Abstract: RFID (radio frequency identification) technology has been widely applied to the communication between RFID readers and dynamic electronic labels,but the data captured by RFID readers often tends to be noisy．In order to provide a better support for high-level RFID’s applications,it is necessary to clean the collected data．Considering the characteristic of frequent movement of labels,this paper put forward an improved Kalman filter model by combining the sliding window technique with Kalman filter model and then proposed a method of RFID data cleaning based on an improved Kalman filter．The method can not only guarantee the accuracy of data cleaning but also effectively solve the problem of time delay caused by dynamic electronic tags．Thus this method is more adaptable to the situation where labels are moved frequently．The experiment’s result shows this approach can improve the efficiency and accuracy of the data cleaning．

Key words: Kalman filter,RFID,Data cleaning,Slide window,Dynamic label