AI/ML Positioning Data Quality Monitoring Using SLA and MIE

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing AI/ML-based positioning systems lack a quantitative method to assess and ensure the quality of data collection, particularly for UE positioning, and do not provide a fair monitoring metric for ground truth-based performance monitoring.

Innovation Solution

A solution that quantifies data collection quality using a spatial label analysis (SLA) indicator and minimum inference error (MIE) indicator, determining a monitoring quality criteria to apply actions on the AI/ML model if the criteria are not satisfied, ensuring accurate and realistic performance monitoring.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If AI/ML models are deployed for positioning without quantitative data quality assessment, then positioning functionality is provided, but data collection quality cannot be ensured and monitoring metrics are unfair

Engineering Contradiction:
Improvepositioning accuracyVSAvoiddata quality assessment
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The patent introduces quantitative parameters (SLA indicator and MIE indicator) to assess data quality. The SLA indicator measures spatial distribution uniformity by comparing actual label distribution against expected uniform distribution, while the MIE indicator quantifies positioning error thresholds. These parameter transformations enable objective data quality assessment that was previously unavailable, directly resolving the contradiction between providing positioning functionality and ensuring data quality.

Inventive Principle:
Principle #35Parameter changes

2Ease of operation

If monitoring metrics are established without spatial label analysis, then performance monitoring is simplified, but the monitoring is not realistic or fair

Engineering Contradiction:
Improvemonitoring simplicityVSAvoidmonitoring accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent implements a feedback mechanism where the SLA indicator continuously monitors whether label distribution meets uniformity thresholds. When the SLA indicator falls below the threshold, the system triggers data collection actions to improve spatial distribution. This feedback loop provides realistic and fair monitoring while maintaining operational simplicity through automated threshold-based decision making.

Inventive Principle:
Principle #23Feedback

3Productivity

If data collection proceeds without spatial distribution verification, then data collection speed is maintained, but spatial uniformity of positioning labels is not ensured

Engineering Contradiction:
Improvedata collection speedVSAvoidspatial distribution uniformity
Core Design Contradiction:
ProductivityVSStability of the object's composition

Solution Approach 1:

The patent performs preliminary verification of spatial distribution uniformity using the SLA indicator before data collection proceeds. By calculating the SLA indicator and comparing it against thresholds in advance, the system ensures spatial uniformity is achieved before deployment, preventing subsequent positioning accuracy issues while maintaining efficient data collection processes.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4603863A1Data collection quality indication in ai/ML positioning
Publication Date: 2025.08.20 NOKIA TECHNOLOGIES OY
  • EP4603863A1 patent drawingFigure 1
  • EP4603863A1 patent drawingFigure 2~3C
  • EP4603863A1 patent drawingFigure 4

AI summary

Example embodiments of the present disclosure are related to data collection quality indication in artificial intelligence/machine learning, AI/M, positioning. A first apparatus obtains a monitoring metric for an AI/ML model that is configured for direct positioning or assisted positioning of a terminal device within a communication network, the monitoring metric being determined based on an error between a positioning inference result of the AI/ML model for a model input and a ground-truth positioning label for the model input within a monitoring dataset; determines a monitoring quality criteria for the monitoring dataset based on a spatial label analysis, SLA, indicator and a minimum inference error, MIE, indicator of the monitoring dataset; and in accordance with a determination that a monitoring quality criteria is dissatisfied by the monitoring metric, determines an action is to be applied on the AI/ML model.