Title: A three-dimensional localisation algorithm for underwater acoustic sensor networks

Authors: Dae-Ho Won; Yeon-Mo Yang

Addresses: School of Electronic Engineering, Kumoh National Institute of Technology, 1 Yahoho-Dong, Gumi, Korea. ' School of Electronic Engineering, Kumoh National Institute of Technology, 1 Yahoho-Dong, Gumi, Korea

Abstract: Simultaneous localisation mapping (SLAM) is a scheme for location-aware technology that can be applied in vehicles or autonomous robots in outdoor environments. The SLAM scheme helps in the localisation of AUVs (autonomous underwater vehicles) used in deep ocean exploration. In this paper, we propose a 3D SLAM scheme that utilises an extended Kalman filter (EKF) based on landmarks in underwater sensor networks (UWSNs). We obtained our research results by the following three steps. First, we designed and implemented the AUV model using kinematics. Second, we analysed the EKF for use in SLAM and derived an optimisation model of the EKF using information on a moving path and landmarks in an underwater environment. Finally, we implemented the map through the derived EKF, as applied to SLAM in UWSNs. We have thoroughly compared the value of a pre-defined path to the predicted value and confirmed the average position error in terms of AUV speed, number of landmarks, and the sampling time. Through Matlab simulations, we have shown that the proposed scheme achieves a smaller position error under a change in vehicle speed and number of landmarks. We have also confirmed the minimum performance of the 3D-SLAM technique for use in underwater applications.

Keywords: underwater sensor networks; UWSNs; positioning; simultaneous localisation mapping; SLAM; autonomous underwater vehicles; AUVs; extended Kalman filter; EKF; acoustic sensor networks; deep ocean exploration; simulation.

DOI: 10.1504/IJCVR.2011.042840

International Journal of Computational Vision and Robotics, 2011 Vol.2 No.3, pp.218 - 236

Published online: 06 Oct 2011 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article