Technical Description
This invention relates to a machine learning method for fingerprint-based user equipment (UE) positioning in wireless networks. By incorporating estimated label errors derived from positioning parameter variations alongside channel impulse response (CIR) data, the system mitigates the impact of non-line-of-sight multipath interference and data labeling inaccuracies.
Problems Addressed
- Lacks Indoor Viability
- Inefficient Conventional Systems
- Severe Multipath Interference
- Restricted Positioning Precision
- Vulnerable Labeled Data
Tech Features
- Enhanced Localization Accuracy
- Robust Error Estimation
- Intelligent Model Training
- Optimised Indoor Tracking
- Precise Position Mapping
Target Audience
- Wireless Communication Industries
- Telecommunication Infrastructure Sectors
- Autonomous Vehicle Industries
- Research & Development
Tech ID: P16-2117 TRL 4 Patent Status: Published Available For Exclusive and Non-exclusive License
