Book - 2016 Rectangular Dielectric Resonator Antennas Theory and Design

RajveerS.
Yaduvanshi
HarishParthasarathy
Rectangular
Dielectric
Resonator
Antennas
Theory and Design
Rectangular Dielectric Resonator Antennas
Rajveer S. Yaduvanshi Harish Parthasarathy

Rectangular Dielectric
Resonator Antennas
Theory and Design
123
Rajveer S. Yaduvanshi Harish Parthasarathy
Department of ECE Department of ECE
Ambedkar Institute of Advanced Netaji Subhas Institute of Technology
Communication Technologies and New Delhi, Delhi
Research India
New Delhi, Delhi
India
ISBN 978-81-322-2499-0 ISBN 978-81-322-2500-3 (eBook)

DOI 10.1007/978-81-322-2500-3
Library of Congress Control Number: 2015942632
Springer New Delhi Heidelberg New York Dordrecht London

Springer India 2016
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microlms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specic statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made.
Printed on acid-free paper
Springer (India) Pvt. Ltd. is part of Springer Science+Business Media (www.springer.com)

Preface
Microwave dielectric resonator antenna (DRA) materials or ceramics were dem-

onstrated by Richtmeyer in 1939. Richtmeyer showed that these dielectric ceramics
can resonate. Theory of DRA was expanded by Okaya and Brash in 1960. More
experimental work on DRAs, done by Long in 1980, proved that DRAs can become
efcient radiators and can be used as antennas. S.A. Long experimentally imple-
mented DRAs of different shapes and sizes as a low-prole antenna.
Analysis and studies on characteristic equation, radiation patterns, and excitation
methodology made DRAs popular by providing a new avenue compared to tradi-
tional patch antennas suffering from low gain and low bandwidth. Aldo Petosa
made DRAs a very successful candidate as functional antennas. Both the limitations
of low gain and low bandwidth in patch antennas can be eliminated by the use of a
rectangular dielectric resonator antenna (RDRA) operating in higher modes and
hybrid modes.
The modes theory of RDRA gives an important analysis on current distribution,
impedance, and radiation patterns of an antenna. Modes form a real, orthogonal
basis function for currents on the antenna. These are dened by boundary value
problems using eigenvalues and eigenvectors. The scope of this book has been
restricted to RDRAs, however, the concept can be extended to other geometries. In
RDRAs, once the excitation is given, the total distributed current on the antenna
structure becomes a weighted sum of eigen currents or a superposition of various
modes at any instant of time.
Resonant modes in RDRAs can be classied as dominant and higher modes.
Dominant modes correspond to lowest resonant frequency. These are called as TE,
TM, and HEM modes. E and H eld formats inside the RDRA at any instant of time
at a known frequency are termed as resonant modes. Modes excitation is directly
related to the surface current densities of the structure due to applied RF current.
This current gets converted into modal elds based on Maxwells equations. These
elds are restricted by RDRA boundary conditions. Reflection and refraction of
electromagnetic waves takes place because of dielectric interface at the boundary.
The generation of higher modes generally depends on RF excitation, device
dimensions, permittivity of dielectric material and coupling techniques used in
v
vi Preface
design of the antenna. The higher-order modes and hybrid modes have much
flexibility and design space in RDRA for different applications, but the excitation
techniques are complex. Rectangular DRA has a high degree of design flexibility
due to two aspect ratios (a/d and b/d), low cost, simplicity, and ease of fabrication.
It can retrot to the existing patch antenna technology for gain improvements.
Researchers have long felt the need for a rigorous theoretical analysis on reso-
nant modes of RDRA, and resonators have become a demanding eld for industry
and academia. This is because knowledge of resonant modes gives physical insight
to the antenna designer, based on which input impedance and radiation character-
istics can be predicted. We hope that this book will help to ll the gap.
The investigations and theory developed are based on applying waveguide
theory models. Propagation of electromagnetic elds has been taken along z-axis,
i.e., exp(c z). Initially, these are exploited via the Maxwells curl equations and
then manipulating them to express the transverse components of the elds in terms
of the partial derivatives of the longitudinal components of the elds w.r.t. x and
y (i.e., the transverse coordinates).
Waveguide models of four different boundary conditions lled with homoge-
neous as well as inhomogeneous dielectric materials with linear and nonlinear
permittivity, permeability, and conductivity have been developed to determine TE
and TM propagating electromagnetic elds. These have resulted in different sine
cosine combinations. TE modes generation required Hz elds as longitudinal elds
and Ex, Ey, Hx, and Hy elds as transverse elds.
If input excitation is applied along x-axis as partial elds, y-axis will have xed
variation and z-axis will have desired variation in propagating elds. For example,
TE d13 . Similar cases can be developed for TM modes so as to propagate Ez elds
as longitudinal and Ex, Ey, Hx, and Hy as transverse elds. Hz eld will get vanished
because of boundary conditions.
An equivalent but computationally simpler way to pass on from waveguide
@
physics to resonator physics is to just replace (c) by ( @z ) in all the waveguide
formulae that express the tangential eld components in terms of the longitudinal
components. This is done after solving the full 3D Helmholtz equations using
separation of variable as x, y, z.

x2 Ez
r 2
2
0
c H z
The discrete modes (mnp) enable us to visualize the resonator as collection of

L, C oscillators with different L, C values. The outcome of all this analysis enables
us to write down the E and H elds inside the resonator, as superposition of four
and three vector-valued basis functions.
Preface vii
X
1 n o
E x; y; z; t Re Cmnpejxmnpt wEmnp x; y; z
mnp1
X
1 n o
E x; y; z
Re Dmnpejxmnpt / mnp
mnp1
and
X
1 n o
H x; y; z; t Re Cmnpejxmnpt wH
mnp
x; y; z
mnp1
X
1 n o
H x; y; z
mnp1
We note that there are only two sets {C(mnp)} and {D(mnp)} of linear com-
bination of coefcients using from the Ez and Hz expansions. The vector-valued
complex functions are as follows:
E ; wH ; /
wEmnp ; / H 2 R3
mnp mnp mnp
where R is autocorrelation and contain components fcos; sing fcos; sing

fcos; sing functions and hence for m0 ; n0 ; p0 6 m; n; p, each function of the set,
where m, n, p are integers.
n o
; wH ; /
wEmnp ; /
E H
mnp mnp mnp
is orthogonal to each function of the set:

n o
E ; wH ; /
wEm0 n0 p ; / H
0
mnp0 mnp 0
mnp 0 0
w.r.t. the measure of dx dy dz over surface of RDRA [0, a] [0, b] [0, d], where
a, b, and d are RDRA dimensions. The exact form of the function / E; /
H ; wE ; wH
depends on the nature of RDRA boundaries.
Excitation of RDRA plays very important role for modal analysis. To calculate
the amplitude coefcients {C(mnp)} and {D(mnp)}, we assume that at z = 0, an
e e
excitation Ex x; y; t or Ey x; y; t is applied for some time say t [0, T] and then
removed, as usually is done in L, C oscillators. Then, the Fourier components in this
excitation corresponding to the frequencies {mnp} are excited, and their solutions
are the oscillations for t > T. The other Fourier components decay within the
resonator.
viii Preface
{Cmnp, Dmnp} magnitude components can be determined based on principle of

orthonormality:
X
jxmnpt E
Re Cmnpe wmnp X x; y; 0
mnp

E x; y; 0 E e x; y; t
Re Dmnpejxmnpt / mnp X x
and
X
Re C mnpejxmnpt wEmnp Y x; y; 0
mnp

E x; y; 0 Ee x; y; t
Re Dmnpejxmnpt / mnp Y y
E x; y; 0}; for different (m, n),

By using orthogonality of {wEmnp X x; y; 0, / mnp X
we write p xed and likewise of {wEmnp Y x; y; 0;/ E x; y; 0}; in addition, we
mnp Y
need to use KolmogorovArnoldMoser (KAM) type of time averaging to yield:
E x; y; 0
C mnpwEmnp X x; y; 0 Dmnp/ mnp X
ZT
lim 1
Exe x; y; tejxmnpt dt
T ! 1 2T
T
and likewise

CmnpwEmnp Y x; y; 0 Dmnp/
E
x; y; 0
mnp Y
ZT
lim1 e
EY x; y; tejxmnpt dt
T ! 1 2T
T
In this book, RDRA resonant modes theoretical as wells as practical aspects have
been investigated along with rigorous mathematical analysis for TE, TM, and HEM.
Higher modes generation and control of resonant modes have been experimented.
Shifting of dominant mode toward higher modes and vice versa is desired phenom-
enon for recongurability, merging of neighboring resonant modes have been
exploited with simulation results. Use of higher modes for practical applications in
antennas has been described. Merging of neighboring modes signicantly increased
antenna bandwidth. The device miniaturization using high-permittivity materials has
been described. The devising control on modes has imparted reconguration of
operating frequency, beam pattern, beam width, polarization, gain, and bandwidth.
Higher modes radiation pattern, sensitivity analysis by changing dimensions, and
permittivity analysis by changing permittivity have been mathematically modeled, and
each is supported with simulated and experimental results. Selecting and cancelling a
particular resonant mode has also been described. The concept of modes has been
Preface ix
supported with practically implemented case studies. Devising control on resonant

modes in RDRA can be used for software-dened radios and military applications,
where frequent change of antenna parameters is operational requirement. For auto-
mation on modes control, microcontrollers equipped with lookup table can be used.
The modes have been modeled by R, L, C networks. Antenna far elds patterns
and impedance have been computed and measured. Analysis on hybrid modes in
RDRA has been discussed. Hybrid modes are complex to determine. Their math-
ematical formulations have been described. These modes are diversied.
The excitation of hybrid modes is complex, and their effective control can
revolutionize the antenna technology. Detailed study of mathematical modeling of
hybrid modes has been described. Hybrid modes are more popular for azimuthally
eld variations. The transcendental equation and characteristic equation for RDRA
modes are used for determining propagation constants and then resonant frequency.
The solution of resonant modes can be obtained using the following:
(a) Hz and Ez elds are expressed as umnp(x, y, z), vmnp(x, y, z) and xmnp based on
solving Maxwells equations with given boundary conditions.
(b) At z = 0, surface (x, y) excitation with applied surface current density is

Jsx x; y; t; Jsy x; y; t
(c) Surface current density is equated with generated magnetic elds

Js x; y; d Jsx ; Jsy ^z H Hy ; Hx ;
at z = 0; amplitude coefcients (Dmnp and Cmnp) are obtained on expansion of

Hz is terms Dmnp, and Ez terms as Cmnp.
(d) Equate tangential component of Ez at boundary, i.e., Ey jz0 to zero, and
compute the coefcients Dmnp for Hz and Cmnp of Ez.
(e) Excited by mnp and arbitrary feed position in xy plane x0 ; y0 /0 ; h0
" #
X n o
H? ~
Re Dmnp e jxmnpt
r? ~umnp x; y; z
mnp
X n o
Re C~ mnp ejxmnpt r? ~umnp x; y; z
mnp
and similarly E? .
Depending on the boundary conditions, four cases have been developed.
In RDRA, these four walls are assumed as perfect magnetic conductors and top and
bottom walls are taken as perfect electric conductors.
umnp sin sin sin Ez
vmnp cos cos cos Hz

x Preface
Sidewalls and top walls all are perfect electric conductors
umnp sin sin cos Ez
vmnp sin sin sin Hz
Sidewalls and top walls all are perfect magnetic conductors
umnp cos cos sin Ez
vmnp sin sin cos Hz
Top and bottom walls are perfect magnetic conductors, and all four sidewalls are
PEC
umnp cos cos cos Ez
vmnp sin sin sin Hz
Transcendental equation is used to solve propagation constants, i.e., kx, ky, and kz.
The propagation constant gives rise to resonant frequency with the help of charac-
teristic equation. These wave numbers kx, ky, and kz are in x, y, and z-directions,
respectively. The free space wave number is k0. The resonant frequency can be
determined from combined solution of transcendental equation and characteristic
equation of rectangular DRA. Time-averaged electric energy = time-averaged
magnetic energy
er k02 kx2 ky2 kz2

0
e0 k02 kx2 ky2 kz2
0
kz 6 pp=d
kz
tankz d q ;
k0 er 1 kz2
2
the nal result of transcendental equation is thus achieved.

The contents of this book are the outcome of our research work on RDRA
higher-order resonant modes. In this book, analyses have been restricted to rect-
angular resonators higher modes, however, the concept can be extended to other
geometry resonators, such as cylindrical, conical, and hemispherical. With this
book, we hope to ll the gap for rigorous theoretical analysis on RDRA resonant
modes. The work is supported with live projects data and their case studies. This
book should be very useful for antenna designers, both in research and development
and for practical implementations. This book is written in a simple and reader
friendly manner and can be easily understood with an initial knowledge of basic
Preface xi
electromagnetic theory. All the chapters are self-reliant, and no initial specialization
is required to understand the contents. We hope that this book will help open the
design space for a new class of antenna implementations.
This book is organized into 12 chapters including rigorous theoretical analysis of
modes along with case studies and design data annexure. Introduction along with
history of RDRA is given in Chap. 1. Introduction of resonant modes is explained
in Chap. 2. Mathematical derivations for modes and the generation of TE/TM
modes have been discussed in Chap. 3. Chapter 4 presents the derivation of RDRA
transcendental equations. In Chap. 5, mathematical description of amplitude coef-
cients of even and odd modes is presented. Chapter 6 contains radiation param-
eters and mathematical explanations of RDRA. Chapter 7 describes derivations of
higher-order resonant modes and their applications for high-gain antenna designs.
Chapter 8 explains the effect of angular variation on excitation to produce various
types of radiation patterns to meet military requirements. Chapter 9 discusses
sensitivity analysis and mathematical modeling of radiation pattern solutions in
RDRA. Chapter 10 presents the excitation of hybrid modes in RDRA and their
possible applications. Chapter 11 covers inhomogeneous solution along with
measurements. Basic RDRA resonant frequency formulations, materials required,
and their sources are given in the annexures. Complete and detailed solutions of
RDRA have been explained in case studies. Design data are provided in the
annexures. Chapter 12 discusses case studies.
New Delhi, India Rajveer S. Yaduvanshi

Harish Parthasarathy
Acknowledgments
We would like to express special thanks to reviewers and editors for sparing their
valuable time in reviewing this manuscript. We acknowledge the support and
cooperation extended by all our family members, without whom it would not have
been possible to complete this book.
We in person acknowledge the healthy discussion with our students and col-
leagues. Special thanks to Sujata, R. Gupta, Promod, Vipin, and Chander P. for
imparting their help and support in preparing this manuscript, i.e., typing work and
proofreading. We extend special thanks to Springer for publishing this book.
Rajveer S. Yaduvanshi
Harish Parthasarathy
xiii
Contents
1 Rectangular DRA Fundamental Background . . . . . . . . . . . . . . . . 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 History of DRA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Working Mechanism of RDRA. . . . . . . . . . . . . . . . . . . . . . . 2
1.4 Antenna Radiation Parameters . . . . . . . . . . . . . . . . . . . . . . . 4
1.5 Advantage of RDRA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.6 Resonant Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.7 Characterization of Resonant Modes . . . . . . . . . . . . . . . . . . . 6
1.8 Magnetic Dipole Moment . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.9 Spring Resonator of Length L. . . . . . . . . . . . . . . . . . . . . . . . 8
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2 Rectangular DRA Resonant Modes and Sources. . . . . . . . . . . . . . 11

