资 源 简 介
Naidu的这本《Sensor Array Signal Processing》是学习阵列信号处理的必看书籍之一，内容生动，属雷达、声呐、地震传播学等学科精品SENSORARBAY SIGNALPROCESSINGPrabhakar s aiduCRC PressBoca Raton London New York Washington, D. C.1195/Disclaimer Page I Monday, June 5. 2000 3: 20 PMLibrary of Congress Cataloging- in-Publication DataNaidu. Prabhakar sSensor array signal processing /Prabhakar S NaiduIncludes bibliographical references and indexISBN0-8493-1195-0(alk.pa1. Singal proccssing-Digital tcchniqucs. 2. Multisensor data fusionTitleTK51029N3520006213822dc2100-030409CIPThis book contains information obtained from authentic and highly regarded sources. Reprinted materialis quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonableefforts have been made to publish reliable data and information, but the author and the publisher cannotassume responsibility for the validity of all materials or for the consequences of their useNeither this book nor any part may be reproduced or transmitted in any form or by any means, electronicor mechanical, including photocopying, microfilming, and recording, or by any information storage orretrieval system, without prior permission in writing from the publisherThe consent of CRC Press llC does not extend to copying for general distribution, for promotion, focreating new works, or for resale. Specilic permission Imust be oblained in writing Iron CrC PreSs llCfor such copying.Dircct all inquiries to CRC Prcss LLC, 2000 N.w. Corporatc Blvd., Boca Raton, Florida 33431Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and areused only for identification and explanation, without intent to infringeo 2001 by CRC Press LLCNo claim to original U.S. Government worksInternational Standard Book Number 0-8493-1195-0Library of Congress Card Number 00-030409Printed in the United states of america 1234567890Printed on acid-free paperPrologueAn array of sensors is often used in many diverse fields of science andengineering, particularly where the goal is to study propagating wavefieldSome examples are astronomy (radio astronomy), medical diagnosis, radar,communication, sonar, nonrestrictive testing, seismology, and seismicexploration(see [1] for different applications of the array signal pirocessing)The main goal of array signal processing is to deduce the followinginformation through an analysis of wavefields(a) Source localization as in radar, sonar, astronomy, and seismology, etc(b) Source waveform estimation as in communication, etc(c) Source characterization as in seismology(d)Imaging of the scattering mediun as in nedical diagnosis, seisImicexploration, etcThe tools of array signal processing remain the same, cutting across theboundaries of different disciplines. For example, the basic tool ofbeamformation is used in many areas mentioned above. The present book aimsat unraveling the underlying basic principles of array signal processing withouta reference to any particular application. However, an attempt is made toinclude as many tools as possible from different disciplines in an order whichreflects the underlying principleIn the real world different types of wavefields are used in differentapplications, for example, acoustic waves in sonar, mechanical waves inseismic exploration, electromagnetic waves in radar and radio astronomyFortunately, all wavefields can be characterized under identical mathematicalframework. This common mathematical framework is briefly summarized inchapter 1. Here we have described the basic equations underlying differentwavefields and the structure of array signals and the background noise when thenoise sources follow some simple geometrical distribution. The topics coveredare wavefield in open space, bounded space including multipath propagation andlayered medium. Also covered is the weak scattering phenomenon which is thebasis for tomographic imaging. In chapter 2 we study different types of sensorconfigurations. The emphasis is however on commonly used uniform lineararray (ULA), uniform circular array (UCA). Many practical sensor arraysystems can be studied in terms of the basic ULA and UCA SysteIns(cylindrical array in radar and sonar, cross array in astronomy and seismology)Like sensors, the sources can also be configured in the form of an array TheSource array is useful in synthesizing a desired wavefront and/or waveform. Inchapter 3 we examine the issues connected with the design of 2d digital