Nielsen, JohnLachapelle, GérardLi, Yuqi2017-04-212017-04-2120172017http://hdl.handle.net/11023/3718The ever-growing demand for Location Based Services has significantly boosted the research and development need for indoor positioning systems. Of various indoor positioning solutions, techniques making use of Received Signal Strength (RSS) of wireless signals of opportunity have gained extensive interest due to the ubiquitous wireless signal infrastructure and the readily available RSS measurements with standard mobile devices. However, the performance of RSS-based indoor positioning systems is highly affected by significant uncertainties in RSS due to many factors affecting wireless propagations. To enhance the performance of an RSS-based indoor positioning system, from a Bayesian filtering theory perspective, a better estimation of the a posteriori distribution of position is needed. This can be done through a better modelling of RSS measurements to mitigate uncertainties and/or incorporating prior information. This thesis specifically explores mitigating RSS uncertainties by modelling those due to human body shadowing and incorporating prior information from widely available security cameras and building maps. The characterization of RSS measurements indoors is first demonstrated using data collected in various environments. Experimental results characterize the RSS sensitivity to location and the uncertainty incurred by body shadowing effects on RSS measurements. Based on the characterization, an empirical model with a small number of parameters estimated from training data is proposed to model the RSS loss due to body shadowing. An estimator based on this model is proposed to improve positioning. Experimental results show that when the user heading is known, the positioning obviously improves. When the heading is unknown, and thus needs to be jointly estimated, the improvement becomes less apparent. This thesis then investigates the use of security cameras and building maps to enhance RSS-based positioning. An estimator based on computer vision processing is proposed to estimate user’s heading in corridors. Based on this, a camera-aided RSS system based on Kalman-filter is proposed and it is experimentally shown that a 37.5% improvement in horizontal position estimation occurs. To further incorporate building map information, a map-camera-aided RSS system based on particle filters is proposed. Experimental results indicate that the use of map constraints further bring 44.4% improvement in the across track direction.engUniversity of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission.EngineeringEngineering--Electronics and ElectricalIndoor-positioningRSSSecurity-cameraMap-matchingdoctoral thesis10.11575/PRISM/25912