Outdoor Localization via WiFi Manifold Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current outdoor localization methods face challenges in achieving accurate localization beyond fingerprinted zones and are inefficient due to high energy consumption and signal loss in urban areas, with existing solutions failing to provide satisfactory accuracy using WiFi infrastructure.
Innovation Solution
A framework that utilizes manifold learning on crowdsensed hotspot labels from outdoor WiFi networks to construct manifolds, allowing for improved localization accuracy by filtering outliers and synthesizing labeled and unlabeled data, and processing location queries efficiently using semi-supervised techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If WiFi fingerprinting methods are used for outdoor localization, then localization service can be provided without GPS, but localization accuracy is not satisfactory and does not work beyond fingerprinted zones
Solution Approach 1:
The patent divides the outdoor localization problem into two parts: using GPS for accurate positioning in open areas, and using WiFi fingerprinting as a supplementary method in urban areas where GPS signals are blocked. This segmentation allows each method to operate in its optimal environment, resolving the contradiction between service availability and accuracy.
Solution Approach 2:
The patent introduces an intermediary system that combines GPS and WiFi fingerprinting methods, using GPS as the primary localization method and WiFi as a fallback when GPS is unavailable. This intermediary approach maintains both service availability and acceptable accuracy by selectively applying the appropriate method based on environmental conditions.
2Measurement precision
If GPS sensors are used for outdoor localization, then high localization accuracy can be achieved, but energy consumption is high
Solution Approach 1:
The patent implements a dynamic localization system that automatically switches between GPS and WiFi fingerprinting methods based on signal availability and environmental conditions. The system dynamically selects the most energy-efficient method that still provides acceptable accuracy, using GPS only when necessary and relying on WiFi in GPS-denied urban areas.
Solution Approach 2:
The patent changes the operational parameters of the localization system by adjusting which method (GPS or WiFi) is active based on external conditions such as GPS signal strength and urban environment characteristics. This parameter change allows the system to optimize energy consumption while maintaining localization functionality.
3Reliability
If supplementary location indicators such as cellular and FM signals are used, then localization can be provided in areas without GPS, but localization accuracy is very low due to sparse deployment
Solution Approach 1:
The patent merges WiFi fingerprinting with traditional supplementary location indicators (cellular and FM signals) to create a hybrid localization system. By combining multiple signal sources and methods, the system achieves both service availability in GPS-denied areas and improved accuracy compared to using sparse cellular or FM signals alone.
Solution Approach 2:
The patent creates a composite localization system that integrates multiple different localization technologies (GPS, WiFi fingerprinting, cellular signals, FM signals) into a unified framework. This composite approach leverages the strengths of each individual method while compensating for their weaknesses, achieving both availability and acceptable accuracy.
Data Source
AI summary
Described herein is a framework for outdoor localization. In accordance with one aspect of the framework, a set of hotspot labels are received from one or more user devices connected to an outdoor wireless local area network. Manifold learning may be performed based on the set of hotspot labels to construct one or more manifolds. Using the one or more constructed manifolds, the framework may then estimate a location of a particular user device associated with a query record received from during an online location query.