Professional Documents
Culture Documents
Inverse Synthetic Aperture Radar Imaging With Matlab Algorithms 2Nd Edition Caner Ozdemir Full Chapter PDF
Inverse Synthetic Aperture Radar Imaging With Matlab Algorithms 2Nd Edition Caner Ozdemir Full Chapter PDF
https://ebookmass.com/product/advanced-algorithms-for-mineral-
and-hydrocarbon-exploration-using-synthetic-aperture-radar-1st-
edition-maged-marghany/
https://ebookmass.com/product/nonlinear-ocean-dynamics-synthetic-
aperture-radar-1st-edition-maged-marghany/
https://ebookmass.com/product/introduction-to-synthetic-aperture-
radar-concepts-and-practice-e-david-jansing/
https://ebookmass.com/product/modern-information-optics-with-
matlab-yaping-zhang/
Specialty Imaging: Arthrography 2nd Edition Julia Crim
Md
https://ebookmass.com/product/specialty-imaging-arthrography-2nd-
edition-julia-crim-md/
https://ebookmass.com/product/fundamental-chemistry-with-
matlab-1st-edition-daniele-mazza/
https://ebookmass.com/product/signal-processing-for-multistatic-
radar-systems-adaptive-waveform-selection-optimal-geometries-and-
pseudolinear-tracking-algorithms-dogancay/
https://ebookmass.com/product/data-structures-and-algorithms-
in-c-2nd-edition/
https://ebookmass.com/product/cryptography-algorithms-2nd-
edition-early-release-massimo-bertaccini/
Inverse Synthetic Aperture
Radar Imaging with
MATLAB Algorithms
Wiley Series in Microwave and Optical Engineering
Kai Chang, Editor
Texas A&M University
A complete list of the titles in this series appears at the end of this volume.
Inverse Synthetic Aperture Radar Imaging
with MATLAB Algorithms
Second Edition
10 9 8 7 6 5 4 3 2 1
To:
My wife Betül,
My three daughters,
My brother,
My father,
and the memory of my beloved mother
vii
Contents
2 Radar Fundamentals 35
2.1 Electromagnetic Scattering 35
2.2 Scattering from PECs 38
2.3 Radar Cross Section 39
2.3.1 Definition of RCS 40
2.3.2 RCS of Simple-Shaped Objects 43
2.3.3 RCS of Complex-Shaped Objects 44
2.4 Radar Range Equation 44
2.4.1 Bistatic Case 46
2.4.2 Monostatic Case 49
2.5 Range of Radar Detection 50
2.5.1 Signal-to-Noise Ratio 51
2.6 Radar Waveforms 53
2.6.1 Continuous Wave 53
2.6.2 Frequency-Modulated Continuous Wave 56
2.6.3 Stepped-Frequency Continuous Wave 59
2.6.4 Short Pulse 61
2.6.5 Chirp (LFM) Pulse 62
2.7 Pulsed Radar 69
2.7.1 Pulse Repetition Frequency 69
2.7.2 Maximum Range and Range Ambiguity 69
2.7.3 Doppler Frequency 70
2.8 Matlab Codes 74
References 82
Contents ix
4 Inverse Synthetic Aperture Radar Imaging and Its Basic Concepts 162
4.1 SAR versus ISAR 162
4.2 The Relation of Scattered Field to the Image Function in ISAR 166
4.3 One-Dimensional (1D) Range Profile 167
4.4 1D Cross-Range Profile 172
4.5 Two-Dimensional (2D) ISAR Image Formation (Small Bandwidth, Small
Angle) 176
4.5.1 Resolutions in ISAR 180
4.5.1.1 Range Resolution 181
4.5.1.2 Cross-Range Resolution: 181
4.5.2 Range and Cross-Range Extends 181
4.5.3 Imaging Multibounces in ISAR 182
4.5.4 Sample Design Procedure for ISAR 185
4.5.4.1 ISAR Design Example #1: “Aircraft Target” 189
4.5.4.2 ISAR Design Example #2: “Military Tank Target” 193
4.6 2D ISAR Image Formation (Wide Bandwidth, Large Angles) 197
4.6.1 Direct Integration 198
4.6.2 Polar Reformatting 201
4.7 3D ISAR Image Formation 205
4.7.1 Range and Cross-Range resolutions 209
4.7.2 A Design Example for 3D ISAR 210
4.8 Matlab Codes 217
References 243
Appendix 619
Index 628
xvi
In the first edition of the book, I tried to cover most of the aspects of inverse syn-
thetic aperture radar (ISAR) imaging starting from Fourier analysis to some
advanced ISAR concepts such as range-Doppler ISAR processing and ISAR motion
compensation techniques. The main goal was to present a conceptual description
of ISAR imagery and the explanation of basic ISAR research topics. Although the
primary audience would be graduate students and other interested researchers in
the fields of electrical engineering and physics, I hoped that colleagues working in
radar research and development or in a related industry might also benefit from
the book.
It has been more than eight years since the publication of the first edition. Since
then, I have been really grateful that I have received positive responses from the
researchers and colleagues that are interested and/or involved in radar imaging,
and especially ISAR imaging. Undoubtedly, ISAR has been gaining more attention
among researchers, scholars, and engineers as emerging new developments in
ISAR research have been reported by various colleagues day by day.
In this second edition of the book, I have tried to include the recent progress
made in ISAR imaging research and also give insights to more advanced concepts.
Therefore, in this edition of the book, I have made the following alterations and
additions:
• All the chapters in the first edition have been revised including all the texts,
equations, and figures with some additions. Typos in the first edition have also
been corrected.
