Design and implementation of an ultrasonic localization system for wireless sensor networks using angle-of-arrival and distance measurement

Abstract

This paper presents a localization system for Wireless Sensor Networks (WSN) based on ultrasonic (US) Time-of-Flight (ToF) measurements. The participants send out US pulses while a central localization unit measures the Time-Difference-of-Arrival (TDoA) between four US sensors to calculate the Angle-of-Arrival (AoA). The radio frequency (RF) transceiver of the sensor nodes enables distance measurements using TDoA (US vs. RF) in addition. This improves the localization accuracy significantly since the estimated distance from triangulation suffers excessively from even small angle errors. Several filter stages including Kalman-filtering minimize the number of outliers and fluctuations of the calculated distances and angles. Those computed polar coordinates (angle / distance) are converted into a Cartesian form and forwarded to a base station which is connected to a PC. The mean error and standard deviation of the angle and distance measurements (1.36 ° ± 0.39 ° / 1.00 cm ± 0.14 cm) lead to a small mean localization error of 4.21 cm and a standard deviation of 0.57 cm. © 2012 The Authors. Published by Elsevier Ltd.