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2 Type of Modes (TE, TM, HEM). . . . . . . . . . . . . . . . . . . . . . 14
2.3 Solutions of Helmholtz Equation. . . . . . . . . . . . . . . . . . . . . . 15
2.4 Rectangular Waveguide Analysis . . . . . . . . . . . . . . . . . . . . . 17
2.5 Two-Dimensional Resonator. . . . . . . . . . . . . . . . . . . . . . . . . 20
2.6 Basic Mathematical Representation of Resonant Modes . . . . . . 21
2.7 Voltage Source Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.8 Resonant Modes Generation . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.9 MATLAB Simulated Results . . . . . . . . . . . . . . . . . . . . . . . . 29
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3 Mathematical Analysis of Rectangular DRA. . . . . . . . . . . . . . . . . 33

3.1 Rectangular DRA with Homogeneous Medium . . . . . . . . . . . . 34
3.2 Rectangular DRA Mathematical Modeling . . . . . . . . . . . . . . . 36
3.2.1 Model-1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2.2 Model-2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
xv
xvi Contents
3.2.3 Model-3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.2.4 Model-4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.2.5 Basic Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4 Mathematical Analysis of Transcendental Equation

in Rectangular DRA . . . . . . . . . . . . . . . . . . . . . . . . . . . ....... 57
4.1 Case-1: Top and Bottom Walls as PMC and Rest
of the Four Walls are PEC . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.2 Case-2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.3 MATLAB Simulation Results. . . . . . . . . . . . . . . . . . . . . . . . 78
4.4 Resonant Frequency of RDRA for Experimentations . . . . . . . . 87
5 Mathematical Analysis of RDRA Amplitude Coefficients . . . . . . . 103

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
5.2 Amplitude Coefficients Cmnp . . . . . . . . . . . . . . . . . . . . . . . . 104
5.3 RDRA Maxwells Equation-Based Solution . . . . . . . . . . . . . . 106
5.4 RDRA Inhomogeneous Permittivity and Permeability . . . . . . . 112
5.5 RDRA with Probe Current Excitation . . . . . . . . . . . . . . . . . . 118
5.6 RDRA Resonant Modes Coefficients
in Homogeneous Medium . . . . . . . . . . . . . . . . . . . . . . . . . . 121
5.7 RDRA Modes with Different Feed Position . . . . . . . . . . . . . . 123
5.8 R, L, C Circuits and Resonant Modes . . . . . . . . . . . . . . . . . . 125
5.9 Resonant Modes Based on R, L, C Circuits . . . . . . . . . . . . . . 130
6 Mathematical Analysis of Radiation Pattern of RDRA . . . ...... 135

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 135
6.2 Radiation Pattern of RDRA Due to Probe Current i(t)
and Probe Length dl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
6.2.1 Radiation Pattern. . . . . . . . . . . . . . . . . . . . . . . . . . 139
6.3 Poynting Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
6.4 Moat-Shaped RDRA Radiation Pattern . . . . . . . . . . . . . . . . . 141
6.5 Quality Factor of RDRA . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
7 Rectangular DRA Higher-Order Modes and Experimentations . . . 147

7.1 Introduction to Higher Modes. . . . . . . . . . . . . . . . . . . . . . . . 148
7.2 Resonant Frequency and RDRA Structure . . . . . . . . . . . . . . . 156
7.2.1 Fields in Rectangular DRA. . . . . . . . . . . . . . . . . . . 158
7.3 Modes (Resonant) Mathematical Solution. . . . . . . . . . . . . . . . 165
7.4 Top-Loading RDRA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
7.5 Simulated HFSS Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
7.6 Modes at Varying Heights of RDRA . . . . . . . . . . . . . . . . . . . 168
7.7 Distortions Due to Overlap of Dipole Moment . . . . . . . . . . . . 168
7.8 Prototype and Anechoic Chamber Experimentations . . . . . . . . 168
7.9 Adjacent Modes Combination for Broadband Applications . . . . 169
Contents xvii
7.10 Effect of Air Gap Between RDRA and Ground Plane . . . . . . . 169
7.11 Effect of Asymmetrical Wells Inside RDRA. . . . . . . . . . . . . . 170
7.12 Effect of Moat Insertion Inside RDRA. . . . . . . . . . . . . . . . . . 170
7.13 Effect of a/b and d/b Aspect Ratio . . . . . . . . . . . . . . . . . . . . 171
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
8 RDRA Angular Excitation Mathematical Model

and Resonant Modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 181
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 181
8.2 Angular Shift in Excitation. . . . . . . . . . . . . . . . . . . . . ..... 185
8.3 Radiation Pattern Based on Angle 0 ; /0
Variation in xy Plane . . . . . . . . . . . . . . . . . . . . . . . . . ..... 189
8.4 Replacing Probe with Slot of Finite Dimensions
(Ls, Ws) at an Angle h0 ; /0 . . . . . . . . . . . . . . . . . . . ..... 190
8.5 HFSS Computed Radiation Pattern with Shifted hi ; /i
Slot Positions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 192
8.6 Experimentations . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 193
9 Sensitivity Analysis of Rectangular DRA . . . . . . . . . . . . . . . . . . . 199

9.1 MATLAB Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
9.2 HFSS Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
9.2.1 HFSS Result. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
9.3 Radiation Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
10 Hybrid Modes in RDRA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211

10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
10.2 Mathematical Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
10.3 Modes in Homogeneous Medium with Source Terms . . . . . . . 218
10.4 Current Density in RDRA . . . . . . . . . . . . . . . . . . . . . . . . . . 219
10.5 E and H Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
10.6 Mathematical Modeling of Hybrid Modes . . . . . . . . . . . . . . . 221
10.7 General Solution of Hybrid Modes (HEM) . . . . . . . . . . . . . . . 225
10.8 HFSS Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
10.9 Prototype RDRA Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
11 Inhomogeneous Permittivity, Permeability, and Conductivity

Solution in Rectangular DRA . . . . . . . . . . . . . . . . . . . . . . . .... 233
11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 233
11.2 Mathematical Model . . . . . . . . . . . . . . . . . . . . . . . . . . .... 234
11.3 Applications: Hybrid Modes Generation Inside RDRA
Can Be Used for Polarization Diversity . . . . . . . . . . . . . .... 250
11.3.1 RF Measurements for Antenna Parameters . . . . .... 250
xviii Contents
12 Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251

12.1 Structure and Hardware Experimentations . . . . . . . . . . . . . . . 251
12.1.1 RDRA Antenna Results . . . . . . . . . . . . . . . . . . . . . 251
12.2 RDRA with ManganeseManganese Material as Dielectric . . . . 256
12.3 Dual-Feed RDRA with Measurements Results . . . . . . . . . . . . 269
12.4 Isolated and Grounded RDRA . . . . . . . . . . . . . . . . . . . . . . . 279
12.4.1 S11 Plot. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
12.4.2 Gain Plot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
12.4.3 Impedance (Z) Plot . . . . . . . . . . . . . . . . . . . . . . . . 281
12.4.4 Design of RDRA with Ground Plane . . . . . . . . . . . . 282
12.4.5 S11 Plot. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
12.4.6 Gain Plot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
12.4.7 Impedance Plot . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
12.4.8 Comparison of DRA With and Without
Ground Plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
12.4.9 Detailed Design of Aperture-Coupled DRA . . . . . . . 285
12.4.10 Return Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
12.4.11 Radiation Pattern. . . . . . . . . . . . . . . . . . . . . . . . . . 287
Annexure-1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Annexure-2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
Annexure-3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Annexure-4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Annexure-5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
Annexure-6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
About the Authors
Dr. Rajveer S. Yaduvanshi is working as an associate professor in the

Department of ECE at Ambedkar Institute of Advanced Communication
Technologies and Research Delhi for the last 7 years. Before that, he was working
as a senior scientic ofcer in Ministry of Defence, Government of India. He has a
total of 31 years of experience in teaching and research. He is the author of a book
on MHD Antenna, Design and Applications and has published 27 papers in reputed
journals and conferences. He has supervised 151 B.Tech. projects and 11 M.Tech.
projects and is currently supervising seven Ph.D. students. He holds an M.Tech.
degree from NIT Allahabad and a Ph.D. degree from NSIT, Delhi (Delhi
University). His research interests include design of Dielectric resonator antennas,
MHD embedded antennas and analysis of higher modes in DRA.
Dr. Harish Parthasarathy is working as a professor in the Department of ECE at
Netaji Subhas Institute of Technology. He is the author of more than 11 books and
has guided several Ph.D. students. He holds his Ph.D. and B.Tech. degrees from IIT
Delhi. His research involves antenna and signal processing with specialization in
mathematical modeling.
xix
Chapter 1
Rectangular DRA Fundamental
Background
Abstract This chapter introduces rectangular dielectric resonator antenna.

Working mechanism of rectangular DRA (RDRA) has been explained. Survey
work along with citations on related works based on the available literature has
been described. RDRA as a new candidate in the eld of antennas whose com-
parison has been made with existing patch antennas. Their advantages have been
listed. Mathematical solution of one-dimensional resonator has been derived.
Keywords RDRA (rectangular DRA)

Working mechanism Survey

Characteristics Advantages One-dimensional resonator
1.1 Introduction
Antenna is usually visualized as metallic device for radiating and receiving elec-
tromagnetic waves. It is an interface (transducer) between space and communica-
tion device. For wireless communication system or radar system, antenna is used to
couple radio energy from transmitter to space in transdirection, and space to
receiver in receive direction. Antennas are frequency dependent. The design of
antenna corresponds to specic bandwidth and resonant frequency. These are
purely designed as per requirements. The antenna rejects all signals beyond their
bandwidth. An antenna is an integral part of any wireless communication. Hence,
its development must be in synchronization with communication system. There
have been revolutionary developments in communication systems since last dec-
ades. The emergent requirements are being felt in antenna development. The
Gigabytes of data transmission at very high speeds are todays communication
requirements. To match todays advanced communications requirements, rectan-
gular DRA (RDRA) is the most suitable candidate. Rectangular dielectric resonator
antenna is new kind of antenna, which is different from traditional metal or patch
antenna. The patch or metal antennas generally suffer from low bandwidth, high
conducting loss and low gain. RDRA has high gain and wide bandwidth antenna.
Springer India 2016 1

R.S. Yaduvanshi and H. Parthasarathy, Rectangular Dielectric
Resonator Antennas, DOI 10.1007/978-81-322-2500-3_1
2 1 Rectangular DRA Fundamental Background
1.2 History of DRA
Dielectric resonator antenna is a microwave antenna consists of block of ceramics

material having permittivity greater than 10 F/m. In 1939, R.D. Richtmeyer showed
that non-metalized dielectric material objects can resonate and function as an
antenna, these are called as dielectric resonator antenna [1, 2]. There were no
practical applications of these DRAs until 1960. Dielectric resonator antenna was
rst introduced by S.A. Long in 1980 [3]. Since then, vast research has been carried
out for analysis of DRA material properties, and their effective use as DRA. Various
shapes and excitation methods for DRA have been developed. Many research
papers have been published in reputed journals by researcher such as Kishk [4], Lee
[5], Leung [6], Luk [7], Mongia [8], Shum [9], Junker [4], Antar [10], and Petosa
[11] till date. No rigorous theoretical analysis for RDRA is available in the literature
so for. It is felt that if good literature along with sound mathematical analysis on
RDRA is made available, it can benet the society in large. Only few books are
available on introduction of DRA, but no book is available for sound theoretical
analysis supported with mathematical computations of RDRA.
1.3 Working Mechanism of RDRA
The electromagnetic waves were generated by rapid oscillations of electrons in

atoms causes acceleration or deaccelerations which become electromagnetic wave
radiation. The radio waves are introduced into ceramics forming resonator as shown
in Fig. 1.1 from RF transmitter circuits. These RF waves bounce back and forth
between resonator walls, thus forming standing waves, hence stores electrical
energy. Oscillating current introduces oscillating magnetic elds, H elds, and
oscillating electric elds, E elds. The time-varying eld radiates away from
antenna into space due to accelerrating currents. The walls of ceramic formed
partially transparent magnetic walls, and the magnetic energy leaks through these
transparent walls due to fringing effect. Thus, radio power is radiated into space.
Let RDRA having dimensions a, b, d lengths is excited by external electric elds
Eix(x, y) and Eiy(x, y) in x, y plane. The equivalent circuit as shown in Fig. 1.2 is
Fig. 1.1 Ceramics rectangular DRA with a, b, and d dimensions

1.3 Working Mechanism of RDRA 3
Fig. 1.2 RDRA equivalent

RLC circuit
drawn based on the electrical properties of this ceramic RDRA. This results into
longitudinal E z and H z elds. The probe currents are equated with RDRA radiating
currents as per principle of conservation of energy. In other words, time-average
(KAM) electric energies inside the RDRA are equated with time-average magnetic
energies. Figure 1.2 presents RDRA equivalent RLC circuit for computing quality
factor of RDRA.
Z Z
jE j2 dV jH j2 dV; 1:1
V V
h2mn c2 h2 ; 1:2
where c jpp
d .
These elds are computed using Helmholtz equations, taking into account of
source and RDRA boundaries. Mathematical solution of transverse and longitudinal
elds is obtained by half-wave and full-wave Fourier analysis, taking inside
medium and outside medium into consideration.
Figure 1.3 is shown as RDRA placed on innite ground plane. The image theory
can be applied to this RDRA for reducing its height. Varying sinusoidal in time,
energy flow in particular direction can be treated as power radiated per unit solid
angle (energy per unit area per unit time).
Fig. 1.3 RDRA with ground

plane
1.4 Antenna Radiation Parameters
Antenna radiation parameters are as follows:

Antenna radiation pattern
Power
Gain
Polarization
Impedance
Efciency radiation
1.5 Advantage of RDRA
Lower conduction losses due to use of dielectric material;

