Automated WLAN Radio Map Construction via FSA State Labeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for constructing WLAN radio maps in indoor spaces are costly and time-consuming, particularly due to the reliance on GPS signals which are not available indoors and require additional sensors, limiting their accuracy and availability.
Innovation Solution
An automated method and system using Finite State Automata (FSA) and machine learning to label WLAN fingerprints collected by smartphones without location information, dynamically determining the number and placement of Access Points (APs) based on signal propagation models and algorithms like A* and EM, allowing for automatic radio map construction without manual operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If GPS signals are used as reference locations for constructing radio map, then location accuracy can be maintained, but GPS signals are not available in indoor spaces
Solution Approach 1:
The patent introduces FSA-based state models as an intermediary between WLAN fingerprints and location information. Instead of directly using GPS signals which don't work indoors, the system creates a mediating representation (FSA states) that captures spatial relationships and transitions, allowing indoor location recognition without direct GPS dependency
Solution Approach 2:
The patent replaces the GPS-based mechanical positioning system with a WLAN fingerprint-based electromagnetic signal system. By substituting the physical GPS satellite signal mechanism with WLAN radio frequency signals and FSA-based probabilistic modeling, the system achieves indoor positioning capability where GPS is unavailable
2Measurement precision
If manual construction of radio map is performed, then accurate location data can be collected, but much cost and time are required
Solution Approach 1:
The system enables self-service by allowing mobile devices to automatically contribute WLAN fingerprints during normal operation. Users inadvertently participate in radio map construction by allowing their devices to collect and upload WLAN signal data, eliminating the need for dedicated manual data collection teams while maintaining data quality through automated FSA-based processing
Solution Approach 2:
The patent applies preliminary action by pre-processing and organizing WLAN fingerprints into FSA state models before they are needed for location recognition. The system proactively builds and updates the radio map structure in advance, so that when location queries occur, the processing is already optimized and ready, reducing both construction and query time
3Loss of information
If additional sensors are used to tag collection points, then location information can be obtained, but additional costs are incurred
Solution Approach 1:
The patent extracts location information purely from WLAN fingerprint patterns without requiring additional sensors. By taking out the dependency on extra hardware and focusing solely on analyzing WLAN signal characteristics (strength, MAC addresses, signal quality), the system obtains location data using only the standard wireless network interface already present in mobile devices
Solution Approach 2:
The system makes the WLAN interface universal by using it for multiple purposes: both for normal wireless communication and for location determination. The same wireless card that provides network connectivity also captures fingerprints for positioning, eliminating the need for separate sensors and reducing device complexity while maintaining information quality
Data Source
AI summary
An automated WLAN radio map construction method and system is provided. The automated WLAN radio map construction method includes: collecting WLAN fingerprints obtained by mobile device in an indoor space, machine-learning a learning model which is generated based on a state diagram in which divided areas of an indoor map are expressed by location states, arranging the collected WLAN fingerprints in corresponding location states, and storing a result of the arranging. Accordingly, collection locations of WLAN fingerprints collected in a plurality of unspecific smartphones without reference location information such as GPS signals can be automatically labeled.