STGCN Model Corrects GNSS Positioning Errors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current GPS systems face limitations in accuracy and cost due to dependency on reference stations and satellite geometry errors, making them difficult and expensive to implement, especially for precise location determination in mobile devices.
Innovation Solution
The implementation of a spatio-temporal graph convolution network (STGCN) architecture for correcting GPS positions using deep learning techniques, which eliminates the need for reference stations and predicts satellite trajectories to improve accuracy without relying on traditional GPS correction processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional GPS correction systems (WAAS, DGPS) are used to improve accuracy, then positioning accuracy is improved, but system complexity and cost increase due to requirement for ground stations and reference infrastructure
Solution Approach 1:
The patent replaces the mechanical/physical infrastructure of ground stations and reference stations with a computational neural network model. The STGCN model processes satellite ephemeris and almanac data to predict and correct position errors, eliminating the need for physical correction infrastructure while maintaining high positioning accuracy.
Solution Approach 2:
The patent creates a virtual copy of the correction system through the neural network model. Instead of using physical reference stations to provide correction data, the STGCN model learns correction patterns from training data and replicates the correction function computationally, reducing infrastructure requirements.
2Measurement precision
If reference stations and ground infrastructure are deployed to correct GPS errors, then positioning accuracy is improved, but implementation cost and difficulty increase
Solution Approach 1:
The patent substitutes physical ground-based correction infrastructure with a software-based neural network model. The STGCN architecture processes satellite data and outputs corrected positions without requiring deployment of reference stations or ground infrastructure, significantly simplifying implementation.
Solution Approach 2:
The neural network model performs self-correction of GPS errors using patterns learned from training data. The STGCN model autonomously processes satellite ephemeris and almanac information to predict and compensate for positioning errors, eliminating the need for external reference stations or manual correction processes.
3Measurement precision
If satellite geometry errors are accounted for using traditional correction methods, then positioning accuracy is improved, but system complexity increases due to dependency on multiple reference stations
Solution Approach 1:
The patent replaces complex multi-station reference systems with a single neural network model that processes satellite data. The STGCN model takes satellite ephemeris and almanac information as input and outputs corrected positions, simplifying the system architecture while maintaining accuracy.
Solution Approach 2:
The neural network model serves multiple functions: it processes satellite ephemeris data, corrects positioning errors, and predicts satellite trajectories. The STGCN architecture integrates these functions into a single unified model, reducing system complexity compared to separate correction systems.
Data Source
AI summary
Embodiments including a method and apparatus for correction of a global navigation satellite system (GNSS) are described. In one example, the apparatus includes a communication interface and a processor. The communication interface is configured to a plurality of GNSS signals. The GNSS signals may include at least one almanac value and at least one ephemeris value. The processor is configured to generate a spatio-temporal graph model based on the at least one almanac value, the at least one ephemeris value, and a predetermined offset value for a base location. The spatio-temporal graph model analyzes subsequent GNSS signals to determined a predicted offset or a corrected GNSS position.