Most suitable at microwaves and millimeter waves;
Compact in size and portable;
kg
Dimensions of RDRA are of the order of p ; choosing higher er RDRA size
er
can be reduced signicantly.
Ease of fabrication;
No frequency drift due to change in temperature;
High-power handling capability;
High gain and high bandwidth;
Can be integrated with MIMC;
RDRA has advantage of two aspect ratios. Hence, various modes can be gen-
erated by varying any of the aspect ratio;
Simple coupling schemes;
Bandwidth can be variable by choosing dielectric constant; and
High Q factor.
1.6 Resonant Modes
In RDRA, resonant modes represent the radiating phenomena with the help of
E and H eld patterns. These elds inside RDRA are presented azimuthally. With
the knowledge of modes, radiation characteristics of an antenna can be predicted.
The designer can get insight of antenna design and hence can provide correction in
the antenna design. Resonant modes are real current vectors. These modes are
found by orthogonal Fourier basis functions. These are generated based on the
current distribution on the surface of antenna due to eld perturbations. These can
be classied as TE or TM modes. The loss tangent d introduced is due permit-
tivity of the material. The principle conservation of energy is applied, in which,
1.6 Resonant Modes 5
time-average electrical energy is equated to magnetic energy at any instant of time

to compute radiated elds. RDRA is excited by input radio frequency currents at
proper impedance match at input port. The transverse components are dened in
terms of longitudinal components and vice versa. The modal eld equations are
developed using Fourier basis functions of cosine or sine terms appearing based on
the RDRA boundary conditions, i.e., six walls of RDRA can be PMC, PEC, or any
combination of these PMC and PEC walls. Hence, resonant modes bring physical
insight into the radiating phenomena taking place inside the RDRA. The resonant
modes form a set of orthogonal functions to compute total current on the surface of
RDRA.
Figure 1.4 shows the resonant modes conguration generated into RDRA. Wave
can only propagate if wave vector k > kc, where kc is cutoff frequency. The lowest
resonance is called dominant mode.
This is solved and the solution consists of a superposition of a source (particular
solution) term and a homogeneous term (i.e., general solution of the homogeneous
Fig. 1.4 Resonant modes in

yz plane
part). Two constants in the homogeneous part are determined by applying the
vanishing boundary conditions on Hz and Ez at top and bottom surfaces, i.e., at
z = 0, d.
1.7 Characterization of Resonant Modes
The radiation can be identied as magnetic dipoles. Any function can be decom-
posed or separated by projecting that function into basis function, i.e., inner product
with basis function
Zb
hF1 ; F2 i rF2 ~
F1 ~ r dR; 1:3
a
The resonance modes are E and H eld patterns inside the RDRA. Figure 1.5 has
shown that electric elds are always associated with magnetic elds and vice versa.
These can be three types, i.e., TE, TM, and HEM modes. The amplitude coefcients
and phase of RDRA are Cmnp ; wmnp and Dmnp ; /mnp : Ez and Hz elds are based on
the orthonormality. These can be determined by applying principle of orthonor-
mality. The characteristics equations of RDRA are given as follows:
er k02 kx2 ky2 kz2 ; 1:4
where k0 is free space wave and kx ; ky ; kz are propagation constants in x-, y-, and
z-directions, respectively. Also, k02 x20 l0 e0 ; hence, resonant frequency in free
space can be determined based on the free space wave number. To determine
propagation constants, i.e., kx ; ky and kz , knowledge of transcendental equation is
required. The transcendental equation is developed for RDRA when elds are
propagating in z-direction and given below as
q
d
kz tan kz er 1k02 kz2 ; 1:5
2
Fig. 1.5 Electric elds and magnetic elds are associated

1.7 Characterization of Resonant Modes 7
The solution resonant frequency of RDRA can be determined depending upon

the resonant mode generated, i.e., TE111 ; TM111 ; TE11d ; TE1d1 ; and TE11d . Fields
are expanded into summation of their modal functions, which may be by Cmnp and
Dmnp amplitude coefcients. Applying continuity equation across regional inter-
faces tangential elds, current distribution along surfaces of an antenna can be
computed as Je , i.e., current density. Input impedance Zin and radiation pattern Prad
can be computed based on the current distribution. Eigenvectors or Eigen functions
are formed as characteristic modes. Modes are orthogonal over source region.
E z electric elds produced by J n characteristic currents on the surfaces. These
modes are mainly dependent on the RDRA boundary and excitation.
Electrical walls of RDRA:
E tan n E 0; 1:6
H nor n H 0; 1:7
Magnetic walls of RDRA:
H tan n H 0; 1:8
E nor n E 0; 1:9
The solution of resonant modes shall vary in terms of sine and cosine as these are
dependent on PEC and PMC walls of RDRA. E z and H z fields can be determined
as linear combinations of these functions sin or cosine in xy plane and z-component
of source, to get these propagation constant. Propagating elds in particular
direction x or y or z is assumed to be continuous inside and outside the RDRA.
While taking into account inside the resonator both, reflected as wells as propa-
gating elds are available, outside the RDRA only outgoing eld components are
taken and reflected component is negated. This solves the transcended equation for
RDRA. The modal characteristics of antenna give rise to elds, i.e., resonant
modes. These are also known as eigenvector and eigenvalues. Eigenvectors are
current amplitudes Cmnp and Dmnp , and eigenvalues are resonant frequencies xmnp .
The resonant frequency can be given as follows:

r
c mp2 np2 pp2
fr m; n; p p ; 1:10
2p el a b d
This book contents are lucid, simple, and pedagogical.

1.8 Magnetic Dipole Moment
The radiation in RDRA is taking place due to short magnetic dipole formation.
X
d e r; 1:11
where
d dipole moment
e charge
r distance between two charges
dX X
d er e v; 1:12
dt
dX
d :: e v; 1:13
dt
Hence, charges can radiate only if they move with acceleration. There will be no
radiation even if they move with xed or uniform velocity.
1.9 Spring Resonator of Length L
As shown in Fig. 1.6, one single string of AB length is applied with external
excitation to produce oscillations. These oscillations will give rise to resonant
frequency of the resonator.
x00 t x20 xt C2 e jxt ; 1:14
xt C1 e jxt
x20 x2 C1 C2 , where C1 and C2 are constants and f is the function of length L.
Fig. 1.6 Simple spring resonator

1.9 Spring Resonator of Length L 9
Hence, C1 x2Cx
2
2
0
jxt
xt xC22x
e
2 , if x0 x; then, x(t) will be 1.
0
Now, x x0 d; when d is small deviation,
C2 e jxt
; 1:15
x0 xx0 x
Hence, the solution of spring resonator in one dimension is given as follows:
C2 e jxt

d2x0

@2 1 @2
f x; t 0; at boundaries
@x2 c2 @t2
f 0; t 0 and f L; t 0
Taking Fourier Transform of above equation,

@ 2 x2 ^
f x; x 0; 1:16
@x2 c2
Writing above terms in sine and cosine form, we have

xx xx
C1 sin C2 cos 0 or ^f 0; x 0 or ^f L; x 0
c c
xL
sin c 0. Hence, kL np; sine values to be zero.
npc
x kc ; when n 1; 2; 3. . . where k x=c; 1:17
L
String length 2L, dominant frequency x1 , length L, dominant frequency is 2x1 .

Length is 2L/3, dominant frequency 3x1 ; Eqs. (1.1)(1.17) used in this chapter
presented the mathematical concept of topic.
References
1. Harrington RF (1961) Time harmonic electromagnetic elds. Wiley, New York

2. Richtmyer RD (1939) Dielectric resonator. J Appl Phys 10:391398
3. Wakino K, Tamura H, Sudo T (1987) Dielectric resonator materials and their applications.
Microw J 6:133148
4. Junker GP, Kajfez D, Kishk AA, Lisson AW (1995, May 11) Effect of aperture lling on slot
coupled dielectric resonator antennas operating in HEM11 mode. Electron Lett 31(10):774775
5. Lee RQ, Simons RN (1994) Bandwidth enhancement of dielectric resonator antennas. In:
IEEE antennas and propagation society international symposium, Seattle, WA, June 1994,
pp 15001503
6. Luk KM, Leung KW, Chow KY (1997) Bandwidth and gain enhancement of a dielectric
resonator antenna with the use of stacking element. Microw Opt Technol Lett 14(4):215217
7. Leung KW, Chow KY, Luk KM, Yung EKN (1997) Excitation of dielectric resonator antenna
using a soldered through probe. Electron Lett 33(5):349350
8. Mongia RK, lttipiboon A, Cuhaci M, Roscoe D (1994) Radiation Q-factor of rectangular
dielectric resonator antennas theory and experiment. In: IEEE antennas and propagation
society international symposium, Seattle, WA, pp 764767, June 1994
9. Shum SM, Luk KM, Leung WK, Wa K (1994) Mutual impedance of hemisphere dielectric
resonator antennas. IEEE Trans Antennas Propagat 42(12):16521654
10. Antar YMM, Chang D, Sequin G, Henry B, Keller MG (1998, Oct 5) Modied wave guide
model (MWGM) for rectangular dielectric resonator antennas. Microw Opt Technol Lett 19
(2):158160
11. Petosa A, littipiboon A, Cuhaci M, Larose R (1996) Bandwidth improvement for a micro strip
fed series array of dielectric resonator antennas. Electron Lett 32(7):608609
Chapter 2
Rectangular DRA Resonant Modes
and Sources
Abstract Basics of resonant modes have been described. Their mathematical

analysis for generation of different resonant modes have been presented in this
chapter. Realization of resonant modes based on MATLAB has also been worked.
Modes are generated by applying voltage source. Various types of resonant modes
have been described along with all possible applications.

Keywords Cavity resonator Resonant modes Type of modes Wave guide

analysis Mathematical description of resonant modes Simulated work
2.1 Introduction
In the early 1960s, Okaya and Barash [1] reported the rst ever DRA in the form of
a single-crystal TiO2. Since then, no rigorous theoretical analysis has been devel-
oped so far in the literature to evaluate the resonant modes in Rectangular DRA.
Based on Cherenkov principle of radiations, an external electric eld brings the
charges of the molecules of the dielectric into a certain ordered arrangement in
space and creates acceleration phenomenon in dielectric material itself. The
dielectric polarization P is equal to the total dipole moment induced in the volume
of the material by the electric elds. In most cases, the magnitude of polarization is
directly proportional to the intensity of the electric eld at a given point of a
dielectric. The relative permittivity is related to the dielectric susceptibility.
A dielectric resonator is dened as object of dielectric material which functions as
a resonant cavity by means of reflections at the dielectric air interface. The
discontinuity of the relative permittivity at the resonator surface allows a standing
electromagnetic wave to be supported in its interior at a particular resonant fre-
quency, thereby leading to maximum connement of energy within the resonator.
Certain elds distribution or modes will satisfy Maxwells equations and
boundary conditions. Resonant modes are eld structures that can exist inside the
DRA. Modes are the pattern of motion which repeat itself sinusoidally. Innite
number of modes can excited at same time. Any motion is superposition or

12 2 Rectangular DRA Resonant Modes and Sources
weighted sum of all the modes at any instant of time by combining amplitudes and
phases. As in the case of all resonant cavities, there are many possible resonant
modes that can be excited in dielectric resonators. The boundary conditions are
n H = 0; where H denotes the electric eld intensity and n denotes the normal to
the surface of the resonator.
And, n E = 0, is not necessarily satised at all the surfaces of the RDRA by all
the modes. Different resonant modes have distinct electromagnetic eld distribu-
tions within the DRA, and each mode may provide a different radiation pattern.
Operation of DRA is based on the process that if excitation is applied, then a
high magnetic eld is created inside the dielectric object placed on a ground plane.
Phenomena which occur like a charge particle passing to the eld create the
physical environment like any metal ball passing through liquid. Thus, there will be
change in the eld, contraction, and expansion, which causes fringing effect. This
way dielectric object starts to radiate. Another phenomenon that occurs is that there
might be reflection of the eld from sidewalls of dielectric object due to change in
the refractive index of the medium. The dielectric object acts as an oscillator.
Theory of characteristic modes can be applied in the design of antenna or DRA.
These modes give insight into physical phenomenon taking place inside device in
terms of current vectors as maxima and minima. This helps to locate the feeding
point and desired dimension of RDRA.
In 1968, modes were introduced by Garbacz and later by Harrington. Inagaki
gave simpler theory on modes for radiation mechanism in an antenna. It requires lot
of computation, for loading quality factor, double feeding to improve bandwidth,
and circular polarization. Characteristic modes are current modes or eigenvectors,
which are the solution of characteristic equation. These are orthogonal functions
that can predict total current on surfaces of body of the antenna. Also, desired mode
can be excited for specic radiating pattern. Excitation of mode mainly depends on
feeding arrangement, geometry of the device, and dielectric material used. In time
domain, varying electric eld can produce magnetic elds and vice versa. By
applying RF, excitation currents in RDRA get converted into surface current
density distributed over the surfaces, i.e., RF excitation with proper impedance
match can generates J. This probe current produced magnetic vector potential A.
The radiated magnetic elds are presented in the form E-electric eld intensity
using Lorentz gauge transformation. An antenna can propagate electromagnetic
elds, if wave vector k > kc . The cutoff wave vector kc determines the cutoff
frequency. There can be dominant resonant frequency or higher-order resonant
frequency. The propagation takes place along x-axis if propagation constant k [ np b.
There will not be any propagation if k np b , as it will lead to formation of standing
waves. Similar conditions persist for propagation along y-axis and z-axis.
Maxwells equations dene the behavior of electromagnetic wave propagation,
while the solution of Maxwell equation is dened by Helmholtz equation. The
radiated power is given by Parsevals power theorem. Half-wave Fourier analysis is
used to determine the time domain behavior of antenna radiations. The magnitude
and phase of the radiated eld is given by Poynting vector (S = E H). The image
2.1 Introduction 13
theorem can be applied for antenna size reduction. It can be implemented by

extending ground plane to an isolated RDRA. Resonance in RDRA is created due
to formation of standing waves inside the device. Frequency xmnp is the spectral
solution of an antenna, and this can determine the base half-wave Fourier analysis.
Principle of orthonormality is used to determine radiation parameters by equating
electric time average energy equal to magnetic time average energy by KAM
(KalmogorovArnoldMoser).
At any instant of time, n number of modes exist. The particular mode can be
excited by increasing weighted amplitude of desired mode. More than one mode
can also be excited into RDRA. Blocking modes can take place if Ez and Hz elds
of same frequency are available in RDRA at any instant of time. Hence, mode
spectrum will result into corresponding resonant frequency generation. Wave
propagation can be dened by Helmholtz equation. The Maxwells equation
describes the behavior of electromagnetic elds and forms the basis of all EM
classical phenomena. The size of antenna can be reduced to half by image theorem,
converting isolated cavity into innite ground plane. Dielectric resonator antenna is
formed with high permittivity substrate. The abrupt change in permittivity due to
change in medium forms standing waves. These waves establish resonance as they
bounce back and forth in-between two walls due to elds perturbation. Modes are
spectral resolution of electromagnetic elds of waves radiated by the RDRA. Modal
excitation mainly depends on:
(a) Position of probe;
(b) Magnitude of probe current; and
(c) Phase of input current.
To compute resonant modes, vector principle of orthonormality on half-wave
Fourier analysis has been applied, i.e., radiated magnetic energies time averaged are
compared with applied electrical energies time averaged in the case of resonator
antennas. More number of modes along z-axis in RDRA can be generated either by
increasing electrical height d of RDRA or by increasing resonant frequency of
DRA. Figure 2.1 depicts the prototype RDRA with moat under fabrication.
Fig. 2.1 RDRA prototype with neat sketch

Depending on the nature of the surfaces, different linear combinations of the c

modes are formed. The propagation constant c itself is taking discrete values. This
forces the natural frequencies of the eld oscillations to take discrete values (mnp)
indexed by three positive integers, namely m, n, and p. The solutions of the
waveguide problem yield discrete values of c, i.e., cm; n; x for a given frequency
x by applying boundary conditions to the electromagnetic elds on the sidewalls.
The corresponding eld amplitudes are solutions to the 2-D Helmholtz equations
corresponding to the transverse Laplacian r2? : These amplitudes are called the
waveguide modes and are of the form given below in Sect. 2.2.
2.2 Type of Modes (TE, TM, HEM)
EM waves are of four types given below:

