Mobile Device Localization via Signal Difference Variance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fingerprint-based localization systems face challenges due to device-dependent antenna gains, which affect signal measurements and complicate the estimation of a mobile device's location, as different devices have varying reception characteristics.
Innovation Solution
A method that estimates location by calculating the variance in signal differences between current measurements and fingerprinted measurements across multiple beacons, which factors out the antenna gain of the mobile device, using statistical spread measures such as variance or standard deviation, reducing complexity and susceptibility to device-specific variations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional fingerprint-based localization methods are used, then location estimation can be performed, but device-dependent antenna gains cause measurement variations that reduce accuracy
Solution Approach 1:
The patent extracts and removes the antenna gain factor from the localization calculation by using variance of signal differences. Instead of directly using signal strength measurements that are contaminated by device-specific antenna gains, the method calculates the variance of differences between current and fingerprinted measurements, which mathematically factors out the antenna gain component, leaving only location-dependent variations.
Solution Approach 2:
The patent changes the parameter used for localization from raw signal strength (RSSI) to the variance of signal differences. This parameter transformation converts a measurement that is highly sensitive to antenna gain variations into one that is invariant to such variations, thereby improving reliability across different device types while maintaining location estimation capability.
2Measurement precision
If comprehensive signal measurements are taken to account for device variations, then accuracy may improve, but computational complexity increases
Solution Approach 1:
The patent extracts only the essential location-dependent information by calculating the variance of signal differences, discarding the device-specific antenna gain component. This extraction approach simplifies the computational model from one that would require device-specific calibration and correction to a universal model that works across different devices with minimal computation.
Solution Approach 2:
Instead of trying to directly measure and correct for antenna gain effects on signal strength, the patent inverts the approach by using the variance of differences, which automatically eliminates the antenna gain factor. This inversion transforms a complex device-specific calibration problem into a simple universal calculation.
Data Source
AI summary
A method comprising: obtaining a respective measurement of a beacon signal transmitted wirelessly between the mobile device and each of a plurality of the reference nodes; for each of a plurality of different reference locations within the environment, distinct from the reference nodes, obtaining a corresponding measurement of a test signal previously transmitted or modelled to have been transmitted wirelessly between the reference location and each of the plurality of reference nodes; for each of the reference locations, determining for each of said plurality of beacon nodes a difference between the respective measurement of the beacon signal and the measurement of the test signal; for each of the reference locations, determining a value of a measure of statistical spread in these differences; and estimating the location of the mobile device in dependence on a comparison of the values of the measure of statistical spread determined for the different reference locations.