• Chapter 3 that is devoted to the issues of synthetic aperture radar (SAR) has been
extended to include the SAR focusing/processing algorithms such as range-
Doppler algorithm (RDA), back-projection algorithm (BPA) and frequency-
wave number algorithm (ω-kA). The Matlab codes for these algorithms are
being provided with the associated numerical examples. Brief explanations of
Preface to the Second Edition xvii
• A total of three new chapters have been written to cover the topics that were not
considered in the first edition and also to include more detailed subjects of ISAR
imaging to be able to reflect the recent research studies. These are listed below:
– The “Bistatic ISAR (Bi-ISAR) Imaging” concept is covered in Chapter 9. While
the ISAR imaging algorithms presented in previous chapters are based on
monostatic usage of ISAR imaging, we introduce the formulation of ISAR
imaging for the bistatic usages by presenting key aspects such as resolutions
in range/cross-range directions and usage limitations. Also, extension of Bi-
ISAR to multistatic ISAR (Mu-ISAR) imaging is derived with the associated
Matlab examples. A general assessment of Bi-ISAR and Mu-ISAR imaging
to conventional monostatic ISAR imaging is being made throughout the chap-
ter by comparing the outcomes of quantitative metrics and giving the conclud-
ing statements about their advantages and disadvantages based on these
measurable evaluations.
– In Chapter 10, we have added a new and exciting research topic of ISAR called
“Polarimetric ISAR Imaging.” As the traditional ISAR imaging algorithms are
based only on a single polarization of the backscattered electric field, we dem-
onstrate in this chapter that very exciting features of the target can be
extracted with the use of other possible polarizations for the reflected wave.
Polarization decomposition techniques are being introduced and Pauli
decomposition scheme is taken as the tool to be applied to the different polar-
ization ISAR images in this book. The formulation and the usage of Pauli
decomposition technique are presented together with its Matlab codes. Var-
ious realistic simulation examples based on linear polarization, circular polar-
ization, and also Pauli decomposition are given together with obtained
polarimetric ISAR images. It has been demonstrated through the examples
that polarized ISAR images definitely increase the recognition and classifica-
tion of targets by providing increased number of extracted target features.
– Thanks to the recent development in the microwave circuit technology and
antenna design, ISAR imaging algorithms have been started to be used in the
near-field region. Therefore, I have added a new part entitled “Near-field
ISAR imaging” as Chapter 11. The near-field ISAR imaging algorithms are
being introduced. Two of them called “Focusing operator” and the back-
projection based focusing algorithms are given by presenting their theoretical
formulation and algorithm steps together with corresponding Matlab codes.
xviii Preface to the Second Edition
Also, numerical and measured examples based on real scenarios are being
shared.
Acknowledgments
I would like to address special thanks to the people below for their help and sup-
port during the preparation of this book. First, I am thankful to my wife, Betül and
my three children for their patience and continuous support while writing this
book. I am very grateful to Dr. Hao Ling, Emeritus Professor in Engineering of
the University of Texas at Austin for being a valuable source of knowledge, ideas,
and also inspiration throughout my academic carreer. He has been a great advisor
since I met him, and his guidance on scientific research is priceless to me.
I would like to express my sincere thanks to my former graduate students; Dr.
Şevket Demirci, Dr. Enes Yiğit, Dr. Betül Yılmaz, Dr. Deniz Üstün, Özkan Kırık,
and Dr. Hakan Işıker who have helped carrying out some of the research presented
in this book. I would also like to thank my graduate student Rasheed Khankan for
his help in preparing references.
Last but not least, I would like to convey my special thanks to Dr. Kai Chang for
inviting me to write the first and then second edition of the book. Without his kind
offer, this book project would not have been possible.
Caner Özdemir
xx
Acronyms
1D One-dimensional
2D Two-dimensional
3D Three-dimensional
ACSAR Antenna coupling synthetic aperture radar
ADC Analog-to-digital converter
ANN Artificial neural network
ASAR Antenna synthetic aperture radar
ATC Automatic target classification
ATR Automatic target recognition
Bi-ISAR Bistatic inverse synthetic aperture radar
BPA Back-projection algorithm
CAD Computer aided design
CDF Cumulative density function
CFAR Constant false alarm rate
CO Contrast optimization
Co-pol Co-polarization
CP Circular polarization
Cross-pol Cross-polarization
CSA Chirp scaling algorithm
CW Continuous wave
DCR Dihedral corner reflectors
DFT Discrete Fourier transform
DTV Digital television
EFIE Electric field integral equation
EM Electromagnetic
ESM Exploding source model
FM Frequency modulated
FMCW Frequency modulated continuous wave
Acronyms xxi
FT Fourier transform
GO Geometric optics
GPR Ground-penetrating radar
GPS Global positioning system
GWN Gaussian white noise
H Horizontal
HH Horizontal–horizontal
HSA Hyperbolic summation algorithm
HV Horizontal–vertical
I Inphase
IDFT Inverse discrete Fourier transform
IFT Inverse Fourier transform
IMU Inertial measurement unit
InSAR Interferometric SAR
ISAR Inverse synthetic aperture radar
JTF Joint time-frequency
KB Kbytes
KMA Kirchhoff migration algorithm
L Left
LFM Linear frequency modulated
LFMCW Linear frequency modulated continuous wave
LHCP Left-hand circular polarized
LHEP Left-hand elliptically polarized
LL Left–left
LOS Line of sight
LP Linear polarization
LR Left–right
MB Mbytes
MDA Map-drift autofocus
MFIE Magnetic field integral equation
MIMO Multiple-input multiple-output
MOCOMP Motion compensation
Mu-ISAR Multi-static inverse synthetic aperture radar
PEC Perfect electric conductor
PGA Phase gradient autofocus
P-ISAR Passive inverse synthetic aperture radar
PO Physical optics
PolSAR Polarimetric synthetic aperture radar
PPP Prominent point processing
PRF Pulse repetition frequency
xxii Acronyms
Fourier transform (FT) is a common and useful mathematical tool that is utilized
in innumerous applications in science and technology. FT is quite practical espe-
cially for characterizing nonlinear functions in nonlinear systems, analyzing ran-
dom signals, and solving linear problems. FT is also a very important tool in radar
imaging applications as we shall investigate in the forthcoming chapters of this
book. Before starting to deal with the FT and inverse Fourier transform (IFT), a
brief history of this useful linear operator, and its founders are presented.
