operation while supporting expanded network capacity and diverse services.
Solution Approach 2:
The patent introduces an AI system as an intermediary between the complex network infrastructure and human operators. This intermediary layer automatically processes vast amounts of network data, translates complex technical issues into actionable insights, and coordinates maintenance activities, thereby shielding operators from the increasing complexity while enabling high network capacity.
3Reliability
If more processing models are deployed to handle diverse network activities, then the system intelligence improves, but the model management and coordination complexity increases
Solution Approach 1:
The patent segments the AI system into multiple specialized processing models, each designed to handle specific network activities such as handover decisions, power saving, load balancing, and traffic steering. This segmentation allows each model to be optimized for its specific function, improving overall reliability while making management more systematic through modular organization and independent deployment of each specialized model.
Data Source
AI summary
Disclosed are methods and apparatus for artificial intelligence (AI) application in radio access network (RAN). One embodiment of the subject application provides a method performed by a training host (or a wireless network node including a training host) includes receiving a request associated with a wireless network activity, identifying a primary processing model for accomplishing the wireless network activity, and transmitting first information associated with the primary processing model to a second wireless network node to accomplish the wireless network activity, herein the first wireless network node is capable of identifying the primary processing model from a plurality of processing models to accomplish the wireless network activity.