Transverse electric and magnetic (TEM) mode
Transverse electric (TE) mode
Transverse magnetic (TM) mode
Hybrid electric and magnetic (HEM) or HE odd and EH even mode
Modes propagation depends mainly on following conguration:
1. Excitation
2. Dimensions
3. Coupling
4. Medium
5. Point of excitation
6. Input impedance
Cross-polarization solution can be the outcome of modes. They can be merged,
separated and mixed depending upon the requirements. Half Fourier analysis can be
used to describe modes of propagation and excitation. Even and odd modes can be
studied. They can be analyzed with magnetic dipole moments. They help to predict
far eld radiation patterns. Modulated bandwidth and gain control can be achieved.
High gain at higher modes can be used for hardware implementations. Device
dimensions can be minimized by proper selection of modes for resonant frequen-
cies. In case of milli metric (mm) wave, device size can be enlarged for easy
hardware development or hardware implementation. The solution is based on
waveguide method when boundaries have been all six electrical walls. The solution
is based on solution of Maxwells equations and then restricted to given boundary
conditions for conned modes of EM waves.
2.3 Solutions of Helmholtz Equation 15
2.3 Solutions of Helmholtz Equation
Helmholtz equation solution with source
@D
rH J 2:1
@t
D E
B lH r A
1
H r A
l
Considering the sources to be natural time harmonic
E Em ejxt 2:2a
H Hm ejxt 2:2b
Now,
@B
rE 2:3
@t
or
r E jxlH jxr A
r E jxA 0 2:4
Using vector identity
r re 0
E jxA re
E jxA re
Using the vector identity
r r A r r A r 2 A 2:5
Now,
r lH rr A r2 A
or
lr H rr A r2 A
From Maxwells fourth equation,
r H J jxE
or
B
r J jxE
l
or
r B lJ jxlE
or
r r A lJ jxlE
or
rr A r2 A lJ jxlE
rr A r2 A lJ jxljxA re
or
rr A r2 A lJ x2 lA jxlre
or
r2 A k2 A lJ jxlre rr A
or
r2 A k2 A lJ rjxle r A
where k2 x2 l.
2.3 Solutions of Helmholtz Equation 17
Using Lorentz condition, i.e.,
r A jxele
or
1
e r A
jxle
r2 A k2 A lJ rjxle jxele 2:6
Hence, r2 A k2 A lJ:
2.4 Rectangular Waveguide Analysis
Propagation in waveguide has been taken along z-axis, and all the four sidewalls of
waveguide are PEC; the elds computed are as follows:
Hx Ex
Hy Ey
Hz Ez
X
1 mpx npy
Ez x; y; z C m; n sin sin expcmn z; 2:7a
m;n1
a b
@Ez
Ez;y ;
dy
@Ez
Ez;x
dx
X
1 mpx npy
Hz x; y; z Dm; n cos cos expcmn z 2:7b
m;n1
a b
Form Maxwells equations
Curl E r E jxlH B; t 2:8a
Curl H r H jxE J D; t 2:8b
Solution of above equations is based on separation of variables solving LHS of both

sides rst

i j k
@ @Ez @Ey @Ez @Ex @Ey @Ex
r E @x @
@y
@
@z i j k
Ex Ey Ez @y @z @x @z @x @y
2:9a

i j k
@ @Hz @Hy @Hz @Hx @Hy @Hx
r H @x @
@y
@
@z i j k
Hx Hy Hz @y @z @x @z @x @y
2:9b
Comparing with RHS in both equations and getting value of Hx, Hy, Hz from (2.9a)
and Ex, Ey, Ez from (2.9b) we get

1 @Ez @Ey 1 @Hz @Hy
Hx ; Ex ; 2:10a
jxl @y @z jx @y @z

1 @Ez @Ex 1 @Hz @Hx
Hy ; Ey ; 2:10b
jxl @x @z jx @x @z

1 @Ey @Ex 1 @Hy @Hx
Hz ; Ez ; 2:10c
jxl @x @y jx @x @y
Ez;y cEy jxlHx
cEx Ez;x jxlHy
@Hz
Hz;y
dy
@Hz
Hz;x
dx
Similarly,
Hz;y cHy j jxEx
cEx Hz;x jxEy
These above equations can be placed in matrix form

jx c Ex Hz;y

c jxl Hy Ez;x
2.4 Rectangular Waveguide Analysis 19
and

jx x Ey Hz;x

c jxl Hx Ez;x
On manipulating them

Ey jxl c Hz;x

Hx c jx Ez;y
Hence, on simplication
jxl c
Ey Hz;x 2 Ez;y
h2m;n hm;n

mpx
npy
c z
X Dm; n np
b cos a sin b e
mn
Ex jxl 2
m;n
hm;n
mp
mpx
npy
c z
X C m; n
a cos a sin b e
mn
c 2
; 2:11
hm;n

mp

X Dm; n np
b cC m; n a mpx npy cmn z
jxl cos sin e
m;n
h2m;n a b
Similarly, we can compute
c2m;n k2 h2m;n ; k 2 lx2

mp2 np2
h2m;n
a b
cmn ! propagation constant
Za Zb mpx npy
2
Eix m; n Eix x; y cos sin dxdy
ab a b
00 2:12a
np
mp
jxl b Dm; n cm;n C m; n

a
;
h2m;n
Za Zb mpx npy
2
Eiy m; n Eiy x; y sin cos dxdy
ab a b
0 0
! 2:12b
jxl mp c Cm; n np
m;n
2 Dm; n b
;
hm;n a h2m;n
" mp cm;n np jxl

#
Eix m; n a h2m;n b h2m;n Cm; n

Eiy m; n c np
hm;n
2 b hjxl
2
mp Dm; n
m;n m;n a
Hence, Cm; n; Dm; n amplitude coefcients can be computed, when the

boundary conditions are given as:
8 9
< x 0; a =
y 0; b
: ;
z 0; d
Incident waves at input of waveguide are Eix x; y, Eiy x; y

np
mp
h mpx npyi
X jxl Dm; n cm;n Cm; n
Eix x; y b a
cos sin ;
m;n
h2m;n a b
2:13a
mp
np
h mpx npyi
X jxl Dm; n cm;n Cm; n
Eiy x; y a b
sin cos :
m;n
h2m;n a b
2:13b
2.5 Two-Dimensional Resonator
Solution is obtained by the application of Helmholtz equation.
@ 2 wx; y; t @ 2 wx; y; t 1 @ 2 wx; y; t

2 0 2:14
@x2 @y2 c @t2
Applying boundary conditions in rectangular plane,
w0; y; t wa; y; t 0
wx; 0; t wx; b; t 0
Let input excitation be some tension T

@2w @ @w @ @w
rdxdy T dy dx T dx dy 2:15
@t2 @x @x @y @y
Y 00 X 00
ky2 ; kx2 ;
Y X
2.5 Two Dimensional Resonator 21
@2w
c2 r 2 w 0 2:16
@t2
Using separation of variables:
wx; y; t X xY yT t 2:17
00
T 00 t 2 X x Y 00 y
x
2
c
T t X x Y y
let
X x sinkx x

Y y sin ky y
x2
kx2 ky2
c2
where kx and ky can be written as:

mp np
kx ; ky 2:18
a b
q
Frequency can be written as: xmn cp ma2 nb2 :
2.6 Basic Mathematical Representation of Resonant

Modes
r2 Ax k2 Ax 0; 2:19
kr 1 far eld pattern

kr 1 near eld pattern
where Az is the magnetic vector potential and k is the wave vector or wave number
along z-axis.

Az C1 cos kx x C1 sinkx xC3 cos ky y

C4 sinky yC5 coskz z C6 sinkz z;
Z
l e jkR 3 0
Az J z0 d z; 2:20
4p R
2p q
kc ; k0 kx2 ky2 kz2 x2 l; where k is the wave number:
k
The wave number can be dened as rate of change of phase w.r.t. distance in the
direction of propagation. Resonant frequency x xmnp in RDRA and its
mathematical expression is given below:
r
mp2 np2 pp2
c
fr m; n; p p ; 2:21
2p
er a b d
where m; n; p are the half-wave eld variations along x, y, z directions.

H rA
dA
E r ; scalar and magnetic vector potential from Lorentz Gauge
dt
conditions:
S E H ; S is Poynting vector energy flow or flux:
Prad
Z 2 input Impedance:
jIj
Ex ; E y ; E z ; H x ; H y ; H z are electric and magnetic fields
I n npx npxo I n
mpy mpyo
L cos ; sin L cos ; sin ; 2:22
a a b b
where L denotes linear components. It turns out that depending on the nature of
wall surfaces (PEC or PMC), four possible linear combinations can appear
(cos sin; sin cos; sin sin; and cos cos).
In rectangular DRA, weve got to applying in additional boundary conditions on
top and bottom surfaces to be the linear combinations as compared to waveguide.
C1 expfcm; n; xzg C2 expfcm; n; xzg
and these cases are cm; n; x pp

d ; when p 1; 2; 3. . . and have two possible
ppz
ppz
linear combinations of sin d and cos d :

Thus, the possible frequencies x obtained by solving cm; n; x pp d ; then
comes out to be:
2 1=2
m n 2 p2
xm; n; p p 2 2 2 : 2:23
a b d
2.6 Basic Mathematical Representation of Resonant Modes 23

@
physics to resonator physics is to just replace c by @z in all the waveguide
components. This is done after solving the full 3-D Helmholtz equations using
separation of variable in x, y, z.

x2 Ez
r2 2 0 2:24
c Hz
The discrete modes xmnp enable us to visualize the resonator as collection of L,

C oscillators with different L, C values. The outcome of all this analysis enables us
to write down the E and H elds inside the resonator, as superposition of four, three
vector-valued basis functions.
X
1 n o
Ex; y; z; t Re Cmnp e jxmnpt wEmnp x; y; z
mnp1
n o 2:25
X
1
E x; y; z ;
Re Dmnp e jxmnpt / mnp
mnp1
and
X
1 n o
H x; y; z; t Re Cmnpe jxmnpt wH
mnp
x; y; z
mnp1
n o 2:26
X
1
Re H x; y; z
Dmnpe jxmnpt / ;
mnp
mnp1
We note that there are only two sets fCmnp g and fDmnp g of linear combination of
coefcients from the Ez and Hz expansions. The vector-valued complex functions
E wH /
are wEmnp , / H R3 (where R is autocorrelation) and contains compo-
mnp mnp mnp
nents fcos; sing fcos; sing fcos; sing; functions and hence for m0 n0 p0 6
mnp; each function of the set:
n o
E ; wH ; /
wEmnp ; / H ;
mnp mnp mnp
is orthogonal to each functions of the set:

n E H
o
wEm0 n0 p ; /m0 n0 p ; wH
mnp
; / m0 n 0 p
;
w.r.t. the measure of dx dy dz over [0, a] [0, b] [0, d].

The exact form of the function / E; / H ; wE ; wH depends on the nature of the
boundaries. The next problem addressed can be on excitations of RDRA. To cal-
culate the amplitudes coefcients {Cmnp} and {Dmnp}, we assume that at z = 0,
e e
excitations Ex x; y; t or Ey x; y; t are applied for some time say t 2 [0, T] and
then removed. Then, the Fourier components in this excitation corresponding to the
frequencies xfmnpg are excited and their solutions are the oscillations, while the
waveguide for t > T. The other Fourier components decay within the resonator.
{Cmnp, Dmnp} are the components of the form:
X
Exe x; y; t Re Cmnpe jxmnpt wEmnp x x; y; 0
mnp
2:27
Re E x; y; 0
Dmnpe jxmnpt / mnp x
and
X
Eye x; y; t Re Cmnpe jxmnpt wEmnp y x; y; 0
mnp
2:28
Re E x; y; 0
Dmnpe jxmnpt / ;
mnp y
E x; y; 0g. For different (m, n),

By using, orthogonality of fwEmnp x x; y; 0; / mnp x
E x; y; 0g:
we write p to be xed and likewise of fwEmnp y x; y; 0; / mnp y
In addition, we need to use KAM type of time averaging to yield eld
components:
C mnpwEmnp x x; y; 0 Dmnp/ E x; y; 0
mnp x
ZT
lim 1
Exe x; y; tejxmnpt dt: 2:29
T ! 1 2T
T
and likewise
C mnpwEmnp y x; y; 0 Dmnp/ E x; y; 0
mnp y
ZT
lim 1
Eye x; y; te jxmnpt dt: 2:30
T ! 1 2T
T
2.7 Voltage Source Model 25
2.7 Voltage Source Model
This method of excitation can be compared with connecting a voltage or current

source to an LC circuit for sometimes and then switching it off. After a sufciently
long time, all frequencies in the LC circuit decay away except the frequency p1
LC
.
We can more generally compare a resonator with the material medium having
non-zero conductivity. Thus, the medium is characterized by the triplet (; b; r
which corresponds to an array of (C, L, R) = RLC circuits.
Such a resonator is analyzed in the same way replacing by 0 jr x , i.e.,
complex permittivity depending on frequency. The resonant frequencies xmnp
now have a non-zero imaginary part corresponding to decay of the eld with time.
Their frequencies and elds may also be determined by applying separation of
variables with boundary conditions to the Helmholtz equations.

