SYSTEM AND METHOD FOR DETERMINATION OF PROXIMITY BETWEEN WIRELESS DEVICES

• US 20140357267A1
• Filed: 08/23/2013
• Published: 12/04/2014
• Est. Priority Date: 05/28/2013
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Hirsch, Dan F., Carranza, Marcos A., Oviedo, Marcos J.
• Granted Patent: US 9,313,722 B2
• Time in Patent Office: Days
• US Class Current: 455/434
• CPC Class Codes: H04W 48/16 Discovering, processing access restriction or access information

Abstract

Methods and systems to determine the proximity of wireless devices to each other. In an embodiment, the proximity may be determined as a membership in a classification, e.g., high, medium, or low proximity of one wireless device to a second. The process for determining proximity may include a training phase, where a set of wireless devices may perform scans to determine the identities and number of wireless access points, such as wireless routers. From this, statistical features may be extracted and used to perform training; the output of this training may be a set of trained models. The trained models may be used in an operational phase. Here, scan data from wireless devices may be submitted to a feature extraction process. The extracted features may then be used to determine the trained model most closely fitting the extracted features.

Claims

1. A system, comprising:

a feature extraction module, configured

to receive training input scan data from a first plurality of wireless devices, wherein the training input scan data comprises information regarding the accessibility of each of the first plurality of wireless devices to one or more wireless access points,

to extract features of the training input scan data,

to create a plurality of training feature vectors that describe the extracted features of the training input scan data; and

a trainer module configured to receive definitions of a plurality of proximity classifications and to produce a plurality of trained models on the basis of the training feature vectors and proximity classifications.

2. The system of claim 1, wherein the information regarding the accessibility of each of the first plurality of wireless devices to one or more wireless access points comprises:

for each wireless device of the first plurality of wireless devices, an identity of the wireless access points accessible to the wireless device;

for each wireless device of the first plurality of wireless devices, the number of the wireless access points accessible to the wireless device; and

for each wireless device of the first plurality of wireless devices, signal strengths of the accessible wireless access points.

3. The system of claim 1, wherein the extraction of features comprises one or more of:

determining a number of common access points, wherein the common access points are those access points accessible by two or more of the first plurality of wireless devices;

determining a number of access points that are not common to two or more of the first plurality of wireless devices;

determining a difference in a signal strength among a pair of wireless devices of the first plurality that detect the access point, the difference determined as a mean squared signal strength deviation for each common access point;

determining the percentage of the access point'"'"'s total signal strength detected by each wireless device of the pair of wireless devices in the first plurality of wireless devices,

determined for each access point;

determining the percentage of access points that are common access points; and

for each pair of wireless devices of the first plurality of wireless devices, determining whether each wireless device in the pair has the same strongest access point as determined by signal strength.

4. The system of claim 1, wherein the production of trained models by the trainer module comprises creation of Gaussian mixture models.

5. The system of claim 1, wherein the feature extraction module is further configured

to receive operational scan data from each of one or more of a second plurality of wireless devices, wherein the operational scan data comprises information regarding the accessibility of each of the one or more wireless devices in the second plurality of wireless devices to one or more of the wireless access points,

to extract features of the operational scan data, and

to create one or more operational feature vectors that describe the extracted features of the operational scan data;

and

the system further comprising;

a classifier module configured to

receive the operational feature vectors and the plurality of trained models; and

to produce a proximity classification for each pair of the one or more wireless devices of the second plurality of wireless devices, using the trained models.

6. The system of claim 5, wherein the information regarding the accessibility of each of the one or more wireless devices in the second plurality of wireless devices comprises:

for each of the one or more wireless devices of the second plurality of wireless devices, an identity of the wireless access points accessible to the wireless device;

for each of the one or more wireless devices of the second plurality of wireless devices, the number of the wireless access points accessible to the wireless device; and

for each of the one or more wireless access devices of the second plurality of wireless devices, signal strengths of the accessible wireless access points.

