Research On Indoor Positioning Technology Based On Rfid

Radio Frequency Identification, referred to as RFID, is a kind of wireless automatic identification technology. It uses radio-frequency signal to identify the target automatically and obtain relevant information. At present, with the development of society, location-aware technology gets more and more favour from scholars and research institutions. For indoor positioning, the traditional positioning technology can not meet the requirement of the indoor environment. Because of the advantages of non-contact, short time delay, high precision and low cost, indoor positioning technology based on RFID becomes the prior choice of indoor positioning. This thesis mainly studies on the localization of RFID tags for indoor wireless propagation environment, and the following contents are included:Firstly, the research condition of indoor positioning technology in China and other countries is analysed by consulting a large number of literature, and the limitations and deficiencies in the study of indoor location are summed up.Secondly, for indoor two-dimensional spatial, the LANDMARC and VIRE localization algorithm are mainly studied in this thesis. For LANDMARC, simulation and analysis of the factors affecting the positioning performance are shown in this thesis, and the trilateral auxiliary localization algorithm and adaptive K neighbor algorithm are analyzed in theory and simulation. Then an improved LANDMARC Algorithm based on preprocessing of reference tags is proposed. Some tags may be affected by the environment and bring in large positioning error, so the improved algorithm filters out these tags through preprocessing, then adjusts the weighting factor to locate tags. The simulation results show that the proposed algorithm can effectively improve location accuracy.Thirdly, Compared with LANDMARC, VIRE introduces the concept of virtual reference tag and approximate map. The simulation results show that the positioning accuracy of VIRE algorithm is higher than LANDMARC algorithm, but it’s positioning accuracy of the boundary is still significantly lower than the center area. So BVIRE algorithm is studied. The BVIRE algorithm can effectively improve the positioning accuracy of the boundary by bring in virtual reference tags in the border. Moreover, the RSSI value of the virtual reference tags is calculated by linear interpolation in VIRE algorithm. This linear interpolation calculation is just a kind of fast algorithm, and it’s quite different from actual situation. Thus it leads to large positioning error. So, the Newton interpolation algorithm is adopted, which calculates the RSSI value of virtual reference tags by Newton interpolation. The simulation results indicate that it not only fits the actual situation, but also offers good positioning effection. In addition, a concise positioning simulator is designed in order to compare the positioning accuracy of the algorithm straightforwardly. In order to use the emulator, you first need to input the relevant parameters, and select algorithm, then locating results are displayed.Fourthly, for indoor three-dimensional positioning, the extension of LANDMARC and VLM are studied. Detailed theoretical analysis and simulation are made respectively. The results shows that LANDMARC is still affected by the density of reference tags and VLM is affected by the density of virtual reference tags. But VLM can locate simply and rapidly by using virtual reference tags with the constraints of inclusive and exclusive. The performance requirements of the reader is not high for VLM because the readers just need to detect the signals of tags rather than receive accurate RSSI values. So it reduces the costs of positioning.Finally, the research work was summarized and the next research topics in future were discussed.