NWDAF Edge MOS Model Training Without AF Dependency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for establishing a Mean Opinion Score (MOS) model for services in live networks require synchronization and cooperation between 5GC NFs and AF entities, making it difficult to implement effectively.
Innovation Solution
A training method for an application MOS model is introduced, where a central NWDAF entity sends subscription requests to an edge NWDAF entity to obtain quality of service MOS levels and network performance indicators, allowing for the establishment of a MOS model without relying on the capabilities of AF and 5GC NFs entities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the NWDAF subscribes to QoS MOS level from AF entity and network performance indicator from 5GC NFs entity to establish MOS model, then the MOS model can be established with external data sources, but the implementation becomes difficult due to measurement synchronization requirements and dependency on AF and 5GC NFs entity capabilities
Solution Approach 1:
The patent extracts the measurement synchronization function from the complex interaction between NWDAF, AF, and 5GC NFs entities, and relocates it to the E-NWDAF entity. This allows the E-NWDAF to independently perform synchronous measurements of both QoS MOS levels and network performance indicators, eliminating the dependency on external entity capabilities and simplifying the implementation process.
Solution Approach 2:
The E-NWDAF entity acts as an intermediary between the C-NWDAF and the measurement functions. It receives subscription requests from C-NWDAF, performs the actual synchronous measurements locally, and returns the correlated data. This intermediary approach resolves the contradiction by making the system adaptable while ease of operation through centralized measurement capability.
2Adaptability or versatility
If the NWDAF relies on capability and cooperation of AF entity and 5GC NFs entity for measurement synchronization, then external data sources can be utilized, but the system complexity increases due to coordination requirements
Solution Approach 1:
The patent segments the NWDAF functionality into central (C-NWDAF) and edge (E-NWDAF) components. The E-NWDAF is specifically designed to handle measurement and synchronization tasks locally, while the C-NWDAF focuses on model establishment and analysis. This segmentation reduces system coordination complexity by localizing the measurement functions.
Solution Approach 2:
The E-NWDAF entity performs self-service by independently conducting synchronous measurements of QoS MOS levels and network performance indicators without requiring external coordination. It autonomously subscribes to necessary data, performs measurements, and returns results to C-NWDAF, thereby reducing system complexity while maintaining data source flexibility.
3Ease of operation
If synchronous measurement is performed within carrier network by E-NWDAF, then implementation is simplified and measurement accuracy is improved, but the E-NWDAF requires enhanced measurement capabilities
Solution Approach 1:
The patent introduces a new dimensional aspect to the NWDAF architecture by deploying E-NWDAF at the edge with enhanced measurement capabilities. This dimensional change allows synchronous measurements to be performed closer to the data sources, simplifying implementation while the additional capability requirements are localized to the edge component rather than the entire system.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
Embodiments of this application provide a training method for an application MOS model, a device, and a system, to resolve a problem that an existing training method for a MOS model of a service needs to rely on a capability and cooperation of an AF entity, and is difficult to implement in a live network. The method includes: A central network data analytics function C-NWDAF entity sends a first subscription request to an edge network data analytics function E-NWDAF entity, where the first subscription request is used to request to subscribe to a quality of service MOS level of a target service and a corresponding first network performance indicator, and the first network performance indicator is a network performance indicator of a transmission network that carries the target service; and the C-NWDAF entity receives the quality of service MOS level and the first network performance indicator from the E-NWDAF entity, and establishes a MOS model of the target service based on the quality of service MOS level and the first network performance indicator.