Terminal Data Collection for AI Model Drift Retraining

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing AI models in mobile communication systems degrade over time due to environmental changes, necessitating retraining, but securing high-quality retraining data is challenging, especially when data causing performance degradation is included.

Innovation Solution

A method and apparatus for collecting and managing training data by transmitting AI/ML information and data storage capabilities between terminals and base stations, allowing for data storage configuration, collection, and retraining based on concept drift detection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If AI model retraining is performed using collected data, then the AI model can be updated with new information, but data causing performance degradation may be included in the training data

Engineering Contradiction:

ImproveAI model adaptability to new environmentsVSAvoidAI model performance stability

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The system performs preliminary actions by collecting and storing data in advance during normal operation, tagging it with metadata indicating normal operation status. When retraining is needed, this pre-collected clean data is available immediately, avoiding the need to collect data during degraded states.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements feedback mechanisms by monitoring AI model performance in real-time and detecting degradation. This feedback triggers selective data collection only for normal operation states, ensuring that training data does not include degraded performance data that would worsen model performance.

Inventive Principle:

Principle #23Feedback

2Quantity of substance

If data is collected continuously for AI model retraining, then sufficient training data can be secured, but data storage requirements increase significantly

Engineering Contradiction:

ImproveVolume of training dataVSAvoidData storage system complexity

Core Design Contradiction:

Quantity of substanceVSDevice complexity

Solution Approach 1:

Instead of uniformly collecting all data, the system applies local quality by selectively collecting only data from normal operation states. The data collection is localized to specific quality conditions (normal vs. degraded), reducing overall data volume while maintaining training effectiveness.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system changes the parameter of data selection criteria based on operational state. By monitoring performance parameters and only collecting data when parameters indicate normal operation, the system reduces data volume while ensuring high-quality training data.

Inventive Principle:

Principle #35Parameter changes

3Use of energy by moving object

If data collection is performed during connection release or idle states, then power consumption is reduced, but data collection timing becomes more restricted

Engineering Contradiction:

ImproveTerminal power consumptionVSAvoidData collection timing flexibility

Core Design Contradiction:

Use of energy by moving objectVSLoss of time

Solution Approach 1:

The system dynamically adapts data collection timing based on operational state. During active connection, data collection is minimized; during connection release or idle states, data collection is activated. This dynamic approach reduces power consumption while ensuring data is collected at appropriate moments.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS20260012778A1Method and apparatus for collecting learning data of intelligence models

Publication Date: 2026.01.08 ELECTRONICS & TELECOMM RES INST

AI summary

A method of a terminal may comprise: transmitting, to a base station, at least one of Artificial Intelligence/Machine Learning (AI/ML) information of the terminal or data storage information which is a data storage capability of the terminal; receiving data storage configuration information generated by the base station based on the data storage information; storing collected data for event(s) included in the data storage configuration information; and transmitting and receiving the collected data with the base station.