ML Positioning Performance Reporting for RF Fingerprint Accuracy

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing wireless communication systems, particularly in the context of 5G, face challenges in accurately determining the location of user equipment (UE) due to the complexity of signal propagation and interference, especially in environments with high-density deployments and the use of higher frequency bands, which affect the performance of positioning reference signals.

Innovation Solution

The implementation of a machine learning model for RF fingerprinting (RFFP)-based positioning, where a neural network is used to analyze radio frequency fingerprints to enhance positioning accuracy by monitoring and reporting the performance of downlink RFFP models, enabling more precise location determination of UEs.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If machine learning models are used for positioning, then positioning accuracy is improved, but model performance degradation over time occurs

Engineering Contradiction:
Improvepositioning accuracyVSAvoidmodel performance stability
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent implements a feedback mechanism where positioning accuracy metrics are continuously monitored and fed back to the machine learning model. This feedback loop enables the model to learn from actual performance data and adjust its parameters, thereby maintaining high positioning accuracy while adapting to changing environmental conditions and signal characteristics over time.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent performs preliminary actions by pre-training the machine learning model with extensive training data and by establishing baseline performance metrics before deployment. This preparation ensures the model starts with optimal performance and has built-in knowledge of various positioning scenarios, reducing the impact of performance degradation during operation.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If RF fingerprinting is used for positioning, then positioning accuracy is improved, but signal propagation complexity increases

Engineering Contradiction:
Improvepositioning accuracyVSAvoidsignal propagation analysis complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces complex mechanical signal propagation analysis with a machine learning-based approach. Instead of manually analyzing and interpreting complex RF signal characteristics and propagation paths, the system uses pre-trained neural network models that automatically extract meaningful features from RF fingerprints. This substitution transforms the complex analytical task into a more manageable pattern recognition problem, reducing the computational and analytical complexity required for positioning.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces machine learning models as intermediary components between the raw RF signals and the final positioning determination. These models act as mediators that process the complex signal data, extract relevant features, and transform them into positioning information. This intermediary layer simplifies the overall system by abstracting away the complexity of direct signal analysis while maintaining high positioning accuracy.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP4515883B1Machine learning model positioning performance monitoring and reporting
Publication Date: 2026.04.22 QUALCOMM INC
  • EP4515883B1 patent drawingFigure 1
  • EP4515883B1 patent drawingFigure 2A
  • EP4515883B1 patent drawingFigure 2B

AI summary

Disclosed are techniques for wireless communication. In an aspect, a network entity receives a provide location information message from a user equipment (UE), the provide location information message including one or more positioning estimates derived by the UE during one or more positioning inference occasions of a machine learning model, wherein the machine learning model is applied to one or more measurements of a wireless channel between the UE and a network node during each of the one or more positioning inference occasions, and transmits a performance report indicating a performance of the machine learning model at least in deriving the one or more positioning estimates during the one or more positioning inference occasions.