Inverse Synthetic Aperture Radar Imaging with MATLAB Algorithms: With Advanced SAR/ISAR
Imaging Concepts, Algorithms, and MATLAB Codes, Second Edition. Caner Özdemir.
© 2021 John Wiley & Sons, Inc. Published 2021 by John Wiley & Sons, Inc.
2 1 Basics of Fourier Analysis
G f = gt
∞ 11
= gt exp − j2πft dt
−∞
where represents the forward FT operation that is defined from time domain to
frequency domain.
To appreciate the meaning of FT, the multiplying function exp(−j2πft) and
operators (multiplication and integration) on the right of side of Eq. 1.1 should
be examined carefully: The term exp − j2π f i t is a complex phasor representation
for a sinusoidal function with the single frequency of “fi.” This signal oscillates
with the single frequency of “fi” and does not contain any other frequency com-
ponent. Multiplying the signal in interest, g(t) with exp − j2π f i t provides the sim-
ilarity between each signal, that is, how much of g(t) has the frequency content of
“fi.” Integrating this multiplication over all time instants from −∞ to ∞ will sum
the “fi” contents of g(t) over all time instants to give G(fi) that is the amplitude of
the signal at the particular frequency of “fi.” Repeating this process for all the fre-
quencies from −∞ to ∞ will provide the frequency spectrum of the signal repre-
sented as G(f). Therefore, the transformed signal represents the continuous
spectrum of frequency components; i.e. representation of the signal in “frequency
domain.”
1.2 FT Rules and Pairs 3
1.1.3 IFT
This transformation is the inverse operation of the FT. IFT, therefore, synthesizes
a frequency-domain signal from its spectrum of frequency components to its
time domain form. The IFT of a continuous signal G(f) where −∞ < f < ∞ is
described as
−1
gt = G f
∞ 12
= G f exp j2πft df
−∞
where the IFT operation from frequency domain to time domain is represented by
−1
.
There are many useful Fourier rules and pairs that can be very helpful when apply-
ing the FT or IFT to different real-world applications. We will briefly revisit them
to remind the properties of the FT to the reader. Provided that FT and IFT are
defined as in Eqs. 1.1 and 1.2, respectively, FT pair is denoted as
gt G f 13
and the corresponding alternative pair is given by
−1
G f gt 14
Based on these notations, the properties of FT are listed briefly below.
1.2.1 Linearity
If G(f) and H(f) are the FTs of the time signals g(t) and h(t), respectively, the fol-
lowing equation is valid for the scalars a and b.
a g t +b h t a G f +b H f 15
Therefore, the FT is a linear operator.
g t − to exp − j2πf t o G f 16
4 1 Basics of Fourier Analysis
exp j2π f o t gt G f − fo 17
1.2.4 Scaling
If the time signal is scaled by a constant a, then the spectrum is also scaled with the
following rule
1 f
g at G , a R, a 0 18
a a
1.2.5 Duality
If the spectrum signal G(f) is taken as a time signal G(t), then, the corresponding
frequency domain signal will be the time reversal equivalent of the original time
domain signal, g(t) as
Gt g −f 19
g −t G −f 1 10
1.2.7 Conjugation
If the conjugate of the time-domain signal is taken, then the frequency-domain
signal conjugated and frequency-reversed.
g∗ t G∗ − f 1 11
1.2.8 Multiplication
If the time-domain signals, g(t) and h(t) are multiplied in time, then their spectrum
signals G(f) and H(f) are convolved in frequency.
gt ht G f ∗H f 1 12
1.3 Time-Frequency Representation of a Signal 5
1.2.9 Convolution
If the time-domain signals, g(t) and h(t) are convolved in time, then their spectrum
signals G(f) and H(f) are multiplied in the frequency domain.
g t ∗h t G f H f 1 13
1.2.10 Modulation
If the time-domain signal is modulated with sinusoidal functions, then the fre-
quency-domain signal is shifted by the amount of the frequency at that particular
sinusoidal function.
1
gt cos 2π f o t G f + fo + G f − fo
2
1 14
j
gt sin 2π f o t G f + fo −G f − fo
2
While the FT concept can be successfully utilized for the stationary signals, there
are many real-world signals whose frequency contents vary over time. To be able to
display these frequency variations over time; therefore, joint time–frequency (JTF)
transforms/representations are being used.
6 1 Basics of Fourier Analysis
0.2
0.1
0
Amplitude
–0.1
–0.2
–0.3
–0.4
–0.5
0.35
0.025
0.25
Amplitude
0.015
0.15
0.01
0.005
0
0 5 10 15 20
Frequency, kHz
are not considered to be suitable for the detailed analysis of those signals.
Therefore, JTF analysis methods were developed to represent these signals both
in time and frequency to observe the variation of frequency content as the time
progresses.
There are many tools to map a time domain or frequency-domain signal onto
the JFT plane. Some of the most well-known JFT tools are short-time Fourier
transform (STFT) (Allen 1977), Wigner–Ville distribution (Nuttall 1988),
Choi–Willams distribution (Du and Su 2003), Cohen’s class (Cohen 1989),
and time-frequency distribution series (TFDS) (Qian and Chen 1996). Among
these, the most appreciated and commonly used one is the STFT or the spectro-
gram. STFT can easily display the variations in the sinusoidal frequency and
phase content of local moments of a signal over time with sufficient resolution
in most cases.
The spectrogram transforms the signal onto two-dimensional (2D) time-
frequency plane via the following famous equation:
STFT g t ≜ G t, f
∞ 1 17
= gτ w τ − t exp − j2πf τ dτ
−∞
This transformation formula is nothing but the short-time (or short-term) ver-
sion of the famous FT operation defined in Eq. 1.1. The main signal, g(t) is multi-
plied with a shorter duration window signal, w(t). By sliding this window signal
over g(t) and taking the FT of the product, only the frequency content for the win-
dowed version of the original signal is acquired. Therefore, after completing the
sliding process over the whole duration of the time-domain signal g(t) and putting
corresponding FTs side by side, the 2D STFT of g(t) is obtained.
