Building a Location Sensing and Positioning Service to Enable Indoor Location Based Services



To provide location based services for mobile devices inside buildings and/or on devices that do not support the global positioning system (GPS), other methods must be chosen to get to know one's current position. Since WiFi covers almost all urban areas, the signal strength of WiFi routers will be a quite a reasonable source of information to calculate ones position. Being able to bind this technique to universitary use, better WiFi coverage is assumed through Eduroam whose routers should be a quite useful source for the task.


- Write a program that aquires data from the WiFi device and uses the signal information of all WiFi routers in the perimeter to triangulate the device.
- Build up a database storing Information about some Eduroam routers for testing purposes. This database will include information as SSID and location of the router possibly along with some other data that is useful for the task at hand.
- Write a program that uses the data provided from the tasks above to calculate the current GPS coordinates of the device. (An alternative may be to establish some local coordinate space (e.g. for a specific area or building) and locate the device within this space.)
- Nice to have #1: Provide the calculated coordinates to the Google API and display the current location on Google Maps


Location based services (LBS) are a very interesting area and techniques aiming to locate a certain person or device are very important as a basic necessity to provide services for specific locations. Using location based services, one is able to request information or services bound or related to a specific location, i.e. a certain place or building or even a specific room. There is a broad variety of possible applications, ranging from information about schedules in public transportation to informations about collegues or friends in proximity. \par

The probably most prominent technique providing localisation of mobile devices - the Global Positioning System (GPS) - is able to locate mobile devices with an accracy of a few meters. Today, the Global Positioning System is being used by many systems and applications like - being the most prominent one in civilian applications - navigation for cars and pedestrians. Being the only global localisation system with global coverage being available to anyone at reasonable prizes, the Global Positioning System became the todays world leading positioning technique. Even though, the Global Positioning System has some major drawbacks when it comes to localisation inside buildings or within dense urban areas, where buildings weaken and sometimes entirely block the signals of the satellites, thus weakening the accuracy of the localisation or rendering it entirely impossible in the worst case. These drawbacks of the Global Positioning System fueled the research on other approaches, making indoor localisation not only possible, but in some cases even more accurate and resilient than the Global Positioning System.

Approaches were proposed with the objective to build rather local but global positioning systems specialising on indoor deployments, covering additional special requirements like multiple floors. Some of those approaches that has been proposed are using data collected from special sensors like gyroskopes, image sensors or infrared vision. These techniques where mainly proposed with a mobile robotic unit in mind, equipped with an array of adequate sensors. To realise the localisation of people - or to be more precise devices people are used to carry around everyday like smartphones or equivalent devices capable of carrying out complex calculations, these approaches are not very practical, since people or their devices would need to be equipped with these kind of sensors, what turns out to be very expensive and - in cases like infrared sensors - quite impractical.

With the growing distribution of radio frequency (RF) based techniques like WiFi (IEEE 802.11a/b/g/n) or Bluetooth (IEEE 802.15.1), proposals were made to use the signals of the fixed WiFi base stations for localisation purposes. In fact, actual deployments already showed that localisation using WiFi, optionally combined with other RF-Techniques like Bluetooth or GSM is indeed possible and quite accurate. One big advantage is that the already existent RF infrastructure is used, so there are no additional costs for sensors or positioning beacons.

Since corperations and institutions like universities often offer wireless networks on IEEE 802.11b/g/n-basis with almost full coverage of the site or campus to provide wireless internet access to their employees and/or students, the signals of these networks can be used as a source to estimate ones current position in the covered area - even and especially indoors. In fact, the idea of using WiFi networks is not new as one of the first proposals was made by P. Bahl and V. Padmanabhan in 2000.


diploma_thesis.pdf (5.43 Mb)

http://developer.android.com/index.html(external link)


Mostafa Akbari

Created by Daniel Herding. Last Modification: Monday, 17. January 2011 16:21:06 by Hendrik Thüs.