Investigators: William P. L. Cully, Simon Cotton, William Scanlon & Jonathan McQuiston
Indoor personnel localization research has generated a range of potential techniques and algorithms. However, these typically do not account for the influence of the user’s body upon the radio channel. Here it is shown that disregarding body effects reduces the accuracy of the algorithms’ location estimates and that body shadowing effects create a systematic position error that estimates the user’s location as closer to the RFID reader that the active tag has line of sight to.
In this work an active RFID based patient tracking system is demonstrated and three localization algorithms are used to estimate the location of a user within a modern office building. The localisation system was based on activCampus, supplied by ACT Wireless Ltd. The selected arrangement consisted of a wrist-worn tag and four activReader RFID reader units operating at 868 MHz. Using this system several scenarios, both stationary and dynamic, were carried out in two environments.
Figure 1 shows the RSSI gathered at one reader as the user wearing the tag walks at constant speed from one point to another and back to the starting location by the same route. It can be seen that the peak RSSI is lower during the return journey.
The result of this on the estimated path can be seen in Figure 2 where the anticlockwise path is skewed towards the centre of the measurement area.
This project is supported by the Centre for Secure Information Technologies (CSIT) and ACT Wireless Limited.
Download a flyer for this project here.