Characterising indoor positioning estimation using experimental data from an active RFID-based real-time location system

Indoor positioning has attracted much research effort due to many potential applications such as human or object tracking and inventory management. Whilst there are a number of indoor positioning techniques and algorithms developed to improve positioning estimation, there is still no systematic way to characterise the estimation. In this paper, we propose a method comprising of three characteristics to characterise indoor positioning estimation. We conducted experiments on an active radio frequency identification (RFID)-based real-time location system in different environmental conditions. We used both a human and a robot to traverse two experimental areas and collected positioning results at different fixed points along the traversal path. Using this basic positioning data, we were able to characterise positioning estimation using three characterisations: position accuracy, centroid consistency and angular distribution. We demonstrate the use of these characteristics for examining different points in a travelling path and different measurements.