Network Function Analytics Using Multi-Model Accuracy Feedback

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The reliability of data analytics results in network functions such as NWDAF is poor due to dependence on a single model, leading to inaccurate predictions.

Innovation Solution

Perform inference on analytics using at least two models, determine accuracy information, and provide feedback to network functions regarding accuracy status and recovery time to improve reliability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If data analytics is performed based on a single model, then the complexity of the system is reduced, but the reliability of the data analytics result deteriorates

Engineering Contradiction:
Improvesystem complexityVSAvoidreliability of data analytics result
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent combines multiple prediction models to perform data analytics, merging their results through weighted aggregation. This approach integrates diverse model perspectives to improve reliability while managing system complexity through structured combination methods.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent creates a composite prediction system by combining multiple models with different characteristics (e.g., different accuracy levels, different computational costs). Each model contributes its strengths to form a more reliable composite prediction system, analogous to composite materials combining different properties.

Inventive Principle:
Principle #40Composite materials

2Reliability

If multiple models are used for inference, then the reliability of the data analytics result is improved, but the device complexity increases

Engineering Contradiction:
Improvereliability of data analytics resultVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the prediction system into multiple independent models, each handling specific aspects of data analytics. This segmentation allows parallel processing and independent optimization of each model, managing complexity through modular architecture while improving overall reliability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements feedback mechanisms where prediction results from multiple models are aggregated and evaluated, with weights adjusted based on performance feedback. This feedback loop enables the system to adapt and optimize model combinations, managing complexity through intelligent control.

Inventive Principle:
Principle #23Feedback

3Reliability

If accuracy monitoring and feedback mechanisms are implemented, then the reliability of data analytics results is improved, but the loss of time increases

Engineering Contradiction:
Improvereliability of data analytics resultVSAvoidtime for accuracy monitoring
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent performs preliminary actions by pre-calculating model weights and establishing prediction frameworks before actual data analytics. This preparation reduces the time required for accuracy monitoring during operation, as the structural foundation is already in place.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements periodic accuracy monitoring instead of continuous monitoring, checking model performance at intervals. This periodic approach maintains reliability through regular validation while significantly reducing the time overhead compared to continuous monitoring.

Inventive Principle:
Principle #19Periodic action

Data Source

PatentEP4694289A1Information processing method and apparatus, and network element
Publication Date: 2026.02.11 VIVO MOBILE COMM CO LTD
  • EP4694289A1 patent drawingFigure 1~2
  • EP4694289A1 patent drawingFigure 3~5
  • EP4694289A1 patent drawingFigure 6

AI summary

This application discloses an information processing method and apparatus, and a network function, and belongs to the field of communication technologies. The method includes: performing, by a first network function, inference on analytics based on at least two models to obtain a first inference result; determining, by the first network function, first accuracy information; and sending, by the first network function, first information to a second network function based on the first accuracy information, where the first information includes at least one of the following: first indication information, where the first indication information is used to indicate that accuracy for an inference result corresponding to the analytics or the first inference result decreases or does not meet an accuracy requirement; the first accuracy information; a metric type corresponding to the first accuracy information; a problem type corresponding to the first accuracy information; second indication information, where the second indication information is used to indicate a recommended operation for the second network function; and third indication information, where the third indication information is used to indicate a time at which the accuracy for the inference result corresponding to the analytics or the first inference result recovers to meet a preset accuracy requirement.