E
r2 jxlr jx z 0; 2:31
Hz
To have sustained oscillations in such a resonator, we must never switch off the
excitation. We may for example apply a surface current source at z d0 , where
0\d0
d: Letting Jsx x; y; x and Jsy x; y; x be this surface current excitations in
the Fourier domain, the current density corresponds to this is given as:
^ Jsy x; y; xY^ dz d0 ;
J e x; y; z; x Jsx x; y; xX 2:32
This current is computed by substituting into the Maxwell curl equations
Curl E jxl H;
Curl H J e r jxE; div H 0
The method of solution is to express it as the sum of a general solution to the

homogeneous equations, i.e., with J e = 0 and a particular solutions for J e 6 0. The
general solutions to the homogeneous problem are the same as earlier explained,
i.e., containing only the frequencies fxmnpg. Particular solution to the
J e 6 0 inhomogenous problem is obtained by taking the curl of the second
equation and substituting the elds into third equation to obtain
r2 H r J e jxlr jxH; 2:33

We express a particular solution to this equation by setting
X
1
L1 m; n; x expcmnxzumn x; y; x
m;n1 2:34a
b1 m; n; x expcmnxzvmn x; y; x for d z [ d; 2:14
X
1
H p x; y; z; x L2 m; n; x exp cmnxzumn x; y; x
m;n1 2:34b
b2 m; n; x expcmnxzvmn x; y; x for 0 z\d; 2:15
Here, x is a continuous variable, unlike fxmnpg, umn x; y; x and vmn x; y; x

are multiples (x-dependent) of
n mpx npxo
mpy mpy
L cos ; sin L cos ; sin
a a p p
To meet the boundary conditions on the sidewalls, if z = 0, d; if the walls are PEC,
Hpz 0; when z = 0, d. That gives use
X
Hpz x; y; z; x Lm; n; x sin hfcm; n; xz d gumnz x; y; x; d\z d;
m;n
2:35
and
X
Hpz x; y; z; x bm; n; x sin hfcm; n; xzgumnz x; y; x; 0 z\d;
m;n
2:36
The elds Hp? x; y; z; x are easily determined from these equations in the
region z > d and z < d by differentiating them w.r.t. x, y, z; wherever c comes in the
@
multiple w.r.t. expcz, we replace it by @z etc.
In this way, we get
X
1
Hpx x; y; z; x L1 m; n; xwmnx x; y; z; x; for z [ d; 2:37a
m;n1
and
X
1
Hpy x; y; z; x L2 m; n; xwmny x; y; z; x; for z\d; 2:37b
m;n1
2.7 Voltage Source Model 27
where, wmnx x; y; z; x and wmny x; y; z; x are obtained by differentiating
umnz x; y; x sin hfcm; n; xz d g w.r.t. x; y; z:
Likewise for z\d, we have expression of the form
X
1
Hpx x; y; z; x
bm; n; x/
E
x; y; z; x; 2:38a
mnp x
m;n1
and
X
1
Hpy x; y; z; x
bm; n; x/
E
x; y; z; x; 2:38b
mnp y
m;n1
The coefcients Lm; n; x and bm; n; x are obtained from the boundary
conditions

^z Hjzd Hjzd Js jzd :
Hence, current density
^ Jsy x; y; xY:
J Jsx x; y; xX ^ 2:39
2.8 Resonant Modes Generation
The Fig. 2.2 presents how the generated modes look like. This will be able to tell us
the number of resonant modes in particular direction. The transverse components of
EM waves are expressed as Ex ; Ey ; Hx ; Hy : If propagation of wave is along
z-direction, Ez; Hz elds are the longitudinal components. These elds are modal
solutions, solved based on Helmholtz equations using standard boundary condi-
tions. The RDRA is basically a boundary value problem. The linear combinations
of sine and cosine terms give rise to TE and TM modes. The generation of various
kinds of modes in an antenna and propagation is very critical issue; it need through
study. Now, rewriting Helmholtz equation for source-free medium (Fig. 2.3)
Fig. 2.2 Rectangular resonator

Fig. 2.3 Resonant modes

generated in RDRA by HFSS.
x x x
a TEd12 , b TEd14 , c TEd16
modes
r2 W k2 W 0;
Here, k is the wave number and k 2 kx2 ky2 kz2 and, W Wx Wy Wz

1 @ 2
W kx2 2:40a
Wx @x

1 @ 2
W ky2 2:40b
Wy @y

1 @ 2
W kz2 2:40c
Wz @z
Solving above function and keeping propagation in +z-direction only, we get

W or Hz or Ez A sin kx x B cos kx x C sin ky y D cos ky y ejkz z
From boundary conditions, we get

Xn mpx npyo
Hz C mn cos cos e jkz z ; C mn Fourier Coefficients; 2:41a
m;n
a b
Xn mpx npyo
Ez Dmn sin sin e jkz z ; Dmn Fourier Coefficients; 2:41b
m;n
a b
Let c jkz and m, n are integers and a, b are dimensions;

mp2 np2
c2 x2 l kx2 ky2
a b

mp 2 np 2
kz2 x2 l
a b
2.8 Resonant Modes Generation 29
Hence, EM wave will propagate in z-direction if

mp 2 np 2
x2 l [0
a b
This gives cutoff frequency as

s
1 mp2 np2
xc p
l a b
It means, waveguide will support all waves having greater than xc to propagate.
Now, rewriting Hz and Ez
Xn mpx npyo
Hz C mn cos cos e jkz z 2:42
m;n
a b
Xn mpx npyo
Ez Dmn sin sin e jkz z 2:43
m;n
a b
Here, Cmn and Dmn are the coefcients of Fourier cosine and sine series.
q
cm;n h2m;n x2 le
Hence, C mn and Dmn gives us relative amplitudes and phase. Hence, we get solution
of possible amplitudes and phase of wave propagating through rectangular wave-
guide called as modes of propagation.
2.9 MATLAB Simulated Results
Results of resonant frequency obtained on various sizes RDRAs using HFSS have
been placed in Table 2.1. The MATLAB programs are being developed for modes
graphical view. Resonant modes and resonant frequencies are being obtained based
on formulations. The programs and simulated results are given below:
Table 2.1 RDRA HFSS fr

S. No. Permittivity Dimension (a b h) mm Resonant frequency
1 10.0 14.3 25.4 26.1 3.5
2 10.0 14 8 8 5.5
3 10.0 15.24 3.1 7.62 6.21
4 20.0 10.2 10.2 7.89 4.635
5 20.0 10.16 10.2 7.11 4.71
6 35.0 18 18 6 2.532
7 35.0 18 18 9 2.45
8 100.0 10 10 1 7.97
The graph shown in Fig. 2.4 represents inverse relationship between height and
resonant frequency as k-wavelength is inversely proportional to resonant frequency
fr. MATLAB simulation shown in Fig. 2.5 represents number of modes generated
in x, y, z directions. The mathematical expression on the topic is expressed in
Eqs. (2.1)(2.31).
2.9 MATLAB Simulated Results 31
Fig. 2.4 Simulated resonant frequency plot for excited modes
1
0.8
0.6
0.4
0.2
0
-0.2 z
-0.4
-0.6
-0.8
-1
3
2
1 x 4 5
y 3
0 1
2
-1 0
-2 -1
-2 -3
-3 -5 -4
TE7,10,6
Fig. 2.5 Resonant modes 3D in RDRA in xyz plane

Reference
1. Okaya A, Barash LF (1962) The dielectric microwave resonator. Proc IRE 50:20812092
Chapter 3
Mathematical Analysis of Rectangular
DRA
Abstract This chapter, mathematical analysis of electromagnetic elds in rectan-

gular dielectric resonator antenna (RDRA) has been introduced. The investigations
are based on the rst applying waveguide theory, then converting it to resonator by
replacing c to d/dz. Initially, these elds are exploited using the Maxwell curl
equations, then manipulating them to express the transverse components of the
elds in terms of partial derivatives of the longitudinal components of the elds
with respect to x and y axis (i.e., the transverse coordinates). Waveguide models of
four rectangular DRAs with specied boundary conditions with linear permittivity
have been realized.
Keywords Electromagnetic elds mathematical modeling Resonator

Waveguide Homogeneous medium Boundary conditions Surface interface
In this chapter, mathematical analysis of electromagnetic elds in rectangular

dielectric resonator antenna (RDRA) has been introduced. The investigations are
based on rst applying waveguide theory, then converting it to resonator by
replacing c to d/dz. Through out this book, electromagnetic eld propagation has
been taken along z-axis, i.e., exp(z). Initially, these elds are exploited using the
Maxwell curl equations, then manipulating them to express the transverse com-
ponents of the elds in terms of partial derivatives of the longitudinal components
of the elds with respect to x and y axis (i.e., the transverse coordinates).
Waveguide models of four different rectangular DRAs with specied boundary
conditions with homogeneous material having linear permittivity have been
mathematically modeled. The elds are realized to determine TE and TM modes of
propagating elds. These have resulted into different sinecosine combinations.
Propagation of these elds have been split as inside the RDRA and outside RDRA.
The interfacing surface is having two different dielectrics. The solution is developed
as transcendental equation, which purely characterized rectangular DRA frequency
and propagating elds in terms of propagation constants and dominant resonant
frequency. TE modes generation required Hz as longitudinal elds and Ex, Ey, Hx,
and Hy as transverse elds. Excitation is applied along x-axis as partial elds, y-axis
will have xed variation, and z-axis will have desired variation in propagating

34 3 Mathematical Analysis of Rectangular DRA
elds, for example, TE d13 and TE d43 . Similar cases can be developed for other
modes, so as to propagate Ez elds as longitudinal and Ex, Ey, Hx, and Hy as
transverse elds. In this case, Hz shall get vanished because of boundary conditions.
Resonant modes, i.e., amplitude coefcient of these elds Cmnp and Dmnp inside the
DRA can be determined by comparing magnetic energies equal to electrical ener-
gies based on principle orthonormality or law of conservation. The derivation for
the quality factor and radiation pattern have been developed for deeper antenna
analysis.
3.1 Rectangular DRA with Homogeneous Medium
In Rectangular DRA as shown in Fig. 3.1, top and bottom walls of RDRA are PMC
and rest of the other walls are PEC. On magnetic walls (PMC), n E = 0, where
E denotes the electric eld intensity and n denotes the normal to the surface of
the resonator. Similarly, n H = 0 is not necessarily satised at all the surfaces of the
DRA by all the modes. Different resonant modes shall have different electromagnetic
Fig. 3.1 a Rectangular DRA with aperture-coupled feed. b RDRA with input excitation
3.1 Rectangular DRA with Homogeneous Medium 35
eld distribution inside the RDRA, and each mode may provide a different resonant
frequency and radiation pattern, i.e., eigen vector and eigen frequency.
Excitation-based resonant modes can generate desired radiation pattern for different
coverage requirements. By making use of this mechanism, internal as well as
associated external elds distribution can be obtained.
Rectangular DRA is better choice due to flexible aspect ratio, i.e., b/a or d/
a options can generate different modes. The existence of two independent aspect
ratios in a rectangular DRA offers better design flexibility. Assuming the ground
plane to be innitely large, image theory is applied to replace the isolated RDRA by
a grounded resonator of half-size. In this RDRA, two of the six surfaces of the
resonator are assumed to be perfect magnetic walls, while the remaining four are
assumed to be perfect electric walls. Electromagnetic theory is then applied to study
its theoretical analysis, and later three more cases have been developed based on the
different boundary conditions. For example, the elds undergo one half-wave
variation along the dimension a and remains constant along dimension b. They
undergo less than a half-wave variation along z-axis, i.e., variation along DRA
height d. The resonant mode is therefore identied as TEz10 d. The propagation
direction has been assumed in z-direction. TEz310 resonant elds undergo three
half-wave variations along length of DRA a and one half-wave variation along
breadth b, and no variation along height d. To adapt these formulae to an DRA,
we note that the propagation constants along z can be c with the linear combi-
nations of coefcients chosen, so as to meet the boundary conditions at z = 0 and
z = d, i.e., the top and bottom surfaces of the RDRA, which have been taken as PEC
(permanent electrical conducting) walls. On a PEC, the tangential components
(n E = 0) of the electric eld and the normal component (n H = 0) of the
magnetic elds get vanished. While on a PMC wall, by directly, the normal
component of the electric eld (n E = 0) and the tangential components
(n H = 0) of the magnetic eld get vanished.
To compute resonant modes, vector principle of orthonormality on half-wave
Fourier analysis has been applied, i.e., radiated magnetic energies are compared
with applied electrical energies in RDRA. More number of modes along z-axis in
RDRA can be generated either by increasing electrical height d of RDRA or by
increasing excitation resonant frequency. Given below are the two rectangular
DRAs with different congurations shown in Fig. 3.1.
In Fig. 3.1, PMC and PEC walls conguration is labeled. The mathematical
solution is developed based on this conguration. The boundary conditions of
interface walls shall form linear combinations of sinecosine terms. Accordingly,
they will decide whether transverse electric elds or magnetic elds will vanish.
Propagation of longitudinal elds shall depend on the direction of excitation.
Excitation of resonant modes in rectangular boundaries are easier as compared to
cylindrical. Transcendental equation and characteristics equations have been
developed for rectangular DRA. This has provided complete solution of resonant
frequency and propagation constants.
3.2 Rectangular DRA Mathematical Modeling
In this chapter, four different solutions are presented, each RDRA is associated with
different boundaries. The resultant eld formed the resonant modes of different
kinds.
Figure 3.2 described E and H elds pattern forming resonant modes, i.e.,
dominant or higher-order excited modes inside the RDRA.
3.2.1 Model-1
(a) Here, top and bottom walls are assumed as PMC and rest of the other four walls
are PEC as per Fig. 3.1.
Given top and bottom surfaces of RDRA as PMC at z 0; d;
)nH 0
n E 0;
Hy Hx 0;
Ez 0;
Rest of the other four walls are PEC.
n E 0;
n H 0;
x 0; a;
Ey Ez 0;
Hx 0;
Fig. 3.2 E and H elds pattern inside RDRA

3.2 Rectangular DRA Mathematical Modeling 37
At,
y 0; b;
Ex Ez 0;
Hy 0;
From separation of variables (Refer Chap. 2),

1 @Hz 1 @ 2 Ez
Ex 3:1
2
jx 1 kc2 @y jxl @z@x

1 1 @ 2 Ez @Hz
Ey 3:2
2
jx 1 kc2 jxl @z@y @x

1 @Ez 1 @ 2 Hz
Hx 3:3
2
jxl 1 kc2 @y jx @z@x

1 1 @ 2 Hz @Ez
Hy 3:4
jxl 1 kc2 jx @z@y @x
2
Solution of second-order differential equation is given as,
wz X xY yZz
where
X x A1 sin kx x A2 cos kx x 3:5
Y y A3 sin ky y A4 cos ky y 3:6
Z z A5 sin kz z A6 cos kz z 3:7
TE mode Ez 0 and H z 6 0
wHz X xY yZz
substituting Ez 0 in Eq. (3.2) to get, Ey

0@Hz
Ey C ;
@x
or,
Ey C0 X 0 xY yZ z;
Now
X 0 x A1 cos kx x A2 sin kx x;
But at
x 0; a; Ey 0;
) 0 A1 cos kx 0 A2 sin kx 0;
or,
mp
A1 0 and kx ;
a
Similarly from Eq. (3.1)

@Hz0
Ex C ;
@y
or,
Ex C0 X xY 0 yZz:
Now
Y 0 y A3 cos ky y A4 sin ky y;
At,
y 0; b; Ex 0;
) 0 A3 cos ky 0 A4 sin ky 0;
or,
np
A3 0 and ky ;
b
From above equation,

1 @ 2 Hz
0
Hx C ;
jx @z@x
or,
Hx C0 X 0 xY yZ 0 z;
Now
Z 0 z A5 cos kz z A6 sin kz z;
At,
z 0; d Hx 0;
) A5 cos kz 0 A6 sin kz 0 0;
pp
A5 0 and kz ;
d
Hence,
mp np pp
Hz A2 A4 A6 cos x cos y cos z 3:8
a b d
Using Eqs. (3.1)(3.4), and (3.8), we get

mppp mp np pp
Hx C 00 A2 A4 A6 sin x cos y sin z ;
a d a b d
np pp mp np pp
Hy C 00 A2 A4 A6 cos x sin y sin z ;
b d
mp a b d
mp np pp
Ey C 00 A2 A4 A6 sin x cos y cos z ;
a a
np mp np pp b d
Ex C 00 A2 A4 A6 cos x sin y cos z ;
b a b d
Similarly, for TM mode H z 0 and Ez 6 0
wEz X xY yZ z;
At,
x 0; a;
Ez 0;
A1 sin kx 0 A2 cos kx 0 0;
mp
) A2 0 and kx ;
a
Also, at
y 0; b;
Ez 0;
A3 sin ky 0 A4 cos ky 0 0;
np
A4 0; and ky
b
At,
z 0; d;
Ez 0;
) A5 sin kz 0 A6 cos kz 0 0;
pp
A6 0 and kz ;
d
Hence,
mp np pp
Ez A1 A3 A5 sin x sin y sin z 3:9
a b d

mppp mp np pp
Ex C 00 A1 A3 A5 cos x sin y cos z ;
a d a b d
np pp mp np pp
Ey C 00 A1 A3 A5 sin x cos y cos z ;
b d
mp a b d
mp np pp
Hy C 00 A1 A3 A5 cos x sin y sin z ;
a a
np mp np pp b d
Hx C 00 A1 A3 A5 sin x cos y sin z ;
b a b d
3.2.2 Model-2
(b) Top and bottom walls are PEC and rest of the other walls are PMC:
Assuming the top and bottom surface plane be at z 0; d;
n E 0;
n H 0;
Ey Ex 0;
Hz 0;
Rest of the other walls are PMC
n H 0;
n E 0;
At,
x 0; a;
Hy Hz 0;
Ex 0;
At,
y 0; b;
Hx Hz 0;
Ey 0;
We also know that

