Research On The Localization Of Passive UHF Radio Frequency Identification Based On Phase

With the appearance of the new concepts "Smarter Planet" and "Industry 4.0",Radio Frequency Identification(RFID),a key technology of Internet of Things(IoT),has been a hot research topic in recent years.Owing to the long read range,fast read speed and low cost,RFID in ultra-high frequency(UHF)using passive tags has achieved great development.This paper is devoted to the research of the localization of passive UHF RFID tags based on phase information.The research contents are as follows:(1)The computer software of the reader is developed to acquire the phase of the reflected signal of RFID tags.It can be used to set the reader parameters,identify the tags,and display/save the data,etc.It has advantages of the simple interface,easy operation and extension,which meets the requirements of the research.(2)The feature of the phase of the reflected signal is studied.An experiment is designed to verify the relation between the phase value and the distance between the tag and the reader antenna.From the experimental results,it can be concluded that the phase of the reflected signal keeps stable and sensitive,which is suitable for localization.(3)The methods of phase unwrapping based on the phase difference are studied.The angle of arrival(AOA)based on the phase difference of double antennas is analyzed by the experiment.Meanwhile,a method for one dimension localization based on the phase of single antenna is proposed,which can avoid the ambiguity of the phase.The experimental results show that this method has a satisfied localization performance.In addition,by measuring the tag signals repeatedly under a fixed environment,the effect of the fixed environment on the tag signals is similar.Hence,the phase can be used as the fingerprint characteristic to identify the position of the tag.(4)A novel approach using the similarity measurement of virtual reference tags for localization is proposed.The 3 ? principle with variable coefficient is presented to eliminate the multipath error.The natural neighbor interpolation(NNI)combined the transition points of the periods is introduced to estimate the value of the virtual reference tags.The experiment demonstrates that the method has high localization accuracy and stability compared the methods based on traditional kNN and BP neural network.