It is obvious that STFT will produce different output signals for different dura-
tion windows. The duration of the window affects the resolutions in both domains.
While a very short-duration time window provides a good resolution in the time
direction, the resolution in the frequency direction becomes poor. This is because
of the fact that the time duration and the frequency bandwidth of a signal are
inversely proportional to each other. Similarly, a long duration time signal will
give a good resolution in frequency domain while the resolution in the time
domain will be bad. Therefore, a reasonable selection has to be bargained about
the duration of the window in time to be able to view both domains with fairly
good enough resolutions.
The shape of the window function has an effect on the resolutions as well. If a
window with sharp ends is chosen, there will be strong side lobes in the other
domain. Therefore, smooth waveform type windows are usually utilized to obtain
well-resolved images with less side lobes with the price of increased main
1.3 Time-Frequency Representation of a Signal 9
14
12
Frequency, kHz
10
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time, s
beamwidth, i.e. less resolution. Commonly used window types are Hanning, Ham-
ming, Kaiser, Blackman, and Gaussian.
An example of the use of spectrogram is demonstrated in Figure 1.3. The spec-
trogram of the sound signal in Figure 1.1 is obtained by applying the STFT oper-
ation with a Hanning window. This JFT representation obviously demonstrates
the frequency content of different syllables when the word “prince” is spoken.
Figure 1.3 illustrates that while the frequency content of the part “prin…” takes
place at low frequencies, that of the part “..ce” occurs at much higher frequencies.
JTF transformation tools have been found to be very useful in interpreting the
physical mechanisms such as scattering and resonance for radar applications
(Trintinalia and Ling 1995; Filindras et al. 1996; Özdemir and Ling 1997; Chen
and Ling 2002). In particular, when JTF transforms are used to form the 2D image
of electromagnetic scattering from various structures, many useful physical fea-
tures can be displayed. Distinct time events (such as scattering from point targets
or specular points) show up as vertical line in the JTF plane as depicted in
Figure 1.4a. Therefore, these scattering centers appear at only one time instant
but for all frequencies. A resonance behavior such as scattering from an open cav-
ity structure shows up as horizontal line on the JTF plane. Such mechanisms occur
only at discrete frequencies but over all time instants (see Figure 1.4b). Dispersive
mechanisms, on the other hand, are represented on the JTF plane as slanted
curves. If the dispersion is due to the material, then the slope of the image is
10 1 Basics of Fourier Analysis
b
Frequency
e
d
f
c
a
Time
Figure 1.4 Images of scattering mechanisms in the joint time–frequency plane. (a)
Scattering center, (b) resonance, (c and d) dispersion due to material, (e and f ) dispersion due
to geometry of the structure.
positive as shown in Figure 1.4c,d. The dielectric coated structures are the good
examples of this type of dispersion. The reason for having a slanted line is because
of the modes excited inside such materials. As frequency increases, the wave veloc-
ity changes for different modes inside these materials. Consequently, these modes
show up as slanted curves in the JTF plane. Finally, if the dispersion is due to the
geometry of the structure, this type of mechanism appears as a slanted line with a
negative slope. This style of behavior occurs for such structures such as waveguides
where there exist different modes with different wave velocities as the frequency
changes as seen in Figure 1.4e,f.
An example of the use of JTF processing in radar application is shown in
Figure 1.5 where spectrogram of the simulated backscattered data from a dielec-
tric-coated wire antenna is shown (Özdemir and Ling 1997). The backscattered
field is collected from the Teflon-coated wire (εr = 2.1) such that the tip of the
electric field makes an angle of 60 with the wire axis as illustrated in Figure 1.5.
After the incident field hits the wire, infinitely successive scattering mechan-
isms occur. The first four of them are illustrated on top of Figure 1.5. The first
return comes from the near tip of the wire. This event occurs at a discrete time
that exists at all frequencies. Therefore, this return demonstrates a scattering
center-type mechanism. On the other hand, all other returns experience at least
one trip along the dielectric-coated wire. Therefore, they confront a dispersive
behavior. As the wave travels along the dielectric-coated wire, it is influenced by
the dominant dispersive surface mode called Goubau (Richmond and Newman
1976). Therefore, the wave velocity decreases as the frequency increases such
that the dispersive returns are tilted to later times on the JTF plane. The dom-
inant dispersive scattering mechanisms numbered 2, 3, and 4 are illustrated in
Figure 1.5 where the spectrogram of the backscattered field is presented. The
other dispersive returns with decreasing energy levels can also be easily
1.4 Convolution and Multiplication Using FT 11
3
30°
1
1 2 3 4
0
16
–10
Frequency, GHz
14
12 –20
10
–30
8
–40
0 5 10 15 dB
Time, ns
Scattering Dispersive
center returns
Figure 1.5 JTF image of a backscattered measured data from a dielectric-coated wire
antenna using spectrogram.
observed from the spectrogram plot. As the wave travels on the dielectric-coated
wire more and more, it is slanted more on the JTF plane, as expected.
Convolution and multiplication of signals are often used in radar signal proces-
sing. As listed in Eqs. 1.12 and 1.13, convolution is the inverse operation of mul-
tiplication as the FT is concerned, and vice versa. This useful feature of the FT is
widely used in signal and image processing applications. It is obvious that the mul-
tiplication operation is significantly faster and easier to deal with when compared
12 1 Basics of Fourier Analysis
to the convolution operation, especially for long signals. Instead of directly convol-
ving two signals in the time domain, therefore, it is much easier and faster to take
the IFT of the multiplication of the spectrums of those signals as shown below:
−1
g t ∗h t = gt ht
1 18
−1
= G f H f
1.5 Filtering/Windowing
Filtering is the common procedure that is used to remove undesired parts of sig-
nals such as noise. It is also used to extract some useful features of the signals. The
filtering function is usually in the form of a window in the frequency domain.
Depending on the frequency inclusion of the window in the frequency axis, the
filters are named low-pass (LP), high-pass (HP), or band-pass (BP).
