Wireless Terminal Zone Estimation Using Pattern Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining whether a wireless terminal is inside a geographic zone are either costly due to the need for GPS systems or require extensive processing and communication overhead, especially when GPS is not viable, such as indoors or in large cities.
Innovation Solution
A pattern classifier, like a neural network or support vector machine, is trained using signal traits to predict whether a wireless terminal is within a geographic zone, with the trained program executed locally on the terminal to estimate its location without the need for extensive infrastructure or GPS.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS-based location system is used, then location accuracy is improved, but device cost and complexity increase
Solution Approach 1:
The patent creates a simplified copy of the GPS location determination function by training a pattern classifier on signal trait data collected at known locations. The trained classifier model is then deployed to the wireless terminal, enabling location estimation without requiring actual GPS hardware. This copying approach achieves acceptable location accuracy while eliminating the need for expensive GPS receivers in the terminals.
2Measurement precision
If extensive processing and storage hardware is deployed in network infrastructure, then location estimation capability is improved, but system cost and complexity increase
Solution Approach 1:
The patent extracts the essential location estimation functionality from the complex network infrastructure and concentrates it in a training phase. The trained pattern classifier is then deployed to lightweight terminals, removing the need for extensive processing and storage hardware in the network. This extraction approach maintains location estimation capability while dramatically reducing system complexity and cost.
Solution Approach 2:
The patent performs all necessary processing and storage operations during the offline training phase, where the pattern classifier learns from collected signal trait data. Once trained, the classifier requires minimal processing power during actual location estimation at the terminal. This preliminary action approach separates the computationally intensive training phase from the lightweight execution phase, reducing ongoing system complexity.
3Measurement precision
If traditional location systems are deployed, then location determination capability is improved, but communication overhead increases
Solution Approach 1:
The patent copies the location determination logic into the terminal device itself in the form of a trained pattern classifier. This enables the terminal to autonomously estimate its location by processing local signal measurements without needing to communicate with external location servers. The copying approach eliminates continuous communication overhead while maintaining location determination capability.
Data Source
AI summary
A method and apparatus are disclosed for estimating whether or not a wireless terminal is in a geographic zone. The illustrative embodiment employs a pattern classifier that is trained on traits of electromagnetic signals at various locations. A computer-executable program is then generated based on the trained pattern classifier, and the program is installed and executed on a subscribed identity module of the terminal.