7. The system of claim 5, wherein the extraction of features of the operational scan data comprises one or more of:

determining a number of common access points, wherein the common access points are those access points accessible by two or more of the second plurality of wireless devices;

determining a number of access points that are not common to two or more of the second plurality of wireless devices;

determining a difference in a signal strength among a pair of wireless devices of the second plurality that detect the access point, the difference determined as a mean squared signal strength deviation over the common access points;

determining the percentage of the access point'"'"'s total signal strength detected by each wireless device of the pair of wireless devices in the second plurality of wireless devices, determined for each access point;

determining the percentage of access points that are common access points; and

for each pair of wireless devices of the second plurality of wireless devices, determining whether each wireless device in the pair has the same strongest access point as determined by signal strength.

8. The system of claim 5, wherein said production of a proximity classification for each pair of the one or more wireless devices of the second plurality of wireless devices comprises:

for each pair of the one or more wireless devices of the second plurality of wireless devices, determining the trained model that most closely models the operational feature vector; and

identifying the proximity classification associated with the determined trained model.

9. A method, comprising:

receiving, at a feature extraction module, training input scan data from a first plurality of wireless devices, wherein the training input scan data comprises information regarding the accessibility of each of the first plurality of wireless devices to one or more wireless access points;

extracting features of the training input scan data;

creating a plurality of training feature vectors that describe the extracted features of the training input scan data; and

at a training module, receiving definitions of a plurality of proximity classifications and producing a plurality of trained models on the basis of the training feature vectors and proximity classifications.

10. The method of claim 9, wherein the information regarding the accessibility of each of the first plurality of wireless devices to one or more wireless access points comprises:

for each of the one or more wireless devices of the first plurality of wireless devices, an identity of the wireless access points accessible to the wireless device;

for each of the one or more wireless devices of the first plurality of wireless devices, the number of the wireless access points accessible to the wireless device; and

for each wireless device of the first plurality of wireless devices, signal strengths of the accessible wireless access points.

11. The method of claim 9, wherein the extraction of features comprises one or more of:

determining a number of common access points, wherein the common access points are those access points accessible by two or more of the first plurality of wireless devices;

determining a number of access points that are not common to two or more of the first plurality of wireless devices;

determining a difference in a signal strength among a pair of wireless devices of the first plurality that detect the access point, the difference determined as a mean squared signal strength deviation over the common access points;

determining the percentage of the access point'"'"'s total signal strength detected by each wireless device of the pair of wireless devices in the first plurality of wireless devices, determined for each access point;

determining the percentage of access points that are common access points; and

for each pair of wireless devices of the first plurality of wireless devices, determining whether each wireless device in the pair has the same strongest access point as determined by signal strength.

12. The method of claim 9, wherein the production of trained models by the trainer module comprises creation of Gaussian mixture models.

13. The method of claim 9, further comprising:

receiving, at the feature extraction module, operational scan data from each of one or more of a second plurality of wireless devices, wherein the operational scan data comprises information regarding the accessibility of each of the one or more wireless devices in the second plurality of wireless devices to one or more of the wireless access points;

extracting features of the operational scan data;

creating one or more operational feature vectors that describe the extracted features of the operational scan data;

inputting the operational feature vectors to a classifier module;

receiving, at the classification module, the plurality of trained models; and

producing, by the classification module, a proximity classification for each pair of the one or more wireless devices of the second plurality of wireless devices using the trained models.

14. The method of claim 13, wherein the information regarding the accessibility of each of the one or more wireless devices in the second plurality of wireless devices comprises:

for each of the one or more wireless devices of the second plurality of wireless devices, an identity of the wireless access points accessible to the wireless device;

for each of the one or more wireless devices of the second plurality of wireless devices, the number of the wireless access points accessible to the wireless device; and

for each of the one or more wireless access devices of the second plurality of wireless devices, signal strengths of the accessible wireless access points.

15. The method of claim 13, wherein the extraction of features of the operational scan data comprises one or more of:

determining a number of common access points, wherein the common access points are those access points accessible by two or more of the second plurality of wireless devices;

determining a number of access points that are not common to two or more of the second plurality of wireless devices;

determining a difference in a signal strength among a pair of wireless devices of the second plurality that detect the access point, the difference determined as a mean squared signal strength deviation over the common access points;

determining the percentage of the access point'"'"'s total signal strength detected by each wireless device of the pair of wireless devices in the second plurality of wireless devices, determined for each access point;

determining the percentage of access points that are common access points; and

for each pair of wireless devices of the second plurality of wireless devices, determining whether each wireless device in the pair has the same strongest access point as determined by signal strength.