The frequency characteristics of an ideal LP filter are depicted as dashed line in
Figure 1.6. Ideally, this filter should pass frequencies from DC to the cut-off fre-
quency; fc and should stop higher frequencies beyond. In real practice, however,
ideal LP filter characteristics cannot be realized. According to the Fourier theory, a
signal cannot be both time limited and band limited. That is to say, to be able to
Filter
amplitude
Ideal
0 dB
Real
–3 dB
Pass-band Stop-band
fR fC Frequency
0.8
0.6
Amplitude
Kaiser
0.4
Hanning
Hamming
0.2
0
0 20 40 60 80 100
Frequency
(a) (b)
1
0.3
0.8
0.6 0.2
0.4
0.1
0.2
0 0
–50 0 50 –40 –20 0 20 40
Time, s Frequency, Hz
(c) (d)
1
0.3
0.8
0.6 0.2
0.4
0.1
0.2
0 0
–50 0 50 –40 –20 0 20 40
Time, s Frequency, Hz
Figure 1.8 Effect of windowing. (a) Rectangular time signal, (b) its Fourier spectrum: a sinc
signal, (c) Hanning windowed time signal, (d) corresponding frequency-domain signal.
corresponding spectrum is depicted in Figure 1.8d where the side lobes are highly
suppressed thanks to the windowing operation. For this example, the highest side
lobe level is now 32 dB below the maximum value of the main lobe which pro-
vides better contrast when compared to the original, non-windowed signal.
A main drawback of windowing is the resolution decline in the frequency signal.
The FT of the windowed signal has worse resolution than the FT of the original
time-domain signal. This feature can also be noticed from the example in
Figure 1.8. By comparing the main lobes of the figures on the right, the resolution
after windowing is almost twice as bad when compared to the original frequency-
domain signal. A comprehensive examination of windowing procedure will be
presented later on, in Chapter 5.
(b)
s[n]
Sampling
instants
0Ts Ts 2Ts 3Ts 4Ts . . . . . t
0 1 2 3 4 ... . . n
Sampling
numbers
converted to the digital versions. This is achieved by applying the common proce-
dure of sampling. Analog-to-digital (A/D) converters are common electronic
devices to accomplish this process. The implementation of a typical sampling proc-
ess is shown in Figure 1.9. A time signal s(t) is sampled at every Ts seconds such
that the discrete signal, s[n], is generated via the following equation:
s n = s nT s , n = 0, 1, 2, 3, … 1 20
Therefore, the sampling frequency fs is equal to 1/Ts where Ts is called the sam-
pling interval.
A sampled signal can also be regarded as the digitized version of the multipli-
cation of the continuous signal, s(t) with the impulse comb waveform, c(t) as
depicted in Figure 1.10.
According to the Nyquist–Shannon sampling theorem, the perfect reconstruc-
tion of the signal is only possible provided that the sampling frequency fs is equal
or larger than twice the maximum frequency content of the sampled signal
... ...
1.7.1 DFT
As explained in Section 1.1, the FT is used to transform continuous signals from
one domain to another. It is usually used to describe the continuous spectrum of an
aperiodic time signal. To be able to utilize the FT while working with digital sig-
nals, the digital or DFT has to be used.
Let s(t) be a continuous periodic time signal with a period of To = 1/fo. Then, its
sampled (or discrete) version is s[n] ≜ s(nTs) with a period of NTs = To where N is
the number of samples in one period. Then, the Fourier integral in Eq. 1.1 will turn
to a summation as shown below.
N −1
S kfo = s nT s exp − j2π k f o nT s
n=0
N −1
k
= s nT s exp − j2π nT s 1 21
n=0
NT s
N −1
k
= s nT s exp − j2π n
n=0
N
Dropping the fo and Ts inside the parenthesis for the simplicity of nomenclature
and therefore switching to discrete notation, DFT of the discrete signal s[n] can be
written as
N −1
k
Sk = s n exp − j2π n 1 22
n=0
N
In a dual manner, let S(f) represent a continuous periodic frequency signal with
a period of Nfo = N/To and let S[k] ≜ S(kfo) be the sampled signal with the period of
Nfo = fs. Then, the IDFT of the frequency signal S[k] is given by
N −1
n n
s = S kfo exp j2π k f o
fs k=0
fs
N −1
n
s nT s = S kfo exp j2π k f o 1 23
k=0
Nf o
N −1
n
= S kfo exp j2π k
k=0
N
1.7 DFT and FFT 17
Using the discrete notation by dropping the fo and Ts inside the parenthesis, the
IDFT of a discrete frequency signal S[k] is given as
N −1
n
sn = Sk exp j2π k 1 24
k=0
N
1.7.2 FFT
FFT is the efficient and fast way of evaluating the DFT of a signal. Normally, com-
puting the DFT is in the order of N2 arithmetic operations. On the other hand, fast
algorithms like Cooley-Tukey’s FFT technique produce arithmetic operations in
the order of N log(N) (Cooley and Tukey 1965; Brenner and Rader 1976; Duhamel
1990). An example of DFT is given is Figure 1.11 where a discrete time-domain
ramp signal is plotted in Figure 1.11a and its frequency-domain signal obtained
by an FFT algorithm is given in Figure 1.11b.
B = N Δf
N
= 1 26
To
1
=
Ts
For the example in Figure 1.11, the signal duration is 1 ms with N = 10 samples.
Therefore, the sampling interval is 0.1 ms. After applying the expressions in
Eqs. 1.25 and 1.26, the frequency resolution is 100 Hz and the frequency band-
width is 1000 Hz. After taking the DFT of the discrete time-domain signal, the first
entry of the discrete frequency signal corresponds to zero frequency and negative
frequencies are located in the second half of the discrete frequency signal as seen in
Figure 1.11b. After the DFT operation, therefore, the entries of the discrete
18 1 Basics of Fourier Analysis
0.6
0.4
0.2
0
0 2 4 6 8 10
Time, ms
(b)
6
4
s[k]
0
0 200 400 600 800 1000
Frequency, Hz
(c)
6
4
s[k]
0
–500 0 500
Frequency, Hz
1.9 Importance of FT in Radar Imaging 19
frequency signal should be swapped from the middle to be able to form the fre-
quency axis correctly as shown in Figure 1.11c. This property of DFT will be thor-
oughly explored in Chapter 5 to demonstrate its use in ISAR imaging.
