Crowdsourced Wi-Fi FTM Ranging Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing Wi-Fi Fine Timing Measurement (FTM) protocols for indoor positioning face challenges such as erroneous range estimation due to device offset and non-line-of-sight (NLOS) reception, requiring pre-calibration and being sensitive to environmental disturbances.
Innovation Solution
A ranging-type positioning system and method based on crowdsourced calibration, which involves arranging base stations in a target field, using a mobile device to collect measurement data, and employing a computing device to calculate measurement distances, execute a particle filter, and optimize a ranging model with offset compensation and NLOS estimation modules to minimize distance and geometric losses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Wi-Fi FTM protocol is used for indoor positioning, then time resolution can reach a few nanoseconds and sub-meter level ranging accuracy can be achieved, but erroneous range estimation occurs due to device offset and non-line-of-sight (NLOS) reception
Solution Approach 1:
The patent introduces an intermediary calibration process using mobile devices as intermediate measurement points. Mobile devices collect FTM measurement data from multiple base stations and serve as mediators to calibrate offset errors and identify NLOS conditions, thereby improving the reliability of range estimation without sacrificing the high time resolution capability of the FTM protocol
Solution Approach 2:
The patent implements a feedback mechanism where measurement data from mobile devices is used to update and refine the calibration parameters of base stations. The system continuously collects FTM measurement records, processes them to detect NLOS conditions and offset errors, and feeds back corrected calibration data to improve subsequent ranging accuracy, creating a closed-loop system that enhances reliability while maintaining precision
2Ease of operation
If all FTM base stations are pre-calibrated and anchored at stationary locations, then positioning can be performed, but the system is easily affected by environmental disturbances
Solution Approach 1:
The patent transforms the static calibration approach into a dynamic system. Instead of fixed pre-calibrated base stations, the system uses mobile devices to continuously collect measurement data and update calibration parameters in real-time. This dynamic calibration process allows the system to adapt to environmental changes and disturbances, maintaining positioning capability while improving reliability through continuous adaptation
3Loss of information
If a particle filter is executed to reconstruct mobile device path and collect FTM data records, then calibration data can be obtained, but computational complexity increases
Solution Approach 1:
The patent segments the calibration process into distinct phases: mobile device movement and data collection, path reconstruction using particle filter, and offline calibration computation. By separating the computationally intensive particle filter execution from real-time positioning operations and performing calibration offline, the system maintains path reconstruction accuracy while reducing the computational burden on mobile devices during actual positioning
Data Source
AI summary
A ranging-type positioning system and a ranging-type positioning method based on crowdsourced calibration are provided. In a crowdsourcing stage, pedestrian dead reckoning (PDR) is performed based on readings of inertial measurement units on a mobile device, a particle filter (PF) is executed to reconstruct a path of the mobile device with map information of the target field, and FTM data records are collected. Then, a ranging model based on a neural network can be used to calibrate and inversely infer approximate locations of unknown base stations. The optimized ranging model can estimate estimated distances and standard deviations based on the FTM data records obtained in the crowdsourcing stage. In a positioning stage, a position of a to-be-positioned mobile device can be positioned by having the ranging model operated in cooperation with the PDR and the PF.