5G Time Synchronization Capability Exposure for Federated Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is a need for time synchronization among user equipment (UE) in a 5G network to support federated learning, and existing systems lack efficient methods for exposing capability information for time synchronization services.
Innovation Solution
A method is provided for a time synchronization service in a 5G network that includes receiving a subscription request for UE capability information and transmitting this information to support federated learning, involving components like the artificial intelligence/machine learning function (AIMLF), network exposure function (NEF), and UE capability information exposure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If existing systems are used without dedicated time synchronization service exposure, then system complexity is reduced, but time synchronization capability for federated learning cannot be effectively supported
Solution Approach 1:
The Network Exposure Function (NEF) acts as an intermediary component that bridges the gap between existing network infrastructure and federated learning applications. The NEF receives subscription requests from AIML AF, retrieves time synchronization capability information from AIMLF, and exposes this information to external applications without requiring changes to the core network architecture, thus enabling capability exposure while maintaining system simplicity.
Solution Approach 2:
The existing network functions (NEF, AIMLF) are extended to serve dual purposes: their traditional network management functions and the new time synchronization capability exposure function. This multi-functionality approach allows the system to support federated learning time synchronization without adding entirely new specialized components, thereby reducing overall system complexity while achieving adaptability.
2Measurement precision
If detailed UE capability information is exposed to support federated learning, then synchronization precision is improved, but information security risks increase
Solution Approach 1:
The NEF serves as a secure intermediary that controls and filters the exposure of UE capability information. It receives authorization through subscription requests, retrieves only necessary time synchronization capability data from AIMLF, and exposes it to authorized applications. This intermediary mechanism ensures that detailed capability information is not freely accessible, thereby maintaining security while providing sufficient precision for synchronization.
Solution Approach 2:
The system exposes only the specific time synchronization capability information that is locally relevant and necessary for federated learning operations, rather than exposing all UE capability information. The subscription request mechanism allows applications to subscribe to specific capability types, ensuring that only appropriate levels of detail are exposed, balancing precision requirements with security concerns.
Data Source
AI summary
Provided is a method of a time synchronization service for federated learning (FL) in a fifth generation (5G) network. The method includes receiving, by an artificial intelligence/machine learning function (AIMLF), a subscription request for UE capability information for an FL time synchronization service from an artificial intelligence/machine learning application function (AIML AF) and transmitting, by the AIMLF, the UE capability information to the AIML AF, in response to the subscription request.