Continuous indoor navigation with RFID and INS

Abstract

Some navigation applications, such as the navigation of blind users, require that a continuous positioning is performed in combined outdoor urban and indoor environments with a certain positioning accuracy. For outdoor urban environments usually GNSS and dead reckoning are employed and has proven to be satisfactory. In indoor environments, however, absolute position determination with an accuracy in the range of 1 to 2 meters is still very challenging. For indoor location determination, a number of different positioning methods have been developed. In our approach, Radio Frequency Identification (RFID) has been selected and investigated. Using RFID most commonly cell-based positioning is performed. Apart from that we have also investigated trilateration and location fingerprinting based on signal strength measurements (i.e., RSSI short for Received Signal Strength Indication) from the RFID tags in the surrounding environment. The disadvantage of these two positioning methods, however, is the required calibration in the off-line or training phase to deduce ranges to the tags from the RSSI measurements in the case of trilateration or the establishment of the RSSI database at known locations throughout the building in the case of location fingerprinting. Therefore we have developed a new location method based on cell-based positioning which makes use of the measured RSSI in the on-line or positioning phase, i.e., the so-called time-based Cell-of-Origin (CoO). Two modifications have been implemented in comparison to common CoO and will be discussed in the paper. The new approach was then be tested in an indoor environment in an office building of the Vienna University of Technology. It could be seen that for a combined positioning of RFID time-based CoO and a low-cost MEMS-based INS positioning accuracies on the 1 to 2 meter level can be achieved. The different experiments performed in the test bed are described and discussed in this contribution.