Similar arguments can be made for the case of IDFT. Considering a discrete fre-
quency-domain signal with a bandwidth of B sampled N times with a sampling
interval of Δf, the time resolution (or the sampling interval in time) after applying
IDFT can be found as
Δt = T s
1
= 1 27
B
1
=
NΔf
The time duration of the discrete time signal is
1
To = 1 28
Δf
For the frequency-domain signal in Figure 1.11b or c, the frequency bandwidth
is 1000 Hz with N = 10 samples. Therefore, the sampling interval in frequency is
100 Hz. After applying IDFT to get the time-domain signal as in Figure 1.11a, the
formulas in Eqs. 1.27 and 1.28 calculate the resolution in time as 0.1 ms and the
duration of the signal as 1 ms.
1.8 Aliasing
The imaging of a target using electromagnetic waves emitted from radars is mainly
based on the phase information of the scattered waves from the target. This is
20 1 Basics of Fourier Analysis
because of the fact that the phase information is directly related to the range dis-
tance of the target. In the case of monostatic radar configuration as shown in
Figure 1.12a, let the scattering center on the target be at R distance away from
the radar. Then, the scattered field E s from this scattering center on the target
has a complex scattering amplitude, A and a phase factor that contains the distance
information of the target as follows:
Es A exp − j2kR 1 30
As is obvious from Eq. 1.30, there exists a Fourier relationship between the wave
number, k, and the distance, R. Provided that the scattered field is collected over a
bandwidth of frequencies (Figure 1.12b), it is possible to pinpoint the distance R by
Radar
(b)
Es
fmin fmax f
(c)
Range
profile
Rmin Rmax
1.9 Importance of FT in Radar Imaging 21
Fourier transforming the scattered field data as depicted in Figure 1.12c. The plot
of scattered field versus range is called the range profile, which is an important phe-
nomenon in radar imaging. Range profile is, in fact, nothing but the one-
dimensional range image of the target. An example is illustrated in Figure 1.13
× 10–3
1
0.8
0.6
0.4
0.2
0
0 2 4 6 8 10 12
Range, m
Another random document with
no related content on Scribd:
were built there in his time, and during his reign and for many a long year
after it was the custom to hire merchant vessels and arm them—if they
were not already armed to protect themselves against pirates—to
augment the national fleet. Religious, no less than national, rivalry
contributed, albeit unconsciously, to the development of the efficacy of the
warship as a fighting unit. The enmity between Britain and Spain
culminated, in Elizabeth’s reign, after a series of daring attacks by
reckless Englishmen upon the Spanish fleet in preparation for the great
attack upon England, in the dispatch of the Armada. Hawkins and one or
two others foresaw that the advantage would lie with the fleet which could
be most effectively manœuvred. The disparity between the Armada and
the British fleet was not so great as many writers have represented, either
in the size or number of the vessels; but the British vessels on the
average were smaller, faster, and better handled; in other words,
efficiency told against sheer weight of numbers. This was the last great
sea-fight on the ocean in which oared ships took part; they were no match
for their smaller and more speedy sailing antagonists.
Structurally, most of the vessels of this time, the larger especially, were
disfigured by high sterncastles, but early in the seventeenth century this
encumbrance and many others had disappeared. Thence to the
nineteenth century the development of warships was marked mainly by
continual increases in their size, improving their form of hull and,
consequently, their speed and buoyancy; augmenting their sail area and
perfecting the square-rigged system; and adding to the number of gun
decks and the number of guns carried; until the grand wooden three-
deckers swept the seas in all their ponderous pride and majesty. Ships of
the line of various ratings played their part, and were ably seconded by
frigates, brigs, cutters, sloops and bomb-ketches. All these were in vogue
less than a century ago, and though not forgotten, are looked upon as
historical and romantic and interesting curiosities.
* * * * * *
In the weapons, no less than in the ships, the changes have been
marvellous. For many centuries after ships were adopted for war, the
fighting was done by soldiers carried aboard them. The human machines,
the rowers, had to attend exclusively to their oars, for on them the safety
or success of the fighting men depended. The main idea was to get to
close quarters and fight hand to hand with javelin or sword, spear or
battleaxe; bows and arrows were used when possible, and missiles
hurled by hand were not despised. The ram, in various forms, affixed to
the bows in such a manner as to strike the enemy’s ship below or above
the water-line, or both, was used with fearful effect in many a stubbornly
fought engagement.
The introduction of artillery in the fourteenth century marked the beginning
of the first great revolution in naval warfare, and the changes in the
projectiles have been no less extraordinary than those in the guns.
* * * * * *
The next great revolution was the introduction of the steam-engine. Its
adoption in the British Navy in 1832 marked the beginning of the end of
the sailing warship. Her last grim battle against inexorable fate was fought
with the same doggedness which had distinguished her in many an
encounter with her nation’s enemies; but the superiority of steam over sail
was recognised. Temporising measures, a patched-up peace, as it were,
lasted for a few years while the steam engine was employed as an
auxiliary. Sail power, however, had reached its apotheosis so far as
warships were concerned. Engineers, animated by practical common
sense and ignoring romantic associations, improved their engines, so that
the steam power was no longer the assistant of sail, but its associate, and
was quick in attaining the position of chief partner and showing that sails
could be dispensed with altogether. The Crimean War sounded the knell
of the wooden battleship as well as of the paddle-wheel war steamer. The
former gave place to the iron-clad vessel, and the latter was supplanted
by the screw-propelled ship. The power of artillery had shared in the
application of scientific knowledge and benefited accordingly. The great
battle between the maker of armour plates and the maker of guns and
projectiles had begun. Iron, the conqueror of wood, had but a short reign.
Where iron was used a few years ago, steel is now invariably employed.
