Existing indoor localization systems based on Wi-Fi Received Signal Strength (RSS) finger-printing often assume the knowledge of a map of the coverage area and involve a tedious manual survey process at a set of sample locations along this map. In this thesis, we describe an automatic graphical floor map and radio fingerprints generation system, called the Intelligent Mobility Mapping System (IMMS), which applies the concepts of crowd-sourcing and Simultaneous Localization and Mapping (SLAM) to construct a floor map and radio map in support of indoor people localization and tracking. The IMMS makes use of high similar patterns in crowd-sourced traces of RSS measurements to identify location segments in the coverage area and to construct a graphical floor map. Based on graph theo...[ Read more ]

Existing indoor localization systems based on Wi-Fi Received Signal Strength (RSS) finger-printing often assume the knowledge of a map of the coverage area and involve a tedious manual survey process at a set of sample locations along this map. In this thesis, we describe an automatic graphical floor map and radio fingerprints generation system, called the Intelligent Mobility Mapping System (IMMS), which applies the concepts of crowd-sourcing and Simultaneous Localization and Mapping (SLAM) to construct a floor map and radio map in support of indoor people localization and tracking. The IMMS makes use of high similar patterns in crowd-sourced traces of RSS measurements to identify location segments in the coverage area and to construct a graphical floor map. Based on graph theory, graph-drawing algorithms are designed and the interconnection relationship is drawn on a plane for vision. Then, the IMMS uses the RSS measurements to construct a radio map on top of the graphical map. With the IMMS, the elaborate off-line manual data collection process for radio fingerprinting can be eliminated. The maps constructed by IMMS will greatly facilitate accurate indoor localization and tracking.