Date of Award


Document Type


Degree Name

Master of Science (MS) in Computer Science


Computer Science


"Location based services are becoming an important part of life. Wide adoption of GPS in mobile devices combined with cellular networks has practically solved the problem of outdoor localization needs. The problem of locating an indoor user has being studied only recently. Much research contributed to the innovative concept of an indoor positioning system. By analyzing different technologies and algorithms, this thesis concluded that, considering a trade-off between accuracy and cost, a Wi-Fi based Fingerprint method is proved to be the most promising approach to determine the location of a mobile device. However, the Fingerprint method works in two phases-an offline training phase (collection of Received Signal Strength signatures) and an online phase in which data from the first phase is used to determine the current position of a mobile user. The number of training points in a certain area has a direct impact on the accuracy of the system. As a result, the offline phase is a tedious and cumbersome process and the positioning systems are only as accurate as the offline training phase has been detailed. Moreover, the offline phase must be repeated every time a change in the environment occurs. To avoid these limitations, we focus on improving the accuracy of the indoor positioning system, without increasing the number of training points. This thesis presents a Wi-Fi based system for locating a user inside a building. The system is named WiFiPoz, which means Wi-Fi positioning system based on the zoning method. WiFiPoz has a novel approach to Fingerprint method that incorporates Propagation and zoning methods. Experimental results show that WiFiPoz is highly efficient both in accuracy and costs. Compared to traditional Fingerprint methods, with the optimization of the accuracy of the location estimation, WiFiPoz reduces the number of training points. This feature makes it possible to quickly adapt to changes in the environment. In order to explore another possible solution, this thesis also developed, implemented and tested an indoor positioning system named GIS (Geometric Information based positioning System), which is based on a model proposed by another researcher. Several experiments were run in the offline phase and results were compared between the traditional Fingerprint method, GIS and proposed WiFiPoz. We concluded that WiFiPoz is a more efficient and simple way to increase the accuracy of the location determination with fewer training points"--Document.


Typescript. Vita.