The thoroughness of the victory is shown by the fact that in the whole of
the British Isles not one iron vessel, large or small, was built in 1909 for
war or commerce.
* * * * * *
The years 1905 and 1906 saw two of the most important steps forward in
the history of warships, for they included the adoption of the turbine
principle of warship propulsion and the “Dreadnought” principle of
armament. The progress of the last fifty years, culminating in the
Dreadnoughts, has been wonderful; already designs of vessels intended
to relegate them to second place are under consideration.
Type after type of battleship, cruiser, scout, gunboat, destroyer and
torpedo-boat has followed in rapid succession of late years, and
submarines have become an accomplished fact. He would be a foolish
man who would prophesy that the end is in sight.
There is nothing more marvellous in the world’s history than the
tremendous development in marine engineering, in warship construction,
in explosives, in armament, and in projectiles that has taken place in the
latter half of the nineteenth century, and especially in the last twenty-five
years.
WARSHIPS AND THEIR STORY
CHAPTER I
FROM ANCIENT EGYPT TO THE INTRODUCTION OF ARTILLERY
When did man first entrust himself afloat for purposes of war? and
what was the type of vessel he employed? are questions which take
us back almost to the earliest stages of historical human progress,
concerning which all the knowledge of the antiquaries is but
conjectural, a stage so remote that scientists have not yet
determined how many thousands of years ago it existed. The earliest
vessels thus employed must have been transports, and nothing else;
but if employed as aids to aggression when the kings of the earth
took counsel together and, impelled by avarice or a desire to assist
in one another’s turbulent love affairs, or, for their own safety,
convinced of the necessity of finding an outlet for the energies of
their restless subjects, invaded the territories of their neighbours, the
ships, whatever their nature, will have been of a size sufficient to
receive any spoil or any prisoners worth the trouble of carrying back
again.
So far, however, as research has disclosed in those parts of the Near
East where civilisation was cradled, there is no indication that man
fought afloat—boat against boat, or fleet against fleet—until after a
comparatively high stage of civilisation had been attained and
shipbuilding had made enormous advances.
Evelyn remarks: “Concerning men of war, fleets, and armadas for
battel, that Minos was reported to be the author, which shows that
manner of desperate combat on the waters to be neer as antient as
men themselves, since the deluge.” Minos, he adds, disputed the
empire of the seas with Neptune, but “these particulars may be
uncertain.”[1]
Among the legendary expeditions, those of Ulysses and Jason are
the best known. Possibly they took place, but the adventurers never
did or saw half the wonders narrated of them. Herodotus, describing
the type of ship attributed to Ulysses by Homer, states that such
ships were made of acacia, of “planks about two cubits in length,”
joined together like bricks, and built in the following manner: “They
fasten the planks round stout and long ties: when they have thus
built the hulls they lay benches across them. They make no use of
ribs, but caulk the seams inside with byblus. They make only one
rudder, and that is driven through the keel. They use a mast of
acacia and sails of byblus. These vessels are unable to sail up the
stream, but are towed from the shore.”[2] Book II. of the Iliad
mentions, in the famous catalogue, hollow ships, well-benched
ships, swift ships, and dark ships, and that Ulysses had twelve red
ships, but Homer, being a poet and a landsman, did not describe
their differences.
Recent excavations and discoveries in Egypt have revealed the
existence of boats of considerable size, so remote in history that
their period is only guessed at, though they are estimated to date
from about 5000 to 6000 years b.c. If the interpretation of the
designs on the pottery recording these old ships be correct, they
were propelled by over a hundred oars or paddles, were steered by
three paddles at the stern, and had two cabins amidships. They
were, moreover, very high out of the water at the ends, having very
long, overhanging bows and counters, and were shallow and flat-
bottomed.
Even at this period the art of shipbuilding was in a comparatively
advanced stage; vessels such as those depicted would be quite as
capable of use in war for carrying warriors or stores, or both, as in
commerce for conveying merchandise. Egypt has many historical
secrets yet to reveal, and, judging by the constant reassignment of
dates in all matters connected with Ancient Egypt which exploration
has entailed, it is not too much to expect that the dates quoted,
assigned approximately by Egyptologists, may be revised and
events placed more remotely still.
Another hieroglyph, discovered in a tomb, ascribed to the year 4800
b.c., shows enormous progress in shipbuilding and also in the art of
representing a ship pictorially.
During the Sixth Dynasty, a certain Un’e, who was a person of note
under three kings, sent, while the second, Pepi I., was on the throne,
an expedition to the quarries of Syene or Assouan, to fetch stone for
his master’s pyramid.
Another expedition, on behalf of Pepi’s successor, merits attention,
as the fact is emphasised on the inscription on the tomb as most
remarkable, and as never having occurred before “under any king
whatever,” that Un’e had to employ twelve ships for freight and but
one warship.[3] The flotilla consisted of “six broad vessels, three tow
boats, three rafts, and one ship manned with warriors.”[4]
The Egyptians evidently had experience of some sort of fighting
afloat, for there has been discovered at Gebel Abu Faida a tomb with
a painting showing a boat with a triangular mast, and a stem
extending forward below the surface of the water and presumably
intended to be used to damage an enemy’s boat by ramming it.[5]
The first sea-fight of which a pictorial representation is known to exist
was fought off Migdol, at the mouth of the Nile, in the time of
Rameses III., first king of the Twentieth Dynasty, which began about
1180 b.c., and lasted to about 1050 b.c. Egypt was invaded from the
East by “warships and foot soldiers,” and the Egyptian monarch
mustered a fleet and attacked them.
ANCIENT EGYPTIAN WARSHIP.
GREEK BIREME.
From a Vase in the British Museum, found at Vulci.
GREEK WAR GALLEYS.
From a Vase in the British Museum found at Vulci.
