Efficient convex optimization method for underwater passive source localization based on RSS with WSN

• Published in: 2012 IEEE International Conference on Signal Processing, Communication and Computing (ICSPCC) pp. 171 - 174
• Main Authors: Yong-sheng Yan, Hai-yan Wang, Xiao-hong Shen, Fu-zhou Yang, Zhao Chen
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.08.2012
• Subjects: Acoustics; Convex functions; convex optimization; least-square; Maximum likelihood estimation; passive source localization; Signal processing algorithms; Signal to noise ratio; underwater wireless sensor networks; Wireless sensor networks
• DOI: 10.1109/ICSPCC.2012.6335673
• ISBN: 9781467321921; 1467321923

The widespread applications of wireless sensor network (WSN) and the advancement of the micro-electro-mechanical systems (MEMS) technology, and wireless communication develop the underwater wireless sensor network (UWSN). The passive source localization application which is an important application in underwater signal processing is focused on in this paper. We model that the underwater source radiates acoustic noise or energy isotropically and utilize the receiving signal energy (RSS) taken at individual sensors of the UWSN to estimate passively the location of the source. Firstly, we choose the node whose energy is maximal as the reference node to decrease the computational complexity compared to our previous work which use division between pairs of sensor energy output. Secondly, the one-step least-square (OS) method is reviewed and another description based on maximum likelihood source location estimator (MLE) is given. Thirdly, a semidefinite programming (SDP) method is developed to covert the nonconvex problem into the convex optimization problem (CVX). Compared to the least-square (LS) method, this CVX-SDP method based on RSS achieves more accurate results and has better robustness with less number of sensor nodes and with lower SNR.