Research On Indoor Positioning System Based On Radio Frequency Identification

With the development of wireless network and intelligent terminals,the demand for indoor position and navigation is increasing.Therefore,location aware systems have also been paid more and more attention to.A variety of indoor positioning methods have been proposed which are based on different sensors,such as ultrasonic,infrared,image,radio frequency identification technology(RFID)and Bluetooth.RFID has become a hot topic in indoor position due to its characteristics of non-contact,multi-target identification and cost-effective.This paper focuses on the study of passive RFID based indoor positioning method.First,the reference tags are laid in a certain way,the target will identify a set of response tag sequence and make use of intersection constraints of identification to get estimation area,then the target location is estimated by solving the centroid of irregular region.In order to simplify the computing process,this paper introduces the concept of virtual reference tag and converts centroid operation into operation of mathematical set by interpolation method.Besides,taking the interference of environmental noise into consideration,this paper use Interval Kalman filter algorithm to develop mathematical model of positioning process when the part of motion model is known.Finally,the simulation results show that the filter algorithm can effectively reduce the impact of noise on the positioning accuracy.In addition,the reasonable tag network layout is critical for positioning system.In this paper the layout of RFID tag is conducted intensive studies,least squares algorithm and Kalman filter algorithm are used to develop mathematical model to analyze the relationship between the tag layout and positioning error.Then we make the simulation experiments in different ways of tag network layout,the results show that the positioning error is minimized when in triangular distribution.