WIRELESS-BASED LOCALIZATION USING A ZONAL FRAMEWORK

• US 20150237471A1
• Filed: 02/20/2014
• Published: 08/20/2015
• Est. Priority Date: 02/20/2014
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Li, Liqun, Shen, Guobin, Zhao, Chunshui, Zhao, Feng
• Granted Patent: US 9,674,656 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G01S 5/02521 using a radio-map
• CPC Class Codes: H04W 4/021 Services related to particular areas, e.g. point of interest [POI] services, venue services or geofences
• CPC Class Codes: H04W 4/029 Location-based management or tracking services
• 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 84/12 WLAN [Wireless Local Area Networks]

Abstract

Disclosed herein are techniques and systems for performing wireless-based localization using a zonal framework. An area (i.e., surface or space) may be partitioned into multiple zones, and one or more signal propagation models for one or more wireless access points (APs) may be generated for each zone. The result is a set of zonal signal propagation models that allow for improved model fitness on a per-zone basis. A process includes receiving a location query associated with a wireless communication device, selecting a target zone among multiple available zones of an area, and estimating a location of the wireless communication device based at least in part on one of a signal propagation model associated with the target zone or a fingerprint-based localization. The signal propagation model associated with the target zone may be generated based on training samples observed exclusively within the target zone.

Claims

1. A computer-implemented method comprising:

receiving a location query associated with a wireless communication device;

selecting, by one or more processors, a target zone among multiple available zones of an area; and

estimating, by the one or more processors, a location of the wireless communication device based at least in part on one of a signal propagation model associated with the target zone or a fingerprint-based localization.

2. The computer-implemented method of claim 1, wherein the signal propagation model associated with the target zone is generated based on training samples observed exclusively within the target zone.

3. The computer-implemented method of claim 2, wherein a number of the training samples observed is to meet or exceed a threshold number of training samples before the signal propagation model associated with the target zone is generated.

4. The computer-implemented method of claim 3, wherein the threshold number is five.

5. The computer-implemented method of claim 1, wherein the selecting the target zone comprises:

receiving a wireless communication sample from the wireless communication device, the WiFi sample reporting a received signal strength (RSS) from one or more wireless access points at the wireless communication device;

using histograms of the one or more wireless access points to calculate a likelihood of observing the wireless communication sample in individual ones of the multiple available zones; and

selecting the target zone as a zone among the multiple available zones that maximizes the likelihood of observing the wireless communication sample.

6. The computer-implemented method of claim 5, further comprising:

filtering out zones of the multiple available zones based on a comparison of wireless access points in the area to the one or more wireless access points reported in the wireless communication sample to obtain a set of candidate zones with at least one common wireless access point to the one or more wireless access points; and

using specific histograms of wireless access points in the set of candidate zones to calculate the likelihood of observing the wireless communication sample in individual ones of the candidate zones.

7. The computer-implemented method of claim 1, wherein the selecting the target zone comprises:

obtaining an estimated location of the wireless communication device using a global signal propagation model associated with the area; and

selecting the target zone as a zone among the multiple available zones with a center that is closest to the estimated location.

8. The computer-implemented method of claim 1, wherein the selecting the target zone comprises:

obtaining an estimated location of the wireless communication device using a global signal propagation model associated with the area;

selecting a subset of closest zones among the multiple available zones based on distance to the estimated location;

estimating the location of the wireless communication device according to respective signal propagation models associated with each of the closest zones; and

selecting the target zone as a zone among the subset of the closest zones that minimizes a model fitness error for the respective signal propagation models.

9. The computer-implemented method of claim 1, wherein the target zone is associated with a floor among multiple floors of the area that makes up an indoor space, and the estimating the location of the wireless communication device includes determining the floor among the multiple floors.

10. A system comprising:

one or more processors;

one or more memories that store a plurality of computer-executable components that are executable by the one or more processors, the components comprising;

a zonal localization component to receive a location query associated with a wireless communication device; and

a zone selection component to select a target zone from multiple available zones of an area; and

the zonal localization component to estimate a location of the wireless communication device based at least in part on one of a signal propagation model associated with the target zone or a fingerprint-based localization.

11. The system of claim 10, wherein the zone selection component is to:

receive a wireless communication sample from the wireless communication device, the wireless communication sample reporting a received signal strength (RSS) from one or more wireless access points at the wireless communication device; and

use histograms of the one or more wireless access points to calculate a likelihood of observing the wireless communication sample within individual ones of the multiple available zones; and

select the target zone as a zone among the multiple available zones that maximizes the likelihood of observing the wireless communication sample.

12. The system of claim 11, wherein the zonal localization component is further to adjust for device gain diversity by:

shifting the RSS reported in the wireless communication sample by multiple gain offsets across a range of possible gain offsets;

selecting a zone of the multiple available zones that maximizes a likelihood of observing the shifted RSS; and

selecting a gain offset among the multiple gain offsets corresponding to a maximum among all likelihoods determined for each shifted RSS.

13. The system of claim 10, further comprising a global model localization component to estimate the location of the wireless communication device based on a global signal propagation model generated for individual ones of multiple wireless access points within the area.

14. The system of claim 13, wherein the zonal localization component is further to:

apply the global signal propagation model to determine an estimated location of the wireless communication device;

compare respective model fitness errors of the signal propagation model associated with the target zone and the global signal propagation; and

resolve a final location estimation using a signal propagation model associated with a minimum model fitness error of the respective model fitness errors.

15. The system of claim 10, wherein the zone selection component is to select at least one other target zone as part of a set of candidate zones, and the zonal localization component is further to:

determine respective signal propagation models associated with each of the candidate zones;

apply the respective signal propagation models to estimate respective locations of the wireless communication device for each of the candidate zones; and

resolve a final location estimation as an average of the respective locations.

16. The system of claim 10, wherein the signal propagation model associated with the target zone is generated based on training samples observed exclusively within the target zone.

17. A computer-implemented method comprising:

partitioning, by one or more processors, an area into multiple zones;

generating, by the one or more processors, one or more signal propagation models for individual ones of one or more wireless access points in each of the multiple zones;

receiving a location query associated with a wireless communication device;

selecting, by the one or more processors, a target zone among the multiple zones of the area; and

estimating, by the one or more processors, a location of the wireless communication device based at least in part on one of a signal propagation model associated with the target zone or a fingerprint-based localization.

18. The computer-implemented method of claim 17, wherein the generating the one or more signal propagation models for individual ones of the one or more wireless access points in each zone includes:

collecting multiple training samples reporting received signal strength (RSS) at different locations within the zone; and

obtaining a set of parameters of a path loss model for individual ones of the one or more wireless access points in the zone based on the collected training samples.

19. The computer-implemented method of claim 17, wherein the partitioning the area into the multiple zones includes choosing a level of a map tile representation of the area, and designating each tile of the map tile representation at a chosen level as separate zones of the multiple zones.

20. The computer-implemented method of claim 17, wherein the partitioning the area into the multiple zones includes:

determining sub-areas within the area that contain at least a threshold number of observed training samples; and

designating each sub-area as a separate zone of the multiple zones.