ReLoc 2.0: UHF-RFID Relative Localization for Drone-Based Inventory Management

Radio frequency identification (RFID) technology attracts extensive attention for industrial applications, especially supply chain businesses and inventory management. In this article, a passive RFID localization scheme based on unmanned aerial vehicles (UAVs) or drones is established management in warehouses. case of the major challenge on-board antenna tracking errors practical scenarios, state-of-the-art (SOTA) numerical methods rely accurate positions, such as synthetic aperture radar (SAR)-based algorithms, which will introduce large positioning become unreliable. To end, we propose new relative method phase, received signal strength indicator (RSSI), readability, little affected by errors. proposed method, tagged items racks are located laterally through pinpointing minimum unwrapped phases, have been hybrid random forest (RF) model. RSSI differences read count each patch antenna's beam utilized to distinguish rack level assets via boosting tree classifier. The experimental results real warehouse show that achieves 27.1-cm mean lateral accuracy orders horizontally vertically with more than 96.7% 98.0% accuracy, respectively.