Cancelli, or shields of basket work, were placed along the sides of
the ships at such a height that the heads of those on board are just
visible. The cancelli bore a striking resemblance to the circular
basket-work boats still to be found on the upper Euphrates; this
supports the supposition that the cancelli may have been used for
other purposes, particularly if they were made comparatively
watertight, as the function of a shield was not only to protect a
warrior in battle, but to help to keep him dry when on shipboard by
being disposed along the sides to prevent the spray from entering
the ships. A forecastle was constructed upon these ships, and upon
each forecastle a look-out man was stationed; and when these
structures came to be built of larger dimensions they served to
accommodate a number of fighting men who, from their superior
position, could throw their missiles with greater effect. The forecastle
had the further advantage of serving as a stronghold in the event of
an attempt being made to capture the ship by boarding it.
Following the Phœnicians, the Greeks are thought to have begun to
build their own warships about 700 b.c., perhaps earlier, but it was
about that time that the first three-banked warship was launched at
Corinth. The three-banked ships were for many years the largest in
existence. During the fourth century b.c. shipbuilding was practised
extensively, four-banked ships being built at Chalcedon, five-banked
at Salamis, and six-banked ships at Syracuse. Ships of ten banks,
according to Pliny, were ordered by Alexander the Great, and about
300 b.c. ships having twelve banks are said to have been built for
Ptolemy, and fifteen-banked ships for Demetrios, for a battle near
Cyprus.
Ptolemy Philopater, who ruled in Egypt from 222 to 204 b.c., is
alleged to have had a forty-banked ship of a length of 280 cubits or,
reckoning the cubit at 18 inches, of 420 feet, and a beam of 57 feet.
While increasing the size and number of oars, it would, nevertheless,
be impossible to augment to any appreciable extent the speed at
which these ships could be rowed, and the more unwieldy would
they become, and the more difficult would it be to keep steering way
upon them. Again, the assertions of the historians are so
contradictory that it is a thankless task to attempt to reconcile all their
stories, especially as they depended much upon hearsay for their
information. For that reason, therefore, a great deal that has been
recorded as to the early ships and their numerous banks of oars is
not to be accepted without careful inquiry and verification.
It has never been established beyond question what is meant by
banks of oars, or whether the Greek text has been interpreted
correctly when it is taken to express forty superimposed banks of
oars. From constructional reasons it may be assumed that a ship
having forty superimposed banks of oars never existed, and it is very
doubtful whether ships having more than a fourth of that number of
banks passed beyond the imaginations of their inventors. In any
case they were soon dispensed with, and in course of time it was
found that the best results were obtained with galleys having two or
three banks of oars.
It is not definitely known how the rowers were disposed in the ships
of anything over seven or eight banks. If any vessels had forty banks
of oars, the upper rows must have been of an absolutely unwieldy
length. Assuming the oars to have been weighted with lead so that
the inborne and outborne portions were equally balanced, they must
nevertheless have been exceedingly difficult to row even by a
number of men, and it was impossible for any rowers to have moved
these great oars at the same speed as the men at the lower banks
moved their lighter and shorter ones. That some such difficulty was
experienced, even in biremes and triremes, is shown by the
arrangement of the oars, whereby all in a bank were not of equal
length, but were graded so that those nearer the ends of the banks
were longer in order that all the blades might enter the water in a
straight line. Each row above must have had its own line in the water
a little farther away from the side of the ship than the row beneath it,
or the blades would have interfered with each other and the rowers
thrown into hopeless confusion. The tremendous amount of lead that
would have to be carried to counterbalance the outborne portions of
several hundred oars would add materially to the dead weight to be
propelled, and, much of it being placed high above the water, the
stability of the vessel would be lessened.
The Athenians used leather or skin aprons or covers over the oar
holes to prevent the water entering, the oar passing through a hole in
the leather, and the apron was bound to the oar in such a way as to
be watertight. This contrivance was widely adopted later. The oar
ports were constructed between the ribs, but the oars instead of
being rowed against the ribs were pulled against thongs fastened to
the next rib, thus minimising the strain upon the ship’s structure and
preventing the oars being lost overboard. One man one oar was
apparently the general rule at that time.
In his most painstaking study of “Ancient Ships” Mr. Cecil Torr has
gone very closely into the subject of the oar equipment of the
galleys. An Athenian three-banked ship would carry two hundred
oars, of which thirty were worked from the upper decking, sixty-two
on the upper bank, and fifty-four to each of the lower. The earliest
two-banked ships had eighteen rowers. An Athenian four-banked
ship might carry two hundred and sixty-six oars. The Roman and
Carthaginian five-banked ships in use about 256 b.c. had three
hundred rowers besides the combatants. The statement is made by
an early historian that in 280 b.c. the Heraclean fleet on the Black
Sea included an eight-banked ship with a hundred rowers on each
file, or one thousand six hundred rowers in all. As usually the fighting
men carried exceeded the rowers in number, the ship must have had
close upon three thousand five hundred men aboard.
Warships of all the early Eastern nations were strengthened by
cables passed longitudinally round them in order to keep the timbers
in place and prevent them from being started under the strain
occasioned by the shock of ramming. Egyptian ships of about 1200
b.c. had cables stretched from stem to stern and passing over the
top of the mast and other posts, but this contrivance was to prevent
the vessel from drooping at the ends, a weakness known as
“hogging.” The shock to the ramming vessel was scarcely less
severe than that to the vessel receiving the blow. To take up the
strain and add to the power of the blow the bows were strengthened
by means of waling pieces which supported the ram proper. The
Greek ships were built with the keel, the stempost, and the lower pair
of waling pieces converging to hold the ram, while higher up the
stem was a smaller ram which in its turn was buttressed by another
pair of waling planks. The catheads, or beams projecting from the
bows on either side by which the anchors were raised, were so
placed on a level with the gangway and gunwale that they would
sweep the upper works of an enemy’s ship and smash its gangway
and hurl into the sea or the hold all the fighting men upon it. Ships of
more than three banks are believed to have carried another ram
level with the catheads, and to have had a ram for every pair of
additional waling beams. The ram heads were generally of bronze
and weighed 170 lb. or more.
“AN ANCIENT BIREME, FROM BASIUS, HAVING ONE TIER OF OARS ONLY.”