1 @Hz 1 @ 2 Ez
Ex ; 3:10
2

1 1 @ 2 Ez @Hz
Ey ; 3:11
2

1 @Ez 1 @ 2 Hz
Hx ; 3:12
2

1 1 @ 2 Hz @Ez
Hy ; 3:13
2
Now, the solution of second-order differential equation is given as
wz X xY yZz;
where
X x A1 sin kx x A2 cos kx x; 3:14
Y y A3 sin ky y A4 cos ky y; 3:15
Z z A5 sin kz z A6 cos kz z; 3:16
TE mode Ez 0 and H z 6 0
wH z X xY yZ z;
At,
x 0; a;
Hz 0;
A1 sin kx 0 A2 cos kx 0 0;
A2 0;
and
mp
kx ;
a
Also, at
y 0; b;
Hz 0;
A3 sin ky 0 A4 cos ky 0 0;
A4 0;
and
np
ky ;
b
At,
z 0; d;
Hz 0;
A5 sin kz 0 A6 cos kz 0 0;
A6 0;
and
pp
kz ;
d
Hence,
mp np pp
Hz A1 A3 A5 sin x sin y sin z ; 3:17
a b d

mppp mp np pp
Hx C 00 A1 A3 A5 cos x sin y cos z ;
a d a b d
np pp mp np pp
Hy C 00 A1 A3 A5 sin x cos y cos z ;
b d
mp a b d
mp np pp
Ey C 00 A1 A3 A5 cos x sin y sin z ;
a a
np mp np pp b d
Ex C 00 A1 A3 A5 sin x cos y sin z ;
b a b d
TM mode H z 0 and Ez 6 0
wEz X xY yZ z;
From Eq. (3.2) after substituting Hz 0, we get

1 @ 2 Ez
0
Ey C ;
jxl @z@y
Ey C0 X xY 0 yZ 0 z;
Now
Y 0 y A3 cos ky y A4 sin ky y;
At,
y 0; b; Ey 0;
0 A3 cos ky 0 A4 sin ky 0;
np
A3 0 and ky ;
b
Similarly, from the above equations,

01 @ 2 Ez
Ex C ;
jxl @z@x
Ex C 0 X 0 xY yZ 0 z;
x 0; a;
Ex 0;
0 A1 cos kx 0 A2 sin kx 0;
A1 0;
and
mp
kx ;
a
Also, from above equations,
Z 0 z A5 cos kz z A6 sin kz z;
At,
z 0; d; Ex 0;
) 0 A5 cos kz 0 A6 sin kz 0;
or,
pp
A5 0 and kz ;
d
Hence,
mp np pp
Ez A2 A4 A6 cos x cos y cos z ; 3:18
a b d
Using Eqs. (3.113.14) and (3.18), we get

mppp mp np pp
Ex A2 A4 A6 sin x cos y sin z ;
a d a b d
np pp mp np pp
Ey A2 A4 A6 cos x sin y sin z ;
np mp np pp d
b d a b
Hx A2 A4 A6 cos x sin y cos z ;
b a b d
mp mp np pp
Hy A2 A4 A6 sin x cos y cos z ;
a a b d
3.2.3 Model-3
(c) Solution of RDRA, when all six walls are PEC (perfect electrical walls):
Using Maxwell equations:
r E jxB jxlH;
r H jxE;
r E jxlH;

^x ^y ^z

@ @ @
@x @y
@z jxlH;
Ex Ey Ez

@Ez @Ey @Ex @Ez @Ey @Ex
^x ^y ^z jxlH;
@y @z @z @x @x @y
On comparing x; y; z components both the sides
@Ez @Ey
jxlHx ; 3:19
@y @z
@Ex @Ez
jxlHy ; 3:20
@z @x
@Ey @Ex
jxlHz ; 3:21
@x @y
Similarly, using r H jxE; We get
@Hz @Hy
jxEx ; 3:22
@y @z
@Hx @Hz
jxEy ; 3:23
@z @x
@Hy @Hx
jxEz ; 3:24
@x @y
Comparing above equations,

1 @Hz 1 @ 2 Ex @ 2 Ez
Ex ; 3:25
jx @y jxl @z2 @x@z

1 @ 2 Ex 1 @Hz 1 @ 2 Ez
Ex 2 2 ; 3:26
k @z jx @y jxl @x@z

c2 1 @Hz c @Ez
Ex 1 2 ;
k jx @y jxl @x
Ey, Hx, and Hy are expressed in Ez and Hz elds:

c2 1 c @Ez @Hz
Ey 1 2 ;
k jx jxl @y @x

c2 1 @Ez c @Hz
Hx 1 2 ;
k jxl @y jx @x

c2 1 c @Hz @Ez
Hy 1 2
k jxl jxl @y @x
Separation of variables with given boundary conditions, solution is obtained.
w X xY yZ z;

A1 sin kx x A2 cos kx x A3 sin ky y Ay sin ky y A5 sin kz z A6 cos kz z;
TM mode of propagation, H z 0;
Boundary conditions
Electrical walls ! Etan 0 n E;
! Hn 0 n H;
At, x = 0;
Ez X x Az cos0; so A2 must be zero:

y 0; Y y Ay cos0; A4 must be zero:
For standing wave in direction of z,

Therefore,
@
Z z 0;
@z
A5 coskz z A6 sinkz z 0;
Therefore, at
z 0; d;
A5 must be zero;
Hence, we are left with
Ez A1 ; A3 ; A5 ; sin kx x sin ky y cos kz z;
Next, boundary conditions are

At,
x a;
X x A1 sin kx a 0;
mp
kx ;
a
At,
Y b; Y y A2 sin ky d 0;
np
ky ;
b
At,
z 0 zz A4 sin kz d 0;
pp
kz ;
d
As, we know that
k02 kx2 ky2 kz2 ;

1 @Hz 1 @ 2 Ez
Ex ;
jx 1 ky2
2
@y jxl @x@z
Ez 0;
1
Ex A1 A3 A5 kx kz cos kx x sin ky y sin kz z ;

y2
1 k2
k2

1 @Hz 1 @ 2 Ez
Ey ;
2
jx 1 ky2 @x jxl @y@z
A1 A3 A5 ky
Ey kz sin kx x cos ky y sin kz z ;

y2
k 1 k2
2

1 @Ez 1 @ 2 Hz
Hx ;
2
jxl 1 ky2 @y jx @x@z
ky A1 A3 A5
2
sin kx x cos ky y cos kz z ;
x2 l 1 k2
y

1 1 @ 2 Hz @Ez
Hy ;
2
jxl 1 ky2 jx @y@z @x
kz kx A1 A3 A5
2
cos kx x sin ky y cos kz z ;
x2 l 1 ky2
kx A1 A3 A5
Hy cos kx x sin ky y cos kz z ;

y2
jxl 1 k2
For, TE mode

w A1 sin kx x A2 cos kx A3 sin ky y Ay sin ky A5 sin kz z A6 cos kz z;
For PEC walls, electric eld components are assumed to be varying with Hz in
direction of (x, y, z)
@
Ex C 0 Hz
@y
C0 X xY 0 yZ z;
y 0; b;
Y 0 y A3 cos ky y A4 sin ky y 0;
A3 0;
np
ky ;
b
@
Similarly, Ey C00 Hz ;
@x
A1 0;
mp
kx ;
a
Z z A5 sin kz z A6 cos kz z;
At,
z 0; d;
A6 0;
pp
kz ;
d
Hz A2 A4 A5 cos kx x cos ky y sin kz z;
Therefore,
1
Ex 2
A2 A4 A5 cos kx x sin ky sin kz;
jx 1 ky2
A2 A4 A5 kx
Ey 2
sin kx x cos ky y sin kz z ;
jx 1 ky2
kx kz A1 A3 A5
Hx 2
sin kx x cos ky y cos kz z ;
k2 1 k2 y
kz ky A1 A3 A5
Hy 2
cos kx x sin ky y cos kz z ;
x2 l 1 ky2
3.2.4 Model-4
(d) When all the six walls of RDRA are assumed to be PMC (permanent magnetic
walls),
wz X xY yZ z where wz is wave function in x, y, and z direction as space.

Or A1 sin kx x A2 cos kx x A3 sin ky y A4 cos ky y A5 sin kz z A6 cos kz z

3:27
where A1A6 are constants and (A1 sin kx x + A2 cos kx x is solution of second-order
differential equation in x direction, i.e., X(x).
When all six walls are PMC, the rectangular DRA solution is
Htan n H 0;
Hnor n E 0;
Applying boundaries,
At,
x 0; a ) Hy and Hz 0; Ex 0;
At,
y 0; b ) Hx and Hz 0; Ey 0;
At,
z 0; d ) Hx and Hy 0; Ez 0;
TE mode of propagation Ez 0; H z 6 0
Using boundary conditions
At,
mp
x 0; a; Hz 0 ) A2 0 and kx ;
a
At,
np
y 0; b; Hz 0 ) A4 0 and ky ;
b
Now,
@ 2 Hz
Hx C 00 C 00 X 0 xY yz0 z
@x@z
z0 z A5 cos kz z A6 sin kz z
At,
z 0; d ) d ) Hx 0;
pp
) Hx 0 ) A5 0; kz ;
d
Hence,
mpx npy ppz
Hz A1 A3 A6 sin sin cos 3:28
a b d
TM mode of propagation Ez 6 0; H z 0
We again look for the conditions, when Hz = 0, i.e., to get the value of Ez
C 0 @Ez
Hz
@y
C X xY y0 Z z;
0
Y 0 y A3 cos ky y A4 sin ky y; 3:29

Hx 0 at Y 0; b;
np
) A3 0 at ky ;
b
Similarly,
@Ez
Hy C 00 ;
@x
C 00 X 0 xY y0 Z z;
) Hy 0 at x 0; a;
) A1 0;
mp
kx ;
a
At,
z 0; d ) Ez 0;
pp
) A5 0 and kz ;
d
At,
z 0; d ) Ez 0;
pp
) A5 0 and kz ;
d
Hence,
mpx npy ppz
Ez A2 A4 A5 cos cos sin : 3:30
a b d
3.2.5 Basic Theory
Depending on the nature of the surfaces, different linear combinations of the c

modes are formed. The propagation constant c itself is taking discrete values. This
forces the natural frequencies of the eld oscillations to take discrete values mnp,
indexed by three positive integers m, n, and p. The solutions of the waveguide
problem yield discrete values of c, i.e., cm; n; x for a given frequency x by
applying boundary conditions to the electromagnetic elds on the side walls. The
corresponding eld amplitudes are solutions to the 2-D Helmholtz equations
corresponding to the transverse Laplacian r2? : These amplitudes are called the
waveguide modes and are of the form
I n npx nnpxoo I n
mpy nmpyoo
L cos ; sin L cos ; sin
a a b b
where L denotes linear components. It turns out that, depending on the nature of
wall surfaces (PEC or PMC), four possible linear combinations can appear
cos sin; sin cos; sin sin; and cos cos.
In rectangular DRA, we have got to applying in additional boundary conditions
on top and bottom surfaces to be the linear combinations as compared to the
waveguide.
C1 expfcm; n; xzg C2 expfcm; n; xzg
and these cases are cm; n; x pp

d ; when p 1; 2; 3. . . and have two possible

ppz
linear combinations of sin ppz

d and cos d .
Thus, the possible frequencies x obtained by solving cm; n; x pp

d and then
comes out to be
2 1=2
m n2 p2
xm; n; p p 2 2 2
a b d

@
physics to resonator physics is to just replace c by @z in all the waveguide
components. This is done after solving the full 3-D Helmholtz equations using
separation of variable in x, y, and z.

x2 Ez
r 2
2
0
c Hz
The discrete modes xmnp enable us to visualize the resonator as collection of

L, C oscillators with different L, C values. The outcome of all this analysis enables
us to write down the E and H elds inside the resonator, as superposition of four or
three vector-valued basis functions.
X
1 n o
Ex; y; z; t Re Cmnpe jxmnpt wEmnp x; y; z
m;n;p1
n o 3:31
X
1
E x; y; z
Re Dmnpe jxmnpt / mnp
m;n;p1
and
X
1 n o
H x; y; z; t Re Cmnpe jxmnpt wH
mnp
x; y; z
m;n;p1
n o 3:32
X
1
H x; y; z
Re Dmnpe jxmnpt / mnp
m;n;p1
We note that there are only two sets of amplitude coefcients {C(mnp)} and
{D(mnp)} of linear combination of coefcients using from the Ez and Hz expansions.
The vector-valued complex functions are wEmnp ; / E ; wH ; / H R3 (where R is
mnp mnp mnp
autocorrelation) and contains components fcos; sing fcos; sing fcos; sing;
functions and hence for m0 n0 p0 6 mnp; each function of the set
n o
E ; wH ; /
wEmnp ; / H ;
mnp mnp mnp
is orthogonal to each functions of the set

n o
E ; wH ; /
wEm0 n0 p ; / H ;
0
mnp0 mnp 0
mnp0 0
w.r.t. The measure of dx dy dz over [0, a] [0, b] [0, d];

The exact form of the function / E; /
H ; wE ; wH depends on the nature of the
boundaries. The next problem addressed can be on excitations of RDRA. To cal-
culate the amplitude coefcients {C(mnp)} and {D(mnp)}, we assume that at z = 0,
e e
an excitation Ex x; y; t or Ey x; y; t is applied for some time say t 2 [0, T] and
then removed. Then, the Fourier components in this excitation corresponding to the
frequencies fxmnpg are excited and their solutions are the oscillations, while the
waveguide for t > T. The other Fourier components decay within the resonator.
{C(mnp), D(mnp)} are components of the form,
X
Exe x; y; t ReCmnpe jxmnpt wEmnp x x; y; 0
m;n;p
n o 3:33
E x; y; 0
Re Dmnpe jxmnpt / mnp x
and
X
Eye x; y; t Re Cmnpe jxmnpt wEmnp y x; y; 0
m;n;p
n o 3:34
E x; y; 0
Re Dmnpe jxmnpt / mnp y
n o
E x; y; 0 , for different (m, n), we
By using orthogonality of wEmnp x x; y; 0; /
n mnp x o
E x; y; 0 ;
can write p to be xed and likewise of wEmnp y x; y; 0; / mnp y
In addition, we need to use KAM (KolmogorovArnoldMoser) type of time
averaging to yield
E x; y; 0
CmnpwEmnp x x; y; 0 Dmnp/ mnp x
ZT
lim 1
Exe x; y; tejxmnpt dt
T!1 2T
T
and likewise

CmnpwEmnp y x; y; 0 Dmnp/
E
x; y; 0
mnp y
ZT
lim 1
Eye x; y; te jxmnpt dt
T ! 1 2T
T
Chapter 4
Mathematical Analysis of Transcendental
Equation in Rectangular DRA
Abstract Mathematical analysis of transcendental equation in rectangular DRA

has been derived. Transcendental equation of rectangular DRA provides complete
solution of propagation constants, i.e., kx, ky, and kz. The propagation constant gives
rise to resonant frequency with the help of characteristic equation. The wave
numbers kx, ky, and kz are in x, y, and z direction, respectively. The free space wave
number is k0. The exact solution of RDRA resonant frequency can be determined
from combined solution of transcendental equation and characteristic equation.
These equations have unique solution. RDRA depends upon boundary conditions.
MATLAB-based simulation has been worked for RDRAs. They have been depicted
with examples. This chapter has given a complete design solution of rectangular
DRAs.

