indoorpos1
Basement Android project for fingerprint-based indoor positioning system
Interpolation mechanism is embedded within broadcast receiver called WiFiScanReceiver.
Simple interpolation based on Euclidean distance has been used as an initial template which may further be
enhanced to involve more complex interpolation models e.g. based on probability distributions.
In order to use positioning it is required to have
-image files representing map(s) of indoor location(s) so that the program may automatically switch from
one map to the other as the user walks indoors; when the user leaves outdoors away from map-covered area,
the program launches GPS-based outdoor positioning and stays in that mode until the user is back indoors
-list of access point (AP) MACs with radio signals reaching at least one out of all indoor locations mentioned above
-radiomap consisting of 2-D points over all indoor locations (can further be extended to 3-D by considering
building floors and net georaphic height of landscape) and keeping information about RSS (Received Signal Strength)
fingerprints from every AP at each of those points
Examples of these file have been included in the project's main trunk (see n5_2f.png, N5-ap.txt, N5-radioMap.txt)
These data are supposed to be fetched on-the-fly in JSON format from a remote storage, preferably NoSQL-based
document-oriented DB to enhance data availability and make the program better fit real-time constraints.
Prior to using the map images it is necessary to tune the program so that it works correctly with dimensions
of Android device and properly shows user's current position on the screen.