Terminal AI/ML Capability Signaling for Adaptive Model Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing communication systems lack effective methods for identifying and managing AI/ML-based functionalities and models in terminals, which hinders the optimization and enhancement of communication quality in 5G and beyond.
Innovation Solution
A method and apparatus for intelligent learning management in a communication system, where terminals and base stations exchange information on supported functionalities and candidate models, enabling model training and reporting of training results, allowing for dynamic adaptation and optimization of AI/ML models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If AI/ML functionalities are integrated into communication systems, then communication quality and performance are enhanced, but device complexity and management difficulty increase
Solution Approach 1:
The patent implements feedback mechanisms where terminals report supported functionalities and training results to base stations, and base stations provide capability inquiry messages and model information. This closed-loop feedback system enables automatic adaptation and selection of appropriate AI/ML models without manual configuration, resolving the management complexity issue while maintaining enhanced communication quality
Solution Approach 2:
The system enables terminals to autonomously determine which candidate models to apply based on their supported functionalities and available resources. Terminals self-manage the AI/ML model selection and training processes without requiring complex external management, thereby enhancing communication quality while reducing management overhead
2Adaptability or versatility
If multiple candidate models are supported for different functionalities, then adaptability and optimization capability improve, but information management and model selection complexity increase
Solution Approach 1:
The patent segments the AI/ML model management into distinct components: capability inquiry messages for functionality identification, separate candidate model information transmission for each supported functionality, and targeted application functionality reporting. This segmentation organizes the complex information exchange, enabling high model adaptability while managing information efficiently without overwhelming overhead
3Productivity
If terminals autonomously determine and apply AI/ML models, then system responsiveness and operational efficiency improve, but processing requirements and energy consumption increase
Solution Approach 1:
The patent implements partial action by having terminals autonomously determine and apply only the specific AI/ML models that are relevant to their supported functionalities and current operational needs. Rather than processing or maintaining all possible models, terminals perform partial model selection and application, thereby improving operational efficiency while limiting energy consumption to only what is necessary for the selected models
Data Source
AI summary
A method of a terminal may comprise: receiving a capability inquiry message from a base station; transmitting, to the base station, information on at least one support functionality supported by the terminal; receiving, from the base station, information on at least one candidate model related to the at least one support functionality; and transmitting information on an application functionality to be applied by the terminal to the base station based on the information on the at least one candidate model.