Research On RFID Collaboration Localization Based On Multi-data Fusion

Location-based services have been widely and deeply explored in academia and industry.However,in the indoor environment,GPS technology is limited to some extent,which arouses great interest of researchers.Among them,using RFID technology for indoor positioning is an effective way to obtain location information.At the same time,the application of embedded RFID reader and writer in mobile devices is constantly promoted,and the positioning environment for multiple mobile readers and writers is gradually formed,which makes the collaboration localization method more and more concerned.Collaboration localization can be divided into two types: centralized and distributed.In the field of collaboration localization,the key factor to achieve accurate location calculation is the accurate estimation of location data and the effective interaction of location information.Therefore,this paper focuses on: 1)accurate estimation of location information.Using different localization methods to carry out efficient location estimation is the basis of accurate location,so it is also the focus of this paper.Then,a centralized collaboration way is used to fuse and process multiple location data to improve the accuracy of location calculation;2)effective interaction of multiple location information.By making use of the availability and universality of mobile RFID reader,the data between multiple mobile readers and tags can be exchanged in a distributed and collaborative way to further improve the accuracy of localization results.In order to achieve the above requirements,this paper proposes a RFID collaboration localization method based on multi-data fusion,which includes two kinds of cooperation methods:tag oriented centralized localization and reader writer oriented distributed localization.The former is mainly for the processing of multi location data,and mainly focuses on the interaction of multi reader writer data.In details,Centralized Collaboration LocalizationCentralized collaboration localization is the focus of this paper,mainly for the fusion of multiple positioning data.In order to realize the accurate estimation of location data,this paper studies the label oriented location detection,and designs an effective location estimation algorithm.(1)A State Detection Approach Based on Doppler FrequencyWith the change of environment or time,the wireless signal at the same place will change,which leads to the static drift phenomenon,that is,when the user is in a static state,the positioningresults will jump.In order to achieve accurate positioning,it is necessary to first determine whether the user produces a movement.In this paper,the influence of user behavior on Doppler frequency under different states is analyzed in detail,and then an effective state detection method is proposed,which can accurately identify five states: static,standing,sitting,lying down and moving.The experiment results show that this method not only solves the static drift problem,but also can accurately perceive the various behavior states of users.(2)Localization using Ranging in Segment Area based on Fuzzy LogicThere is a direct relationship between the intensity of RFID signal and the distance between reader and label in ideal propagation environment,but in actual situation,the measured signal and distance information can not guarantee one-to-one correspondence,which results in positioning error.In this paper,the influence of multi-path propagation and tag of reader antenna on distance estimation and the limitation of simple signal path propagation model are analyzed,and a ranging method in segment area based on fuzzy logic is proposed.The experiment results in different dimensions show that the proposed method not only outperforms the location results calculated by using simple signal path propagation model,but also outperforms the maximum likelihood location method.(3)Fingerprint localization based on spatial correlation influence factorAt present,one of the most important problems of signal fingerprint is the dynamic and unpredictable environment,which makes the RSS signal fluctuate greatly in the actual deployment process,which is manifested in the instability of time domain and the arbitrariness of space domain,which directly affects the positioning accuracy.By studying the spatial relationship between RSS signal and its neighbors,this paper designs a new signal feature,spatial correlation influence degree,to improve the resolution of signal representation and reduce the influence of the original signal fluctuation.At the same time,based on this feature,an effective positioning method is designed and optimized for fingerprint filtering.Experiments show that the performance of this method is better than that of traditional location methods.Distributed Collaboration LocalizationIn this paper,the distributed collaboration is mainly used in the positioning system constructed by the collaborative method.The distributed collaborative positioning is realized by using heterogeneous and independent mobile readers and tags' residual memory.The collaborative methods include direct collaboration and indirect collaboration of readers and writers.During the positioning process,if there is no memory left for the tag to store information,the reader and writer cooperate directly.On the contrary,if there is residual information in the tag as storage space,the reader and writer can cooperate indirectly.(4)RFID Location based on UKF filter and Reader CooperationIn order to make full use of the advantages of various positioning methods,this paper proposes a positioning method based on UKF filter and reader cooperation.On the one hand,the state detection method is used to distinguish the user's dynamic / static state,and the trackless Kalman filter is used to fuse the location results of the above ranging and fingerprint location methods,so as to realize the accurate centralized location.On the other hand,by using the communication between the reader and the tag,we design the positioning method for direct collaboration and indirect collaboration.By using the reader and writer collaboration in different situations,we provide a distributed localization scheme to further improve the localization effect.The experimental results show that the collaboration localization method can not only improve the accuracy of localization,but also change the centralized and coordinated way of RFID system into a distributed and scalable way.To sum up,the RFID collaboration localization method based on multi-data fusion proposed in this paper combines the advantages of centralized and distributed collaboration.According to the experimental results,the collaboration localization system not only improves the localization accuracy,but also it greatly reduces the communication overhead in the localization process.