Hybrid Positioning System Using Machine Learning for Accuracy and Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current positioning systems for mobile devices, particularly in wireless communication systems, face challenges in achieving accurate location determination due to limitations in combining diverse positioning methods and latency issues in network-based fingerprinting techniques.
Innovation Solution
A hybrid positioning system utilizing machine-learning algorithms to integrate multiple positioning estimates from various methods, including RAT-dependent and RAT-independent techniques, and a user equipment (UE)-based fingerprinting system that determines positions using RF measurements, enhancing accuracy and reducing latency by calculating a weighted average of position estimates and averaging positions based on k-nearest fingerprints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If network-based fingerprinting techniques are used for positioning, then positioning can be performed using RF measurements, but latency issues occur and accuracy is limited
Solution Approach 1:
The positioning system is segmented into multiple independent positioning methods (RAT-dependent and RAT-independent techniques) that operate separately and provide individual position estimates, which are then combined through a machine learning model to achieve accurate and timely positioning
Solution Approach 2:
Multiple positioning estimates from different positioning methods are merged together using a machine learning model that calculates a weighted average, combining the strengths of various techniques to improve both accuracy and reduce latency compared to using any single method
2Measurement precision
If a single positioning method is used, then the system is simpler to implement, but positioning accuracy is insufficient
Solution Approach 1:
The positioning system achieves multi-functionality by integrating both RAT-dependent and RAT-independent positioning techniques within a single hybrid system, allowing it to adapt to different network conditions and environments while maintaining manageable complexity through a unified machine learning model
3Measurement precision
If diverse positioning methods are combined, then positioning accuracy improves, but the complexity of integrating multiple methods increases
Solution Approach 1:
A machine learning model serves as an intermediary that automatically integrates multiple positioning estimates from diverse positioning methods. The model calculates a weighted average of position estimates, managing the integration complexity internally while providing accurate positioning results without requiring complex manual integration logic
Data Source
AI summary
Disclosed are methods, systems, and computer-readable medium to perform operations including: receiving a plurality of position estimates for a user device; providing the plurality of position estimates as input to a trained machine learning model; and outputting a hybrid position of the user device.