Research On Precise And Real-time RFID Relative Localization

Radio frequency identification(RFID)as an important promoter of automatic identification technology has been widely used in many smart applications.In some application scenarios,it is necessary to quickly and accurately obtain the relative positional relationship between items to achieve the purpose of item inventory.RFID tag relative localization solution is widely used in book management,pipeline sorting and other application scenarios due to its low cost.The relative localization method proposed so far has a large localization delay,and the accuracy needs to be improved,which greatly limits the application of RFID technology to actual scenarios.Further,in certain application scenarios,it is necessary to quickly and accurately distinguish items in different groups,such as automatic settlement for unmanned supermarket shopping,logistics goods distribution,etc.The relative localization of the RFID tag set can quickly distinguish different groups of RFID tags that can be used to distinguish items of different groups.The two problems raised above are studied separately,and the specific work is as follows.Firstly,this paper proposes a relative localization method for low-latency tags.By collecting tag EPC,phase and time information,using tag phase information and time information data features to achieve the purpose of low-latency and high-precision tag relative localization.Specifically,our method can identify the "V" zone immediately after the antenna passes the tag.And the method only uses the phase readings in the core "V" shaped area,instead of using all the phase readings to calculate the order of the tags.This avoids unnecessary query operations on the tags after determining the order of the tags,thereby reducing the time used in tag conflict elimination.In addition,since the phase reading in the core “V” area can be obtained when the antenna and the corresponding tag are close,and the data quality is better than the data outside the area(for example,the amount of data loss is small and the phase ambiguity is small),thus This ensures that the localization accuracy of the method is high.We also propose a power control method to adjust the interrogation range of the reader antenna to further improve the quality of the data obtained.In addition,using only the data in the core "V" shaped area can also improve the localization accuracy of the boundary tags.Secondly,this paper proposes a relative localization method for tag sets.By collecting tag EPC,RSSI and time information,the data features of tag RSSI information and time information are used to achieve the purpose of relative localization of tag sets with high accuracy.Localization can be divided into two parts:tag information collection and tag collection division.Tag information collection is often not difficult,and the focus of the algorithm is on tag sets division.The algorithm we give mainly divides the tag set in two steps: first,the tag with obvious data features corresponding to the tag motion path are divided into the initial tag set,and then based on the division of the tag set,the tag set boundary is found,and the spatial distribution characteristics of the tag data are used to determine the set of the tag that does not have the incremental characteristics of RSSI data.The experimental results show that our tag sets relative localization algorithm can successfully distinguish the tags of different sets with high accuracy.