Active radio frequency identification asset tags for positioning: an application of clustering and correlation technologies

Abstract

This study presents an algorithm for a positioning system by using 5 active 433 MHz radio frequency identification (RFID) tags and a reader to identify the positions of targets, such as patients and wheelchairs, in buildings. The RFID tags were distributed near elevator exits on the second and third floors, and outside the doors of three laboratories in the building. An RFID reader was set in a wheelchair and connected to the Internet via a laptop. The different amplitudes recorded on each tag clearly identified the positions of the tags. Calculating the error of distances, the differences between self- and cross-correlation coefficients, and the K-means clustering of these tags enables dividing the space under surveillance into several adjoining regions. Thus, this approach effectuates enlarging the space which can be covered.