NR Positioning Data Collection with SRS-Linked Coordinates

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In scenarios requiring high positioning accuracy, such as industrial settings, the acquisition of measurement signals and positioning coordinates for AI-based models is often out of sync due to being performed at different nodes, affecting training effectiveness and accuracy.

Innovation Solution

Ensure time synchronization of measurement signals and positioning coordinates by having terminals send SRS and acquire coordinates within a first time unit, allowing for synchronized data use in AI-based models.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If measurement signal acquisition and positioning coordinate acquisition are performed at different nodes, then device functionality is distributed, but time synchronization between measurement signals and positioning coordinates deteriorates

Engineering Contradiction:
Improvedistributed acquisition capabilityVSAvoidtime synchronization
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The terminal performs positioning coordinate acquisition within a first time unit after sending the SRS, ensuring that the coordinate acquisition is preliminarily timed to match the SRS transmission time. This preliminary timing action ensures that both the measurement signal and positioning coordinate are associated with the same time point, resolving the time synchronization issue while maintaining distributed acquisition architecture

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The terminal sends the acquired positioning coordinate to the network device, creating a feedback loop that allows the network device to receive both the measurement signal from the base station and the positioning coordinate from the terminal. This feedback mechanism enables the network device to perform AI-based positioning using time-synchronized data from distributed nodes

Inventive Principle:
Principle #23Feedback

2Ease of operation

If AI-based positioning model training uses time-asynchronized data, then data collection flexibility is improved, but positioning accuracy deteriorates

Engineering Contradiction:
Improvedata collection flexibilityVSAvoidpositioning accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The terminal acquires the positioning coordinate within a first time unit after sending the SRS, which is a preliminarily defined time window. This preliminary timing constraint ensures that the coordinate acquisition is synchronized with the SRS transmission time while still allowing flexible data collection methods, thus maintaining both ease of operation and positioning accuracy

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4636440A1Positioning data collection method and apparatus
Publication Date: 2025.10.22 BEIJING XIAOMI MOBILE SOFTWARE CO LTD
  • EP4636440A1 patent drawingFigure 1~2
  • EP4636440A1 patent drawingFigure 3~4
  • EP4636440A1 patent drawingFigure 5~6

AI summary

Disclosed in embodiments of the present disclosure are a positioning data collection method and apparatus, applicable to a new radio (NR) system. The method comprises: a terminal device sends a sounding reference signal (SRS) to a base station, the SRS being used for assisting the base station in collecting an SRS measurement signal; the terminal device collects positioning coordinates of the terminal device within a first time unit after sending the SRS; and the terminal device sends the positioning coordinates to a network side device. By implementing the embodiments of the present disclosure, the collected SRS measurement signal and the collected positioning coordinates can be ensured to be synchronous in time, and therefore, using the SRS measurement signal and the positioning coordinates which are synchronous in time as training data to train an artificial intelligence (AI) based positioning model can ensure the reliability of training data, thereby improving the model training effect.