filtersfor wavefield analysis. Since the propagating wavefields possess someinteresting spectral characteristics in frequency wavenumber domain, forexample, the spectrum of a propagating wavefront is al ways on a radial line, itis natural to take into account these features in the design of digital filters forseparation of interfering wavefields. Specifically, we cover in detail the designof a fan filter and quadrant filter. Also, the classical wiener filter as anoptimum least squares filter is covered in this chapterThe theme in chapters 4 and 5 is localization of a source. In chapter 4we describe the classical methods based on the frequency wavenumber spectrumof the observed array output. We start with the Blackman Tukey type frequencywavenumber spectrum and then go on to modern nonlinear high resolutionspectrum analysis methods such as Capons maximum likelihood spectrumwhich is also known as minimum variance distortionless response (MVDR)beamformer and maximum entropy spectrum localization essentially involvesestimation of parameters pertaining to the source position, for exampleazimuth and elevation angles, range, speed if the source is moving, etc In thelast two decades a host of new methods of source localization have beeninvented. We elaborate these new approaches in chapter 5. These includesubspace based methods use of man-made signals such as in communicationand finally multipath environment. Quite often localization must be done in thereal time and it may be necessary to track a moving source. Adaptive techniquesare best suited for such tasks. a brief discussion on adaptive approach isincluded In chapter 6 we look into methods for source waveform separation andestimation. The direction of arrival (DOa)is assumed to be known or has beenestimated. We shall describe a Wiener filter which minimizes the mean squareerror in the estimation of the desired signal coming from a known direction anda Capon filter which, while minimizing the power, ensures that the desiredsignal is not distorted We also talk about the estimation of direction of arrivalin a multipath environment encountered in wireless communicationThe next two chapters are devoted to array processing for imagingpurposes. Firstly, in chapter 7 we look at different types of tomographicimaging systeIns: nondiffracting, diffracting and reflection Loinography. Thereceived wavefield is inverted under the assumption of weak scattering to mapany one or more physical properties of the medium, for example, sound speedvariations in a medium. For objects of regular shape, scattering points play animportant role in geometrical diffraction theory. Estimation of these scatteringpoints for the determination of shape is also discussed. In chapter 8 we studthe method of wavefield extrapolation for imaging, extensively used in seismicexploration. The raw scismic traces are stacked in order to produce an outputtrace from a hypothetical sensor kept close to the source(with zero-offset).asuite of such stacked traces may be modeled as a wavefield recorded in animaginary experiment wherein small charges are placed on the reflector andexploded at the same time. The zero-offset wavefield is used for imaging ofreflectors. The imaging process may be looked upon as a downwardcontinuation of the wavefield or inverse source problem or propagationbackward in time, i.e., depropagation to the reflector. all three view points arevery briefly describedThe book is based on a course entitled "Digital Array Processingoffered to the graduate students who had already taken a course on digital signalprocessing (DSP)and a course on modern spectrum analysis (MSA). It hasbeen my conviction that a student should be exposed to all basic conceptscutting across the different disciplines without being burdened with thequestions of practical applications which are usually dealt with in speciallycourses. The most satisfying experience is that there is a common thread thatconnects seemingly different tools used in different disciplines. An example isbeamformation, a commonly used tool in radar/sonar, which has a closesimilarity with stacking used in seismic exploration. I have tried to bring outin this exposition the common thread that exists in the analysis of wavefieldused in a wide variety of application areas. The proposed book has asignificantly different flavor, both in coverage and depth