Keywords Mathematical analysis Transcendental equation Rectangular DRA

Propagation constant Eigen vectors Effective electrical length Characteristic
equation
Transcendental equation of rectangular DRA provides complete solution of prop-

agation constants, i.e., kx, ky, and kz. The propagation constant gives rise to resonant
frequency with the help of characteristic equation. The wave numbers kx, ky, and kz
are in x, y, and z direction, respectively. The free space wave number is k0. The
exact solution of RDRA resonant frequency can be determined from combined
solution of transcendental equation and characteristic equation. These equations
have unique solution if RDRA boundary conditions are xed. For example, top and
bottom walls are PMC and rest of the four walls is PEC and vice versa, only two
different transcendental equations will be developed.
To get this solution, Hz elds and derivative of Hz elds need to be solved. They
are solved for continuous propagating elds conditions. The elds are assumed
continuous at interface of RDRA. The RDRA along with eigen vectors is shown in
Fig. 4.1a, b.
CASE#1 RDRA solution:
See Fig. 4.2.

58 4 Mathematical Analysis of Transcendental
Fig. 4.1 a Rectangular DRA. b Eigen currents (current vectors) versus wavelength
Fig. 4.2 RDRA under dened boundaries

4 Mathematical Analysis of Transcendental 59
To derive transcendental equation, the elds inside the resonator and outside the
resonator are required.
kz
tankz d q transcendental equation 4:1
r 1k0 2 kz 2
r k0 2 kx 2 ky 2 kz 2 wave equation 4:2
kx mp=a 4:3a
ky np=b 4:3b
kz pp=d 4:3c
where a, b, and d are dimensions; m, n, and p are the indices.

TE11, TE11, and TE11 are dominant modes.
The solution of resonant frequency can be had if solution of ky propagation
constant is obtained from characteristic equation, r k0 2 kx 2 ky 2 kz 2 , and then
substituted in transcendental equation to compute resonant frequency f0.
Boundary condition
Propagation constant, c2mn k0 2 hmn 2
p
k 2p=k x l x=c;
Z Z
E dV
2
H 2 dV
Time average electric energy = time average magnetic energy

r k0 2 kx 2 ky 2 kz 2 4:4
0 k0 2 kx 2 ky 2 kz 0
2
4:5
kz pp=d
Subtracting Eq. (4.1) from Eq. (4.2), we get
k z 0 kz 2 0 k0 0 r k0 2
2 2
kz 02 kz 2 0 l0 x2 r l0 x2
Taking value of 0 1 and l0 l, we get
kz 02 kz 2 x2 l1 r 4:6
4.1 Case-1: Top and Bottom Walls as PMC and Rest

of the Four Walls are PEC
See Fig. 4.3.

Assuming that the top and bottom surface plane be at z 0; d to be PMC
) nH 0
And
nE 0
or,
Hy Hx 0
Ez 0
Rest of the other walls is PEC
) nE 0
And
nH 0
At
x 0; a Ey Ez 0
Hx 0
Fig. 4.3 RDRA with boundaries

4.1 Case-1: Top and Bottom Walls as PMC and Rest of the Four Walls are PEC 61
At
y 0; b Ex Ez 0
Hy 0
We also know

1 @Hz 1 @ 2 Ez
Ex 4:7a
jx 1 kc2
2
@y jxl @z@x

1 1 @ 2 Ez @Hz
Ey 4:7b
2

1 @Ez 1 @ 2 Hz
Hx 4:7c
2

1 1 @ 2 Hz @Ez
Hy 4:7d
2
Now, the solution of second-order differential equation is given as follows:
wz X xY yZz
where
X x A1 sin kx x A2 cos kx x 4:8a
Y y A3 sin ky y A4 cos ky y 4:8b
Z z A5 sin kz z A6 cos kz z 4:8c
For TE mode Ez 0 and Hz 6 0
wHz X xY yZ z
After putting Ez 0; we get

@Hz
0
Ey C
@x
or,
Ey C 0 X 0 xY yZ z
Now
X 0 x A1 cos kx x A2 sin kx x
But at
x 0; a Ey 0
) 0 A1 cos kx 0 A2 sin kx 0
or,
mp
A1 0 and kx
a
Similarly,

@Hz
Ex C 0
@y
or,
Ex C0 X xY yZz
Now
Y 0 y A3 cos ky y A4 sin ky y
But at
y 0; b Ex 0
) 0 A3 cos ky 0 A4 sin ky 0
or,
np
A3 0 and ky
b
from

01 @ 2 Hz
Hx C
jx @z@x
or,
Hx C 0 X 0 xY yZ 0 z
Now
Z 0 z A5 cos kz z A6 sin kz z
At
z 0; d Hx 0
) A5 cos kz 0 A6 sin kz 0 0
pp
A5 0 and kz
d
Hence,
mp np pp
a b d
Using Eqs. (4.1)(4.4), and (4.8a)(4.8c), we get

mppp mp np pp
Hx C00 A2 A4 A6 x cos
sin y sin z 4:10a
a d a b d
nppp mp np pp
Hy C 00 A2 A4 A6 cos x sin y sin z 4:10b
b d a b d
mp mp np pp
Ey C00 A2 A4 A6 sin x cos y cos z 4:10c
a a b d
np mp np pp
Ex C00 A2 A4 A6 cos x sin y cos z 4:10d
b a b d
Now, evaluate Ex and Hz at the boundary walls of the dielectric waveguide.

As we know that at the PMC wall, the tangential component of magnetic eld
and normal component of electric eld are equal to zero at the interface z = 0, d.
Hence,
Hx ; Hy 0
and
Ez 0
Also, for propagation to be possible, we need two normal components of E and H.

Thus, we take Ex and Hy :
Now, the propagating wave is continuous at the interface, i.e., Ex Ex0 :
Therefore,
mp np mp np 0
A cos x sin y C1 e jkz z C2 ejkz z A cos x sin y C20 ejkz z
a b a b
4:11
or,
0
C1 e jkz z C2 ejkz z C20 ejkz z 4:12
But at z = 0, only the inside waveform exists.

Therefore,
C1 e jkz z C2 ejkz z 0
Now substituting the value of z = 0, we get
C1 C2 0
4:13
or; C1 C2
As Hz is continuous at the interface z = d,
@Hz @Hz0
Hz Hz0 and
@x @x
From Eq. (4.9),

mp np
Hz B cos x cos y coskz z 4:14a
a b
and
mp np
Hz0 B cos x cos y cos kz0 z 4:14b
a b
Equating Eqs. (4.14a) and (4.14b), we get

mp np mp np 0

B cos x cos y C1 e jkz z C2 ejkz z B cos x cos y C20 ejkz z
a b a b
or,
0
From Eq. (4.15), i.e., C1 C2 , we get, at z d;
2jC1 sinkz d C02 ejkz d 4:16
Now, equating the derivative of Hz ; we get

0
jkz C1 e jkz z C2 ejkz z jkz0 C20 ejkz z 4:17
or,
0
2kz C 1 coskz d kz0 C 02 ejkz z :
Dividing equation (4.16) by (4.17)
j tan kz d 1
0 4:18a
kz kz
Squaring both sides and substituting the value of kz2 from Eq. (4.3c),
kz 02 kz2 x2 lr 1
and substituting l 1: We get,
kz
tankz d q 4:18b
k02 r 1 kz2
The above equation is the required transcendental equation.
4.2 Case-2
For transcendental equation, we need to compute the elds inside the resonator and
outside it.
kz
tankz d q 4:19
r 1k02 kz2
where r k02 kx2 ky2 kz2 (characteristic wave equation)
mp
kx 4:20a
a
np
ky 4:20b
b
pp
kz 4:20c
d
where a, b, and d are dimensions; m, n, and p are modes.

TE11, TE11, and TE11 are dominant modes.
Boundary condition
p x
Propagation constant, c2mn k02 hmn2 where k 2p
k x l c :
From the energy conservation principle,

Z Z
E 2 dV H 2 dV:
i.e., time average electric energy = time average magnetic energy.

When top and bottom walls are PMC, rest of the other walls is PEC
Assuming that the top and bottom surface plane be at z 0; d
) nH 0
And
nE 0
or,
Hy Hx 0
Ez 0
Rest of the other walls is PEC
) nE 0
And
nH 0
At
x 0; a Ey Ez 0
Hx 0
4.2 Case-2 67
At
y 0; b Ex Ez 0
Hy 0
We also know

1 @Hz 1 @ 2 Ez
Ex 4:21a
2

1 1 @ 2 Ez @Hz
Ey 4:21b
2

1 @Ez 1 @ 2 Hz
Hx 4:21c
2

1 1 @ 2 Hz @Ez
Hy 4:21d
2
wz X xY yZz 4:22
where
X x A1 sin kx x A2 cos kx x
Y y A3 sin ky y A4 cos ky y
Z z A5 sin kz z A6 cos kz z
For TE mode Ez 0 and Hz 6 0
wHz X xY yZz
Ez 0
we get

@Hz
0
Ey C
@x
or,
Ey C 0 X 0 xY yZ z
Now
X 0 x A1 cos kx x A2 sin kx x
But at
x 0; a Ey 0
) 0 A1 cos kx 0 A2 sin kx 0
or,
mp
A1 0 and kx
a
Similarly,

@Hz
Ex C 0
@y
or,
Ex C 0 X xY 0 yZ z
Now
Y 0 y A3 cos ky y A4 sin ky y
But at
y 0; b Ex 0
) 0 A3 cos ky 0 A4 sin ky 0
or,
np
A3 0 and ky
b

0 1 @ 2 Hz
Hx C
jx @z@x
4.2 Case-2 69
or,
Hx C 0 X 0 xY yZ 0 z
Now
Z 0 z A5 cos kz z A6 sin kz z
At
z 0; d Hx 0
) A5 cos kz 0 A6 sin kz 0 0
pp
A5 0 and kz
d
Hence,
mp np pp
a b d

mppp mp np pp
Hx C00 A2 A4 A6 sin x cos y sin z 4:24a
a d a b d
nppp mp np pp
Hy C 00 A2 A4 A6 cos x sin y sin z 4:24b
b d a b d
mp mp np pp
Ey C 00 A2 A4 A6 sin x cos y cos z 4:24c
a a b d
np mp np pp
Ex C00 A2 A4 A6 cos x sin y cos z 4:24d
b a b d
Above equations can also be written as follows:
kx kz
Hx sinkx x cosky y sinkz z
jxl0
ky kz
Hy coskx x sinky y sinkz z
jxl0
Ey kx sinkx x cosky y coskz z
Ex ky coskx x sinky y coskz z
kx2 ky2
Hz coskx x cosky y coskz z
jxl0
Since Hz is continuous, i.e., ddHzz 6 0;

kx2 ky2
Hz0 coskx x cosky y cosk 0z z
jxl0
Now, Hy can be written as follows:
ky kz
Hy coskx x sinky yC1 e jkz d C2 ejkz d
jxl0
But
Hy 0 at d 0
C1 C2 0
or,
C1 C2
dHy
A0 jkz coskx x sinky yC1 e jkz d C2 ejkz d
dz
or,
dHy
C1 jkz coskx x sinky ye jkz d ejkz d
dz
or,
dHy
C1 2jkz coskx x sinky y coskz d
dz
0
Hy0 C10 coskx x sinky yejkz d outside the cavity
For Hz to be continuous,
dHy dHy0

dz dz
or,
0
C1 2jkz coskx x sinky y coskz d jkz0 C10 coskx x sinky yejkz d
4.2 Case-2 71
or,
0
2C1 kz coskz d kz0 C10 ejkz d 4:25
From above equations, we have

Ex ky coskx x sin ky y C1 e jkz d C2 ejkz d
At
d 0; Ex 0;
so,
C1 C2 0
or,
C1 C2

) Ex ky coskx x sin ky y C1 e jkz d ejkz d
or,

Ex 2jC1 ky coskx x sin ky y sinkz d
Also

Ex0 ky coskx x sin ky y cos kz0 z
or,

Ex0 C10 ky coskx x sin ky y ejkz d
For Hz to be continuous,
Ex Ex0
or,
jC1 ky coskx x sinky y sinkz d C10 ky coskx x sinky yejkz d
or,
2jC1 sinkz d C10 ejkz d 4:26

Dividing Eq. (4.16) by Eq. (4.25), we get
2C1 j sinkz d C0 ejkz d

0 10 jk0 d
2C1 kz coskz d kz C1 e z
or,
j tan kz d 1
0
kz kz
or,
kz
j tan kz d
kz0
On squaring and putting kz0 2 kz2 x2 l0 1 r
kz2
tan2 kz d
kz2 x2 l0 1 r
or,
kz2
tan2 kz d
x2 l0 r 1 kz2
kz
tan kz d q 4:27
r 1k02 kz2
With the help of transcendental equation, we can nd the propagation factor. Also
with the help of this equation, we can obtain resonant frequency.
CASE#3
For transcendental equation, we need to compute the elds inside the resonator and
outside it.
q
kz tankz d r 1k02 kz2 ;
r k02 kx2 ky2 kz2 ;
and
kx mp=a 4:28a
ky np=b 4:28b
4.2 Case-2 73
kz pp=d 4:28c
where a, b, and d are dimensions; m, n, and p are the indices.
TE11, TE11, TE11 are dominant modes.

