A research of time and location based RFID data cleaning

Radio Frequency Identification (RFID) is increasingly being used and has been successfully applied to various areas, such as retail, pharmaceutical and pet tracking. Despite its obvious popularity, RFID data reliability is always the problem which should be addressed. In this paper, a data cleaning technique based on time and location is proposed to improve RFID reading´s reliability. For temporal analysis, sliding-window based on tag´s linger time is used, and in the complex node structure the tag´s travel time is compared to improve data reliability. For location analysis, by analyzing the complete path, which route error takes place is determined, and then the adjacent node information is used to clean data. This method is feasible and efficient according to the experiment.