Calibration of a device location measurement system that utilizes wireless signal strengths

• US 20060119516A1
• Filed: 01/30/2006
• Published: 06/08/2006
• Est. Priority Date: 04/25/2003
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Platt, John C., Krumm, John C.
• Granted Patent: US 7,233,286 B2
• Time in Patent Office: Days
• US Class Current: 342/451
• CPC Class Codes: G01C 21/206 specially adapted for indoor navigation
• CPC Class Codes: G01S 5/02528 Simulating radio frequency fingerprints
• CPC Class Codes: H04W 16/20 for indoor coverage or short range network deployment
• CPC Class Codes: H04W 16/225 for indoor or short range network
• CPC Class Codes: H04W 64/00 Locating users or terminals or network equipment for network management purposes, e.g. mobility management
• CPC Class Codes: H04W 88/18 Service support devices; Network management devices

Abstract

An architecture for minimizing calibration effort in an IEEE 802.11 device location measurement system. The calibration technique is based upon a regression function that produces adequately accurate location information as a function of signal strength regardless of gaps in the calibration data or minimally available data. The algorithm takes a set of signal strengths from known room locations in a building and generates a function giving (x,y) as a function of signal strength, which function may then be used for the estimation of new locations. Radial basis functions, which are simple to express and compute, are used for regression. The fact that the algorithm maps signal strength to continuous location makes it possible to skip rooms during calibration, yet still evaluate the location in those rooms.

Claims

1. A method of calibrating a system used to determine new location(s) depending on observed signal properties, comprising:

measuring signal properties related to identified locations; and

generating a regression function based at least in part upon the signal properties, the function estimating new location(s) as a function of observed signal properties.

2. The method of claim 1, the regression function comprising a plurality of kernel functions, each kernel function computing the difference between a vector of observed signal properties and a vector of stored signal properties.

3. The method of claim 2, the kernel functions being isotropic Gaussian kernel functions.

4. The method of claim 2, the vectors of stored properties are computed by clustering a plurality of vectors of measured signal properties.

5. The method of claim 1, the regression function further comprises additional parameters that are computed from a centroid of the identified locations.

6. The method of claim 1, the signal properties are of wireless radio signals transmitted from wireless transceivers, and the signal properties comprise signal strengths.

7. A method of calibrating a wireless device location measurement system, comprising:

accessing at least one wireless transmitting device;

logging signal strength data for each of wireless transmitting device; and

generating a regression function based upon the signal strength data, the regression function utilized to estimate new location(s).

8. The method of claim 7, the signal strength data logged from a plurality of different locations within range of the transmitting device.

9. A location system that uses observed signal properties to determine new location(s), comprising:

a measuring component for measuring signal properties related to identified locations; and

a regression component for generating a regression function based at least in part upon the signal properties, the function estimating new location(s) based upon observed signal property data.

10. The system of claim 9, the regression function comprising a plurality of kernel functions, each kernel function computing the difference between a vector of observed signal properties and a vector of stored signal properties.

11. The system of claim 10, the kernel functions being isotropic Gaussian kernel functions.

12. The system of claim 10, wherein the vector of stored properties is computed by clustering a plurality of vectors of measured signal properties.

13. The system of claim 9, the regression function comprising weights, and the regression component comprising a fitting component for fitting the regression function to the measured signal properties by adjusting the weights.

14. The system of claim 13 the fitting component comprising a least-squares fitting component.

15. The system of claim 13, the fitting component comprising a component for computing a kernel matrix.

16. The system of claim 13, the fitting component receives as input an estimate of the location uncertainty associated with each identified location.

17. The system of claim 13, the weights comprising a plurality of weights for computing an x location and a plurality of weights for computing a y location.

18. The system of claim 13, the regression function further comprising additional parameters that are computed from the centroid of the identified locations.

19. The system of claim 9, the signals are wireless signals transmitted from wireless transceivers, and the signal properties comprise signal strengths.

20. A computer readable medium having stored thereon the computer executable components of the system of claim 9.