Indoor Wireless Positioning Algorithm Based On RFID

Radio frequency identification technology(RFID)is a wireless non-contact system that uses radio frequency signals to automatically identify targets and obtain related information.In the field of indoor positioning,most traditional positioning technologies have disadvantages,such as small coverage and high hardware costs.With the rapid development of the Internet of Things and mobile communication technologies,RFID-based indoor wireless positioning technology has become an attractive technology in the indoor positioning field relying its high sensitivity,non-line-of-sight,and strong adaptability in the environment.Thus,the research on RFID-based indoor positioning technology has important practical significance.However,existing RFID indoor positioning solutions have the drawback of being strongly dependent on the network and low positioning accuracy with noise interference and multipath effects in the indoor environment.In view of the above issues,this paper aims to provide accurate and real-time location information,and research on some key links in RFID indoor location technology is conducted.The main work are as follows:(1)For the characteristics of indoor positioning environment,the framework and working principle of RFID positioning system are studied.Firstly,the common RFID positioning methods are analyzed from the aspects of positioning accuracy,cost and system complexity.Besides,methods of Gaussian filtering and Kalman filtering are adopted in this paper.And then,the collected signal is preprocessed to reduce indoor environmental noise and multipath effect interference.(2)Since that the traditional RFID positioning algorithms have poor positioning accuracy as well as cost laborious efforts constructing and maintain fingerprinting database,a novel Newton-HGWOSVR positioning algorithm is proposed in this paper,which is based on the Newton interpolation method and support vector regression optimized by hybrid grey wolf optimization algorithm.After completing the collection of position fingerprint information in the off-line phase,the acquired position fingerprint database is firstly expanded through Newton interpolation;Then,the HGWOVR algorithm is employed to build the nonlinear relationship between the RSS values of reference points and their locations.Finally,the position of mobile terminal is predicted by the constructed model.The simulation results indicate the average measurement error of the proposed algorithm is about 0.252 m,which has higher positioning accuracy than the LANDMARC positioning algorithm,Therefore,the comprehensive performance of the proposed algorithm is superior.(3)Considering that the RFID technology is extremely dependent on the wired network and consumes a lot of labor in the process of arranging reference labels,a positioning program combined with ZigBee and RFID technology is applied in this paper.Besides,integrate ZigBee module on RFID Reader,ZigBee node transmits the signal which is received by UHF Reader.According to the advantages of close range,low complexity,lowpower consumption and powerful networking capabilities,which can reduce the system hardware cost to some extent.To solve the shortage of VIRE,this paper introduces the signal path loss model to interpolate the grid reference label in the offline phase,and uses dynamic threshold to obtain the location of tags in the positioning stage.The simulation results indicate that the algorithm has lower cost,lower system complexity and better positioning performance than VIRE algorithm,which can meet the basic demand in reality.