Error Mitigation in RSSI-Based Fingerprinting Localization Using Multiple Communication Channels
Francisco Javier Mora-Becerra · Lincoln (University of Nebraska) · 2016
Location data from radio signal strength indication (RSSI) based wireless networks has been used in various applications such as creating smart home behavioral monitoring systems, tracking health care workers for the spread of hospital-associated infections, and providing location-aware tour guide systems. Because RSSI-based systems are inexpensive and can be used with most wireless devices without requiring additional hardware, they are a popular choice for localization. Unfortunately, multipath fading dramatically degrades the performance of an RSSI-based system's ability to locate a target indoors. This thesis endeavors to reduce localization error for RSSI-based fingerprinting localization systems in an indoor environment through frequency diversity by using multiple communication channels. By creating a multichannel fingerprint of the environment using fingerprinting calibration techniques, fine-grained, 5 centimeter, 2-dimensional localization accuracy is achieved in an indoor environment under certain restrictions. Adviser: Lance C. Pérez