| A localization system can save lives in the aftermath of an earthquake; position people or valuable assets during a fire in a building; or track airplanes besides many of its other attractive applications. Global Positioning System (GPS) is the most popular localization system, and it can provide 7--10 meters localization accuracy for outdoor users; however, it has certain drawbacks for indoor environments. Alternatively, wireless networks are becoming pervasive and have been densely deployed for communication of various types of devices indoors, exploiting them for the localization of people or other assets is a convenience. Proximity based localization that estimates locations based on closeness to known reference points, coupled with a widely deployed wireless technology, can reduce the cost and effort for localization in local and indoor areas. In this dissertation, we propose a proximity based localization algorithm that exploits knowledge of the overlapping coverages of known monitoring stations. We call this algorithm Sub-Area Localization (SAL). We present a systematic study of proximity-based localization by defining the factors and parameters that affect the localization performance in terms of metrics such as accuracy and efficiency. Then, we demonstrate that SAL can be used in multi-floor buildings to take advantage of the infrastructure elements deployed across floors to reduce the overall cost (in terms of the number of monitoring stations required) without harming accuracy. Finally, we present a case study of how SAL can be used for spatial spectrum detection in wireless cognitive networks. |
