Near-Field Source Localization by Using Focusing Technique
- PDF / 607,968 Bytes
- 9 Pages / 600.05 x 792 pts Page_size
- 2 Downloads / 202 Views
Research Article Near-Field Source Localization by Using Focusing Technique Hongyang He, Yide Wang, and Joseph Saillard Ecole Polytechnique de l’Universit´e de Nantes, Institut de Recherche en Electrotechnique et Electronique de Nantes Atlantique (IREENA), Rue Christian Pauc, BP 50609, 44306 Nantes Cedex 3, France Correspondence should be addressed to Hongyang He, [email protected] Received 2 June 2008; Revised 30 September 2008; Accepted 7 November 2008 Recommended by M. Greco We discuss two fast algorithms to localize multiple sources in near field. The symmetry-based method proposed by Zhi and Chia (2007) is first improved by implementing a search-free procedure for the reduction of computation cost. We present then a focusing-based method which does not require symmetric array configuration. By using focusing technique, the near-field signal model is transformed into a model possessing the same structure as in the far-field situation, which allows the bearing estimation with the well-studied far-field methods. With the estimated bearing, the range estimation of each source is consequently obtained by using 1D MUSIC method without parameter pairing. The performance of the improved symmetry-based method and the proposed focusing-based method is compared by Monte Carlo simulations and with Crammer-Rao bound as well. Unlike other near-field algorithms, these two approaches require neither high-computation cost nor high-order statistics. Copyright © 2008 Hongyang He et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
1.
INTRODUCTION
In the last two decades, subspace-based methods for estimating the directions of arrival (DOA) of far-field sources impinging on an array of sensors have become very popular. In particular, the MUSIC algorithm [1] and its derivatives have received much attention [2]. Most of these methods make the assumption that the sources are located relatively far from the array so that the waves emitted by the sources can be considered as plane waves. With this assumption, each signal wavefront can be characterized by their DOA. However, when a source is located close to the array (i.e., near field), the wavefront must be characterized by both the azimuth and range. Methods based on the farfield assumption are no longer applicable to this situation. The near-field situation can occur, for example, in sonar, electronic surveillance, and seismic exploration. Recently, many localization methods for near-field sources have been proposed, such as maximum likelihood method in [2, 3], the 2D MUSIC method in [2, 4], the modelfitting method in [5], the linear prediction method in [6], the higher-order ESPRIT method in [1, 7], and so forth. Most of these methods either involve multidimensional spectral peak search, or contain an additional parameters pairing procedure, or high-order statistics computation.
In order to reduce
Data Loading...