16. The method of claim 13, wherein said production of a proximity classification for each pair of the one or more wireless devices of the second plurality of wireless devices comprises:

for each pair of the one or more wireless devices of the second plurality of wireless devices, determining the trained model that most closely models the operational feature vector; and

identifying the proximity classification associated with the determined trained model.

17. One or more computer readable media comprising having computer control logic stored thereon, the computer control logic comprising logic configured to cause a processor to:

receive training input scan data from a first plurality of wireless devices, wherein the training input scan data comprises information regarding the accessibility of each of the first plurality of wireless devices to one or more wireless access points;

extract features of the training input scan data;

create a plurality of training feature vectors that describe the extracted features of the training input scan data;

receive definitions of a plurality of proximity classifications; and

produce a plurality of trained models on the basis of the training feature vectors and proximity classifications.

18. The one or more computer readable media of claim 17, wherein the information regarding the accessibility of each of the first plurality of wireless devices to one or more wireless access points comprises:

for each of the one or more wireless devices of the first plurality of wireless devices, an identity of the wireless access points accessible to the wireless device;

for each of the one or more wireless devices of the first plurality of wireless devices, the number of the wireless access points accessible to the wireless device; and

for each wireless device of the first plurality of wireless devices, signal strengths of the accessible wireless access points.

19. The one or more computer readable media of claim 17, wherein the extraction of features comprises one or more of:

determining a number of common access points, wherein the common access points are those access points accessible by two or more of the first plurality of wireless devices;

determining a number of access points that are not common to two or more of the first plurality of wireless devices;

determining a difference in a signal strength among a pair of wireless devices of the first plurality that detect the common access point, the difference determined as a mean squared signal strength deviation over the common access points;

determining the percentage of the access point'"'"'s total signal strength detected by each wireless device of the pair of wireless devices in the first plurality of wireless devices, determined for each access point;

determining the percentage of access points that are common access points; and

for each pair of wireless devices of the first plurality of wireless devices, determining whether each wireless device in the pair has the same strongest access point as determined by signal strength.

20. The one or more computer readable media of claim 17, wherein the production of trained models at the trainer module comprises creation of Gaussian mixture models.

21. The one or more computer readable media of claim 17, the control logic further comprising logic configured to cause the processor to:

receive operational scan data from each of one or more of a second plurality of wireless devices, wherein the operational scan data comprises information regarding the accessibility of each of the one or more wireless devices in the second plurality of wireless devices to one or more of the wireless access points;

extract features of the operational scan data;

create one or more operational feature vectors that describe the extracted features of the operational scan data; and

receive the plurality of trained models and producing, on the basis of the operational feature vectors and trained models, a proximity classification for each pair of the one or more wireless devices of the second plurality of wireless devices.

22. The one or more computer readable media of claim 21, wherein the information regarding the accessibility of each of the one or more wireless devices in the second plurality of wireless devices comprises:

for each of the one or more wireless devices of the second plurality of wireless devices, an identity of the wireless access points accessible to each of the one or more wireless devices of the second plurality of wireless devices;

for each of the one or more wireless devices of the second plurality of wireless devices, the number of the wireless access points accessible to the wireless device; and

for each of the one or more wireless access devices of the second plurality of wireless devices, signal strengths of the accessible wireless access points.

23. The one or more computer readable media of claim 21, wherein the extraction of features of the operational scan data comprises one or more of:

determining a number of common access points, wherein the common access points are those access points accessible by two or more of the second plurality of wireless devices;

determining a number of access points that are not common to two or more of the second plurality of wireless devices;

determining a difference in a signal strength among a pair of wireless devices of the second plurality that detect the access point, the difference determined as a mean squared signal strength deviation over the common access points;

determining the percentage of the access point'"'"'s total signal strength detected by each wireless device of the pair of wireless devices in the second plurality of wireless devices, determined for each access point;

determining the percentage of access points that are common access points; and

for each pair of wireless devices of the second plurality of wireless devices, determining whether each wireless device in the pair has the same strongest access point as determined by signal strength.

24. The one or more computer readable media of claim 21, wherein the production of a proximity classification for each pair of the one or more wireless devices of the second plurality of wireless devices comprises:

for each pair of the one or more wireless devices of the second plurality of wireless devices, determining the trained model that most closely models the operational feature vector; and

identifying the proximity classification associated with the determined trained model.