ML Model Exchange via SI and RRC Signaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless communication systems lack techniques for user equipment (UE) and network entities to exchange information about which machine learning (ML) models are supported and applicable, leading to inefficiencies in model management and communication reliability.
Innovation Solution
The method involves a UE and network entity exchanging information about ML-based models through system information (SI) and radio resource control (RRC) messages, using model identifiers and interface IDs, allowing for compatible model selection and activation, enabling flexible model updates and improved communication reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If UE and network entity exchange information about supported ML models through signaling messages, then communication reliability is improved, but signaling overhead increases
Solution Approach 1:
The network entity provides an indication of applicable ML models before the actual communication begins. This preliminary exchange of model capability information allows the UE to know in advance which models are available at the network entity, enabling efficient model selection without requiring continuous signaling during active communication.
Solution Approach 2:
The UE transmits first signaling indicating the second set of ML-based network-side models supported by the UE, creating a feedback loop where the UE informs the network entity of its capabilities. This feedback mechanism ensures both sides have accurate information about compatible models, improving reliability while managing signaling overhead through targeted information exchange.
2Manufacturing precision
If UE and network entity manually configure and manage ML model compatibility, then model selection precision is improved, but device complexity increases
Solution Approach 1:
The UE autonomously determines the second set of ML-based network-side models supported by the UE based on its own capabilities. This self-service approach allows the UE to independently manage model selection without requiring complex manual configuration or centralized control, reducing device complexity while maintaining precise model compatibility.
Solution Approach 2:
The system uses model identifiers and interface IDs as parameters to represent ML models and their compatibility relationships. By transforming the complex model management problem into parameter comparison and matching, the system achieves precise model selection without requiring complex management mechanisms.
3Adaptability or versatility
If the network entity provides indications of applicable ML models, then adaptability is improved, but use of energy increases
Solution Approach 1:
The network entity provides an indication of the first set of ML-based network-side models applicable at the network entity in advance. This preliminary provision of model information enables the UE to prepare appropriate models locally without requiring continuous energy-intensive model evaluation or reconfiguration during active communication, thus improving adaptability while managing energy consumption.
Data Source
AI summary
Certain aspects of the present disclosure provide techniques for exchanging information between user equipments (UEs) and network entities regarding which models the UEs and network entities support. A method that may be performed by a UE includes: obtaining an indication of one or more machine learning (ML) based network-side models applicable at a network entity; and transmitting signaling indicating one or more of the ML-based network-side models supported by the UE.