UE Capability Reporting for AI Model Activation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless communication systems face challenges in efficiently managing and activating artificial intelligence (AI) and machine learning (ML) models in telecommunications networks, leading to suboptimal resource utilization and increased errors in beam selection and resource allocation.
Innovation Solution
The implementation of a method where user equipment (UE) and network nodes use UE capability reporting and control signaling to identify and activate a subset of AI/ML model functionalities, enabling unified functionality-based life cycle management and model identification, thereby optimizing resource usage and reducing communication errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If AI/ML models are activated in wireless communication systems, then network performance and beam selection accuracy are improved, but computing resources and power consumption increase
Solution Approach 1:
The patent segments AI/ML model functionalities into distinct, selectable components that can be individually activated or deactivated. The network node and UE identify specific model functionalities from a set of available functionalities based on capability reporting, allowing partial activation rather than all-or-nothing deployment. This segmentation enables the system to activate only the necessary model components for improved beam selection while leaving other functionalities inactive, thereby maintaining accuracy improvements while reducing power consumption and computing resource usage.
2Productivity
If AI/ML model functionalities are deployed, then resource allocation efficiency is improved, but device complexity increases
Solution Approach 1:
The patent implements self-service mechanisms where the UE automatically generates capability reports indicating its supported AI/ML functionalities, and the network node automatically identifies and activates appropriate model functionalities based on these reports. The system performs self-configuration and self-management without requiring manual intervention or complex external orchestration. This automation reduces the operational complexity of deploying AI/ML models while maintaining improved resource allocation efficiency, as the system autonomously manages the complexity internally.
3Measurement precision
If comprehensive UE capability reporting is implemented, then model identification accuracy is improved, but communication overhead increases
Solution Approach 1:
The patent extracts and reports only the essential capability information needed for AI/ML model identification from the UE's full capability set. Rather than reporting all possible UE capabilities, the system specifically extracts and communicates the subset of functionalities relevant to AI/ML operations. This selective extraction maintains accurate model identification by providing the necessary information while reducing communication overhead by excluding unnecessary capability data from the reporting process.
Data Source
AI summary
Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a user equipment (UE) may transmit a UE capability reporting message including information associated with identifying a set of UE conditions associated with a first set of functionalities, wherein the first set of functionalities corresponds to a set of model features. The UE may receive, based at least in part on transmitting the UE capability reporting message, control signaling identifying a second set of functionalities that is a subset of the first set of functionalities. Numerous other aspects are described.