in comparison with theones on the market [1-5. The first book, edited by Haykin, is a collection ofchapters, each devoted to an application. It rapidly surveys the state of art inrespective application areas but does not go deep enough and describe the basicmathematical theory required for the understanding of array processing. Thesecond book by Ziomek is entirely devoted to array signal processing inunderwater acoustics. It covers in great depth the topic of beamformation bylinear and planar arrays but confines to linear methods. Modern array processingtools do not find a place in this book. The third book by pillai [3] has a verynarrow scope as it deals with in great detail only the subspace based methodsThe fourth book by bouvet and Bienvenu(Eds)is again a collection of paperslargely devoted to modern subspace techniques. It is not suitable as a textFinally, the present book has some similarities with a book by Johnson andDudgeon [3 but differs in one important respect, namely, it does not cover theapplication of arrays to imaging though a brief mention of tomography ismade. Also, the present book covers newer material which was not available atthe time of the publication of the book by Johnson and Dudgeon During thelast two decades there has been intense research activity in the area of arraysignal processing. There have been at least two review papers summarizing thenew results oblained during this period. The present book is not a researchmonograph but it is an advanced level text which focuses on the importantdevelopments which, the author believes, should be taught to give a broadpicture of array signal processinghave adopted the following plan of teaching as the entire bookcannot be covered in one semester(about 35 hours)I preferred to cover it in twoparts in alternate semesters. In the first part, I covered chapter 1(exclude $1.6)chapter 2, chapters 4, 5 and 6. In the second part, I covered chapter 1, chapter 2(exclude 52.3), chapter 3(exclude 83.5), chapters 7 and 8. Exercises are given atthe end of each chapter. The solution guide may be obtained from thepublisher).1.S Haykin(Ed), Array Signal Processing, Prentice Hall, Englewood CliffsNJ.1985.2. L.J. 7iomek, Underwater Acoustics, A Linear Systems Theory, AcademicPress, Orlando. 19853. S.U. Pillai, Array Signal Processing, Springer-Verlag, New York, 1989.4. M. Bouvet and G. Bienvenu, High resolution Methods in underwaterAcoustics, Springer-Verlag, Berlin, 1995. D.H. Johnson and D. E Dudgeon, Array Signal Processing, Prentice HallEnglewood CliffS.NJ. 19936 H. Krim and m. Viberg, Two decades of array signl processing, IEEE SignalProc. Mag. July 1996, pp. 67-947. T. Chen(Ed) Highlights of statistical signal and array processing, IEEESignal Proc. Mag. pp 21-64, Sept. 1998Prabhakar s naiduFebruary, 2000Prot, Dept of eCeIndian Institute of scienceBangalore 560012. India.Sensor Array Signal ProcessingContentsChapterAn overview of wavefields1. 1 Types of wavefields and the governing equaLions1. 2 Wavefield in open space3 Wavefield in bounded space1. 4 Stochastic wavefield5 Multipath propagation1.6 Propagation through random medium1. 7 ExercisesChapter TwSensor Array Systems2.1 Uniform linear array qULa)2.2 Planar array2.3 Broadband sensor array2.4 Source and sensor arrays2.5 ExercisesChapter ThreeFrequency Wavenumber Processing3.1 Digital filters in the O-k domain3.2 Mapping of ID into 2D filters3 Multichannel Wiener filters3. 4 Wiener filters for ULA and UCa3.5 Predictive noise cancellation3.6 ExercisesChapter FourSource Localization: frequency wavenumber spectrum4. 1 Frequency wavenumber spectrum4.2 Beamformation4.3 Capons O-k spectrum4. 4 Maxiinuin entropy (o-k spectrum4.5 Exercisespler FiveSource Localization: Subspace Methods5. I Subspace methods(Narrow band)5.2 Subspace methods Broadband3 Coded5. 4 Array calibration5.5 Source in bounded space5.6 Exerciseshapter SixSource estimationers6.2 Minimum variance( Capon method)6.3 Adaptive beamformation6. 4 Beamformation with coded signals6.5 Multipath channel6.6 ExercisesTomographic Imaging7. 1 Nondiffracting radiation7. 2 Diffracting radiation7. 3 Broadband illumination7.4 Reflection tomograph7.5 Object shape estimation7.6 ExercisesChapter eightImaging by Wavefield Extrapolation8.1 Migration8.2 Exploding reflector model8.3 Extrapolation in a-k plane8.4 Focused beam8.5 Estimation of wave speed8.6 Exercises
