Person tracking using Kalman Filter in Wireless Sensor Network

Abstract

Wireless Sensor Network (WSN) is an emerging technology for person detection and tracking. This paper proposes a scheme for detecting and tracking a person using WSN. In an organization where employees possess unique Radio Frequency Identification (RFID) tags, a WSN node detects the presence of a person using a PIR sensor and the identity of the person is obtained from the RFID tag. The accuracy of tracking a person in a WSN is limited due to the sensor's detection capabilities. Also, the sensor packets may be lost in the wireless medium. Moreover, the location information in the sensor is not accurate indoors. This paper proposes a multi-sensor and Kalman Filter (KF) based tracking scheme in a WSN oblivious to localization errors, errors due to missing events caused by failure of nodes etc. The accuracy of the proposed tracking has been validated for different scenarios.