Device based Multi-User Tracking System using Internet of Things

Abstract

In Light Dependent Resistor (LDR) sensor-based user is localized based on the event and the intensity of the room light when a user enters inside a room and switch ON the lights, the intensity goes high, an entry is noti?ed. An exit is noti?ed when a user switches OFF the light and exit the room. Moreover, the model remains prone to more error in multi user localization because multiple users may enter inside same room at same time and the lights of many rooms remain ON which makes more difficult to localize a user. In order to overcome this ambiguity of light sensors, two passive infrared (PIR) sensor with radio frequency identi?cation (RFID) tag-based model has been proposed, where every user has a tag. In this system, 10 PIR sensors and 5 RFID readers were attached to house room (10.0 m * 6.0m). An entry is noti?ed if the following pattern form, the outer PIR detects a motion and waits for few seconds, next the RFID reader reads the tag given to the user and ?nally the inner PIR detects a motion within the given time delay. An exit of a user is noti?ed only if the pattern from inner PIR to outer PIR is followed with the given time delay. The RFID tag is used to identify which user has entered a room at a particular time and also ensures unauthorized entry. The LDR based system gives accuracy nearby 20% but the multi-person tracking in a binary infrared sensor network-based system gives accuracy near about 90%. In this paper, the proposed PIR sensor along with RFID based indoor navigation system gives accuracy near about 94%.