ML Coordination via RRC Signaling in Wireless Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless communication systems face challenges in effectively coordinating machine learning capabilities between next-generation radio access networks (NG-RAN) and user equipment (UE) over the air interface, limiting network and UE performance optimization.
Innovation Solution
The implementation of new radio bearers and signaling mechanisms, such as the use of RRC signaling, new system information blocks, and dedicated data radio bearers, to facilitate the exchange of machine learning configurations, models, and reports between NG-RAN and UE, enabling distributed and federated learning for joint optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If machine learning capabilities are coordinated between NG-RAN and UE over the air interface, then network and UE performance optimization is enabled, but the complexity of coordination and collaboration increases
Solution Approach 1:
The patent segments the machine learning coordination into distinct functional components: ML capability indication, service registration, capability inquiry, configuration transmission, and model updates. Each component is handled through separate signaling messages and procedures, making the overall complex process manageable and implementable through modular mechanisms
Solution Approach 2:
The patent introduces standardized signaling mechanisms as intermediaries between NG-RAN and UE for ML coordination. Specific RRC signaling messages, system information blocks, and data radio bearers serve as intermediary channels that facilitate structured information exchange, reducing the complexity of direct peer-to-peer coordination
2Adaptability or versatility
If new radio bearers and signaling mechanisms are implemented for ML exchange, then distributed and federated learning is enabled, but the device and network complexity increases
Solution Approach 1:
The patent implements universal signaling mechanisms that serve multiple functions: RRC signaling messages handle both traditional radio resource control and ML-specific coordination; system information blocks convey both network configuration and ML capability information; data radio bearers transport both user data and ML models/parameters. This multi-functionality reduces the need for separate dedicated mechanisms
Solution Approach 2:
The patent establishes ML coordination capabilities in advance through capability indication and service registration procedures before actual ML operations begin. The network and UE exchange capability information, register services, and set up signaling channels proactively, so that when distributed or federated learning needs to be executed, the infrastructure is already in place and ready to operate
Data Source
AI summary
This disclosure describes systems, methods, and devices related to collaboration between user equipment (UE) and network for machine learning. A radio access network (RAN) node B device may transmit, to the CE device, an indication that the node B device supports machine learning; identify a service registration, received from the UE device, indicating that the UE device requests machine learning support from the node B device; transmit, to the UE device, a request for information associated with the UE device, the information associated with at least one of hardware capabilities or machine learning capabilities of the UE device; identify the information received from the UE based on the request for information; and transmit, to the UE device, a machine learning configuration for use by the UE device, wherein the machine learning configuration is based on the information.