AI/ML Capability Reporting in Wireless Terminals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless communication systems face challenges in effectively coordinating between wireless terminals and telecommunications networks to report real-time AI/ML model capabilities, leading to potential overconfiguration or underutilization of AI/ML models and functionalities due to unknown device capabilities.
Innovation Solution
The implementation of a message-based system where the network requests and receives AI/ML-related information from wireless terminals, allowing for periodic or triggered updates on run-time capabilities, enabling dynamic adaptation and optimization of AI/ML model support to meet performance KPIs, through UECapabilityInquiry messages and resource restriction responses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the network configures AI/ML models without knowing terminal capabilities, then model deployment can proceed, but overconfiguration or underutilization of AI/ML models occurs
Solution Approach 1:
The terminal performs preliminary capability self-determination before network configuration, identifying its AI/ML model capabilities in advance. This preliminary action allows the terminal to provide accurate capability information to the network, ensuring that subsequent model deployment is neither overconfigured nor underutilized.
Solution Approach 2:
The terminal provides feedback to the network about its AI/ML capabilities through capability information messages. This feedback mechanism enables the network to adjust its configuration strategies based on actual terminal capabilities, resolving the mismatch between configured models and terminal abilities.
2Measurement precision
If the network requests detailed AI/ML capability information from terminals, then configuration accuracy improves, but signaling overhead increases
Solution Approach 1:
The patent extracts only the essential AI/ML capability information that the network needs for configuration, rather than requiring terminals to report all possible details. This selective extraction of capability information reduces signaling overhead while maintaining sufficient accuracy for effective model deployment.
Solution Approach 2:
The terminal reports capability information at an appropriate level of detail - not all possible capabilities are reported, but sufficient information is provided to enable accurate configuration. This partial action approach balances reporting accuracy with signaling efficiency.
3Speed
If AI/ML models are activated without capability verification, then system responsiveness improves, but resource waste increases
Solution Approach 1:
The terminal performs preliminary capability self-determination before model activation, identifying which AI/ML models it can support. This preliminary identification enables rapid, informed activation decisions without subsequent capability verification delays, while avoiding activation of models the terminal cannot execute.
Solution Approach 2:
The terminal autonomously determines its own AI/ML capabilities and makes informed decisions about model activation based on its self-assessed capabilities. This self-service approach eliminates the need for network-orchestrated capability verification, improving activation speed while preventing resource waste through accurate self-knowledge.
Data Source
AI summary
A network includes one or more nodes which comprises processor circuitry and interface circuitry. The processor circuitry is configured to generate at least one message which requests the wireless terminal to report Artificial Intelligence/Machine Learning Model (AI/ML) related information to the network. The interface circuitry configured to transmit the at least one message to the wireless terminal.