Document Type thesis Author Name Heidari, Mohammad URN etd-050405-144528 Title A Testbed for Real-Time Performance Evaluation of RSS-based Indoor Geolocation Systems in Laboratory Environment Degree MS Department Electrical & Computer Engineering Advisors Kaveh Pahlavan, Advisor Allen H. Levesque, Committee Member Wenjing Lou, Committee Member Keywords Performance Evaluation RSS-based fingerprinting algorithm Testbed Indoor Geolocation Indoor Positioning Date of Presentation/Defense 2005-04-28 Availability unrestricted
Recently, there has been an enormous growth of interests in geolocation applications that demand an accurate estimation of the userâ€™s location in indoor areas. The traditional geolocation system, GPS, which was designed for being used in outdoor environments, does not perform well in indoor areas, causing frequent inaccuracies in location estimation. Therefore the need for more accurate positioning systems and even positioning techniques is a motivation for researchers to turn their attention into indoor positioning systems.
In this thesis we present a unique testbed for indoor geolocation systemâ€™s real-time performance evaluation. Then we present a real-time performance evaluation of a sample indoor positioning system. We make a comparison between the simulated results of the performance evaluation of the positioning engine and the real-time performance evaluation of the positioning system. Finally, we perform a sensitivity analysis for Ekahauâ„¢ indoor positioning engine. We show that the simulation with the introduced testbed yields the same results as one would obtain by evaluating the performance of the positioning system by means of massive measurement campaigns. Running the testbed for several measurement campaigns for different scenarios enabled us to compare the results and study the effect of selected parameters on the performance of the positioning system. We also perform primitive error analysis in terms of distance error to verify the validity of the result obtained with the testbed. We show that under the same configuration both real-time performance evaluation and simulated performance evaluation will yield same result with respect to position error. We also use error modeling to determine which error model is best matched to the observed indoor positioning error.
Amongst all of the possibilities of choosing methods of positioning, we focused on the Received Signal Strength (RSS) based method along with fingerprinting. Briefly said, profiles previously gathered by measurement or simulation will decide on the location of mobile terminal if a new profile comes in.
It is worth mentioning that previous work similar to this testbed has been done for outdoor areas according to Ekahau's white paper. Their work is mainly focused on outdoor environment, in which multipath does not exist. In this research effort we tried to analyze the effect of different parameters on sensitivity of indoor positioning systems who suffer from multipath. Different setups for simulating real-time radio channels have been studied in literature, but still not focused on indoor areas.
Browse by Author | Browse by Department | Search all available ETDs
Questions? Email firstname.lastname@example.org