Keoh , Choo Kent
(2017)
Wireless Local Area Network Based Indoor Positioning System.
Masters thesis, Universiti Sains Malaysia.
Abstract
This research is aimed to overcome the poor coverage of the Global Positioning System (GPS) in indoor environment. The indoor positioning system developed is based on the wireless local area network (WLAN) using the IEEE 802.11n standard. The positioning system is designed to self-locating of a user/object with the minimal setup of hardware and adequate accuracy. The system is implemented in a laptop to provide a visualized graphic user interface to shows the location of the user/object on the predefined map. The positioning system used the triangulation concept with the fingerprint method to locate the position of the user/object. An Arduino Wi-Fi Shield CC3000 is connected to an Arduino Mega is used to collect received signal strength (RSS) from individual wireless routers in predefined space to determine the location of the user/object. The data then will go through the predefined lookup table to locate the location of the user/object and the location will send to the MATLAB. The location of the user/object will then show in the graphic user interface in MATLAB. A set of experiment is conducted to identify the behavior of the RSS value toward the router. From the experiment, the potential factors that affect the performance of the positioning system is examined and identified. Besides that, a set of database of the RSS value on the predefined 12 points is collected to develop the lookup table. Lastly, the system is verified, validated and tested with a few experiments to test and determine the accuracy and stability of the system. The positioning system is developed with a fine performance and with a best accuracy of 100% at radius of less than 2m around the predefined point, accuracy of 70% - 60% at radius of 2 - 3 m radius, accuracy of 50% - 40% at 3 - 4 m radius and 30% - 20% at 4 - 5 m radius. When the area of the experiments around the predefined point increased the accuracy drop
Actions (login required)
 |
View Item |
![[img]](http://eprints.usm.my/style/images/fileicons/application_pdf.png)
