On Fast and Accurate 3D RFID Mobile Localization

This paper proposes an ultrahigh-frequency (UHF) radio frequency identification (RFID) based 3D mobile localization system (3DRML) for passive tags and tagged objects. Influenced by factors such as calculation model, grid scale and phase center shift (PCS), prior RFID based 2D and 3D mobile localization methods are subject to certain restrictions in computational time and accuracy. To overcome these limitations, 3DRML has the following features. First, 3DRML achieves grid based mobile localization with low time cost by leveraging the idea of reflection coefficient reconstruction (RCR) which regards each point representing an area as a reflection point and calculates the reflection coefficients from simple matrix operations. Second, a PCS calibration process is performed to compensate the phase shift caused by the antenna phase center change. Third, 3DRML uses the nonlinear optimization algorithm to solve the least square localization model for a quick localization, and then constructs a much smaller grid area to facilitate the grid based real-time accurate localization. The performance of 3DRML is evaluated by simulations with various interferences, and the results show that 3DRML enables fast 3D localization while achieving higher accuracy.