RFID based indoor location tracking

Abstract

Radio Frequency Identi cation (RFID) attracts considerable attention because of its low cost and various area of applications. Location estimations based on received signal strength (RSS) in wireless technologies took the topic of localization using RFID technology into agenda. In our study, changes in the RSS values according to distance have been investigated in an indoor environment based on both passive and active tags. For the system in which active RFID tags are used, localization has been made using reference tags. An online radio map of the environment is created with the help of deployed reference tags. A nearest neighbor algorithm is applied in order to localize the tracked tag. However, unpredictable and non-uniform behavior of tags resulted in inappropriate location estimations. Passive tags, on the other hand, severely su ers from shadowing e ect, creating biases towards to shadowed tags. The tag worn by the subject was undetectable because of the signal absorption by the human body. We have developed an algorithm estimating the location based on the drops in RSS value of the tags as tracked subject is passing by. We achieved a mean error of 0:25 m in a 2x2:5 m2 indoor environment.