Systems and methods for localization
US9584969B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Dec 23, 2014 |
| Grant date | Feb 28, 2017 |
| Priority date | — |
| Expiry date | Dec 23, 2034 |
Classification
- Technology area (CPC H)Electricity
- CPC primaryH04W4/33
- WIPO fieldDigital communication
- WIPO sectorElectrical engineering
Abstract
Systems and methods for a localization system are provided. In one aspect, a RF signature map for a geographical area is determined using a Gaussian Process (“GP”) model. Training RF measurements are taken at some locations within the area to train the GP using the Firefly Algorithm (“FA”). The RF measurements for other locations of the area are predicted using the conditional probabilities of the trained GP and without taking RF measurements at those other locations. The RF signature map is used for fingerprinting localization. In one aspect, a reference RF signature map is constructed for one, some, or all access points (“APs”) covering the area. A location of a user device, such as, for example, a smart phone, is then estimated by comparing the RF signals received by the user device from one or more APs with the determined one or more reference RF signature maps using a combined likelihood function.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.