DEVELOPMENT OF INDOOR POSITIONING SYSTEM, BASED ON LEAFLET MAP, GEOLOCATION AND WI-FI POSITION ESTIMATION WITH ACCURACY ASSESSMENT

Abstract

Indoor positioning systems have a wide variety of applications. They can be used to see and track user location, for real-time navigation and others. They are usually developed for a specific building or set of buildings and may use different positioning techniques to meet the user needs in terms of accuracy and budget constraints. For the current study free and open source data is used for development of the base map � OpenStreetMap and Leaflet library. The data for specific building is added, as the floor plans are transformed to GeoJSON format. As positioning techniques, in this case, geolocation and Wi-Fi position estimation are used and there are three main steps for calculating user coordinates: � The user gets the available Wi-Fi signals strength and based on defined fixed limits the algorithm choose which of the preliminary coordinated and fingerprinted control points in the building is the closest; � Geolocation data for the user (latitude and longitude) is streamed and trough communication channel is imported in the developed system; � The user location is translated to the nearest control point and corrections are calculated as coordinate differences, similar to differential outdoor GNSS measurements with initialization on known points, but with modification only for the rover absolute position, so the indoor position of the user is corrected and shown on the map. The data was collected repeatedly static for 6 points in the building (3 on the first floor and 3 on the second), the corrections are calculated after 3 seconds of observations and applied to the coordinates for 1 minute intervals. All corrected coordinates are compared to the control ones, obtained by geodetic measurements in order to be analyzed the corrections validity. Also, tracks were made, so it could be seen what is the difference between the real moving trajectory, the corrected and the measured ones. The results of the analysis show that the data collected by this indoor positioning system is not very precise, compared to other systems, based on other positioning techniques � the difference compared to absolute position is approximately 1.8 m. Even though, this position estimation is much better than standard accuracy achieved by separate use of geolocation or Wi-Fi positioning with only 2 Wi-Fi sources, without preliminary defined signal strength fingerprint map.