Boundary conditions
Propagation constant, c2mn k02 h2mn
p
k 2p= k x 2 l x=c;
Z Z
E dV
2
H 2 dV
Time average electric energy = time average magnetic energy
0 r k02 kx2 ky2 kz2 4:29a
0 k02 kx2 ky2 kz02 4:29b
kz0 6 pp=d
Subtracting Eq. (4.1) from Eq. (4.2), we get
kz02 kz2 0 k002 0 r k02

kz02 kz2 0 l0 x2 0 r l0 x2
Taking the value of 0 1 and l0 l; we get
kz02 kz2 x2 l0 1 r 4:30
When top and bottom walls are PEC, rest of the other walls is PMC.
Now,
Assuming that the top and bottom surface plane be at z 0; d
) nE 0
and
nH 0
or,
Ey Ex 0
Hz 0
Rest of the other walls is PMC
) nH 0
And
nE 0
At
x 0; a Hy Hz 0
Ex 0
At
y 0; b Hx Hz 0
Ey 0
We also know

1 @Hz 1 @ 2 Ez
Ex 4:31a
2

1 1 @ 2 Ez @Hz
Ey 4:31b
2

1 @Ez 1 @ 2 Hz
Hx 4:31c
2

1 1 @ 2 Hz @Ez
Hy 4:31d
2
wz X xY yZ z
4.2 Case-2 75
where
X x A1 sin kx x A2 cos kx x
Y y A3 sin ky y A4 cos ky y
Z z A5 sin kz z A6 cos kz z
(i) For TE mode Ez 0 and Hz 6 0
wHz X xY yZ z
At
x 0; a Hz 0;
or,
A1 sin kx 0 A2 cos kx 0 0
mp
) A2 0 and kx
a
Also at
y 0; b Hz 0
or,
A3 sin ky 0 A4 cos ky 0 0
np
) A4 0 and ky
b
At
z 0; d Hz 0
) A5 sin kz 0 A6 cos kz 0 0
pp
A6 0 and kz
d
Hence,
mp np pp
H z A1 A3 A5 sin x sin y sin z 4:32
a b d

mppp mp np pp
Hx C 00 A1 A3 A5 cos x sin y cos z 4:33a
a d a b d
nppp mp np pp
Hy C00 A1 A3 A5 sin x cos y cos z 4:33b
b d a b d
mp mp np pp
Ey C00 A1 A3 A5 cos x sin y sin z 4:33c
a a b d
np mp np pp
Ex C 00 A1 A3 A5 sin x cos y sin z 4:33d
b a b d
Now, evaluate Hx and Hz at the boundary walls of the dielectric waveguide.

As we know that at the PEC wall, the tangential component of electric eld and
normal component of magnetic eld is equal to zero at the interface z 0; d:
Hence,
Ex ; Ey 0
and
Hz 0
Also, for propagation to be possible, we need two normal components of E and H.

Thus we take Ey and Hx :
Now, the propagating wave is continuous at the interface, i.e., Hx Hx0 :
Therefore,
mp np mp np 0
A cos x sin y C1 e jkz z C2 ejkz z A cos x sin y C20 ejkz z
a b a b
or,
0
But at z 0, only the inside waveform exists.

Therefore,
C1 e jkz z C2 ejkz z 0
Now, substituting the value of z 0; we get
C1 C2 0
4.2 Case-2 77
or,
C1 C2 4:35
As Hz is continuous at the interface z d:

Therefore,
@Hz @Hz0
Hz Hz0 and
@x @x
mp np
H z B sin x sin y sinkz z 4:36a
a b
and
mp np
H 0 z B sin x sin y sink0 z z 4:36b
a b
Equating Eqs. (i) and (ii), we get

mp np mp np 0
B sin x sin y C1 e jkz z C2 ejkz z B sin x sin y C20 ejkz z
a b a b
or,
0
C1 e jkz z C2 ejkz z C20 ejkz z
From Eq. (1b), i.e., C1 C2 ; we get, at z d;
2C 1 coskz d C 02 ejkz d 4:37
Now, equating the derivative of Hz ; we get

0
jkz C1 e jkz z C2 ejkz z jkz0 C20 ejkz z
or,
0
2jkz C1 sinkz d kz0 C 02 ejkz z 4:38
Dividing equation (iv) by (iii), we get
jkz tan kz d kz0

Squaring both sides and substituting the value of kz02 from Eq. (4.3c), we get
kz0 kz2 x2 lr 1
and substituting l 1, we get isolated DRA case as:

q
kz tankz d r 1k02 kz2 4:39a
DRA with ground plane case as:

q
kz tankz d=2 r 1k02 kz2 4:39b
Hence, the solution of transcendental equation is completely obtained.
4.3 MATLAB Simulation Results
The same can be seen if MATLAB simulation is obtained as given below:
clear
clear all
er=9.8;
c=3*10^8;
d=10*10^-3;
for p=1:1:10
f=c*p*(sqrt(1+tan(p*pi/2).^2))/2*d*(sqrt(er-1));
end
plot(p,f);
title('pvsf')
xlabel('p-------->>>');
ylabel('f-------->>');
grid on;
Relationship between delta distance and its impact on

resonant frequency is shown in Fig. 4.4.
The resonant frequency is increasing as the delta length is increasing as shown in
Fig. 4.4. Also, radiation lobe is increasing as the number of resonant mode is
increasing as shown in Fig. 4.5.
4.3 MATLAB Simulation Results 79
frequency vs pdelta
3.00E+07
frequency vs pdelta
frequency-------->>>
2.50E+07
2.00E+07
1.50E+07
1.00E+07
5.00E+06
0.00E+00
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
pdelta---------->>>
Fig. 4.4 Frequency versus delta distance
Radiation Lobes: RDRA dimensions are given to compute resonant modes

using MATLAB.
Fig. 4.5 Radiation lobes of radiation pattern in RDRA
MATLAB Program for Ez eld
m=5;
n=4;
p=3;
a=10;
b=5;
c=2;
x=linspace(-5,5,51);
y=linspace(-2.5,2.5,51);
z=linspace(-1,1,51);
[xi,yi,zi] = meshgrid(x,y,z);
Ez= (cos(m*pi*xi/a).*cos(n*pi*yi/b)).*sin(p*pi*zi/c);
Ez= Ez.^2;
Ez= sqrt(Ez);
xslice = -4.5; yslice = -2.5; zslice =1;
slice(xi,yi,zi,Ez,xslice,yslice,zslice)
colormap hsv
MATLAB program for transcendental equation and resonant frequency of

RDRA:
d=9;
w=6;
h=7.6;
c=3e8;
cons=9.8;
syms y real
kx=pi/d;
kz=pi/2/h;
ko=sqrt((kx^2+y^2+kz^2)/cons);
f=real(y*tan(y*w/2)-sqrt((cons-1)*ko^2-y^2));
ky=fzero(inline(f),[0,(pi/w)-0.01]);
fresonance = c/2/pi*sqrt((kx^2+ky^2+kz^2)/cons)/1e7;
The MATLAB-simulated resonant modes in Figs. 4.6, 4.7, 4.8, 4.9, 4.10, 4.11
and 4.12 have been drawn, and resonant frequency using transcendental equation is
placed in table form.
1
0.9
0.8
0.7
0.6
z-axis
0.5
0.4
0.3
0.2
5
0.1
0 y-axis
2.5 0
2 1.5 1
0.5
x- axis 0
-0.5
-1 -1.5 -2 -5
-2.5
TE12z
Fig. 4.6 Resonant modes in xy plane

1
0.9
0.8
0.7
0.6
0.5 z
0.4
0.3
0.2
0.1
0
3
2
1
x y 3 4 5
0
1 2
-1 -1 0
-2 -3 -2
-3 -5 -4
TE 22z
Fig. 4.7 Resonant modes in xy plane
1
0.8
0.6
0.4
0.2
z
0
2.5 5
2 4
1.5 3
1 2
0.5 1
0
-0.5 -1
0
y
x
-1
-1.5
-2 -4
-3
-2
-2.5 -5
TE33z
Fig. 4.8 Resonant modes in RDRA in xy plane
Solved examples of RDRA resonant frequency:

Example 1 Calculate the dimension of d in RDRA:
For TE111 mode when
r 100
a 10 mm
b 10 mm
fr 7:97 GHz
1
0.9
0.8
0.7
0.6
0.5
0.4
z
0.3
0.2
5
0.1
0
2.5 2 1.5
0 y
1 0.5 0 -0.5
x -1
-1.5 -2 -2.5 -5
TE12z
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3 z
0.2
0.1
0
3
2
1 4 5
0 2 3
1
0
-1
x -2
-4 -3 -2 -1
y
-3 -5
TE22z
Fig. 4.9 Resonant modes in RDRA in xy plane
TE222
0.5
-0.5
z
-1
2.5
2
1.5 5
1 4
0.5 3
2
0
-0.5 0 y 1
-1 -1
x -1.5
-3
-2
-2 -4
-2.5 -5
Fig. 4.10 Resonant modes 3D in RDRA in xyz plane

Fig. 4.11 TE 341 resonant modes
Fig. 4.12 TE 323 resonant modes

Solution Resonant frequency:

r
c m2 n2 p2
fr p
2 r a b d
s
3 108 12
7:97 109 p 1002 1002
2 100 d
s
12
531:33 20000
d
1
512:167
d
d 1:95 mm
Example 2 RDRA with following data:
r 35
a 18 mm
b 18 mm
fr 2:45 GHz
Solution
r
c m2 n2 p2
fr p
2 r a b d
s
3 108 10002 10002 12
2:45 109 p
2 35 18 18 d
s

2
2
1000 1
9337:222 2
18 d
1
56:252
d
d 17:77 mm
Example 3 Calculate the resonant frequency for

TE111 mode using the given data of RDRA:
r 10
a 14 mm
b 8 mm
d 8 mm
Solution
r
c m 2 n2 p 2
fr p
2 r a b d
s
3 108 10002 10002 10002
fr p
2 10 14 8 8
fr 9:04 GHz
Example 4
r 10
a 14 mm
b 8 mm
d 16 mm
Solution
r
c m 2 n2 p 2
fr p
2 r a b d
s
3 108 10002 10002 10002
fr p
2 10 14 8 16
fr 7:44 GHz
Example 5 Calculate the resonant frequency for the TE11d mode using the given
data:
r 10
a 14 mm
b 8 mm
d 8 mm
Solution
s
c m 2 n2 d 2
fr p
2 r a b d
s
3 108 10002 10002
fr p 0
2 10 14 8
fr 6:82 GHz
Example 6
r 10
a 14 mm
b 8 mm
d 16 mm
Solution
r
c m 2 n2 p 2
fr p
2 r a b d
s
3 108 10002 10002
fr p 0
2 10 14 8
fr 6:82 GHz
4.4 Resonant Frequency of RDRA for Experimentations
The RDRAs can be prototyped with various materials and sizes as per the
requirements.
Table 4.1 consists of list of RDRA materials, permittivity, dimensions, and
computed resonant frequency.
Example 7 Compute resonant frequency when RDRA dimensions are 10 10
10 mm3 and dielectric constant of material used is 10.
r
mp2 np2 pp2
c
fr m; n; p p
2p l a b d
Resonant frequencies in isolated case are 49.7 and 25.8 GHz with ground plane
(Table 4.2).
Table 4.1 RDRA materials, permittivity, dimensions, and computed resonant frequency
S. Material Permittivity RDRA dimension Resonant Resonant
no. (a b h) mm frequency frequency
simulated by calculated
HFSS
Countis Laboratories
1. MgOSiO2TiO2 9.8 9 6 7.6 7.43 7.6757
(CD-9)
2. MgOSiO2TiO2 9.8 14.3 25.4 26.1 3.5 3.7430
(CD-9)
3. MgOCaOTiO2 20.0 10.16 10.2 7.11 4.71 4.6215
(CD-20)
4. MgOCaOTiO2 20.0 10.16 7.11 10.2 4.55 4.5941
(CD-20)
5. MgOCaOTiO2 20.0 10.2 10.2 7.89 4.635 4.4833
(CD-20)
6. MgOCaOTiO2 100.0 10 10 2 4.57 4.2158
(CD-100)
7. MgOCaOTiO2 100.0 10 10 1 7.97 7.7587
(CD-100)
8. MgOCaOTiO2 100.0 12.7 12.7 1 7.72 7.6628
(CD-100)
9. MgOCaOTiO2 100.0 5 10 1 8.85 8.1828
(CD-100)
10. MgOCaOTiO2 100.0 10 5 1 8.5 8.0147
(CD-100)
Emerson & Cuming Microwave Products N.V.
11. Magnesium 10.0 14 8 8 5.5 5.6117
titanate
(ECCOSTOCK@)
12. Magnesium 10.0 14.3 25.4 26.1 3.92 3.7055
titanate
(ECCOSTOCK@)
13. Zirconia 20.0 10.16 10.2 7.11 4.71 4.6215
(ECCOSTOCK@)
14. Zirconia 20.0 10.16 7.11 10.2 4.55 4.5941
(ECCOSTOCK@)
15. Zirconia 20.0 10.2 10.2 7.89 4.635 4.4833
(ECCOSTOCK@)
16. Strontium titanate 100.0 10 10 2 4.57 4.2158
(ECCOSTOCK@)
(ECCOSTOCK@)
18. Strontium titanate 100.0 12.7 12.7 1 7.72 7.6628
(ECCOSTOCK@)
(ECCOSTOCK@)
(continued)
4.4 Resonant Frequency of RDRA for Experimentations 89
Table 4.1 (continued)

S. Material Permittivity RDRA dimension Resonant Resonant
no. (a b h) mm frequency frequency
simulated by calculated
HFSS
(ECCOSTOCK@)
Morgan Advanced Materials
21. CaMgTi (Mg, Ca 20.0 10.16 10.2 7.11 4.71 4.6215
titanate)
(D20)
22. CaMgTi (Mg, Ca 20.0 10.16 7.11 10.2 4.55 4.5941
titanate)
(D20)
23. CaMgTi (Mg, Ca 20.0 10.2 10.2 7.89 4.635 4.4833
titanate)
(D20)
24. ZrTiSn (Zr, Sn 37.0 18 18 9 2.45 2.1617
titanate)
(D36)
Temex Components & Temex Telecom, USA
25. Zr Sn Ti Oxide 37.0 18 18 9 2.45 2.1617
(E2000)
Trans-Tech Skyworks Solutions, Inc.
26. BaZnCoNb 35.036.5 18 18 6 2.532 2.7081
(D-83)
27. BaZnCoNb 35.036.5 18 6 18 2.835 2.3947
(D-83)
T-CERAM, RF & Microwave
28. E-11 10.8 15.2 7 2.6 11.6 10.379
29. E-11 10.8 15 3 7.5 6.88 7.0937
30. E-11 10.8 15.24 3.1 7.62 6.21 6.9440
31. E-20 20.0 10.16 10.2 7.11 4.71 4.6215
32. E-20 20.0 10.16 7.11 10.2 4.55 4.5941
33. E-20 20.0 10.2 10.2 7.89 4.635 4.4833
34. E-37 37.0 18 18 9 2.45 2.1617
TCI Ceramics, Inc.
35. DR-36 36.0 18 18 6 2.532 2.7081
36. DR-36 36.0 18 6 18 2.835 2.3947
90
Table 4.2 Fringing effect along b dimensions increased effective along y-direction of RDRA
S. Permittivity Dimension (a (length) b (width) d (depth)) Resonant Effective width Multiple % change in
no. mm frequency (b) factor width
37. 10.0 14.3 25.4 26.1 3.5 34.22 1.3474 34.7381
38. 10.0 14 8 8 5.5 14.13 1.7665 76.6535
39. 10.0 15.24 3.1 7.62 6.21 8.33 2.8872 168.7230
40. 20.0 10.2 10.2 7.89 4.635 15.31 1.5014 50.1419
41. 20.0 10.16 10.2 7.11 4.71 15.15 1.4858 48.5797
42. 35.0 18 18 6 2.532 24.12 1.34 33.9973
43. 35.0 18 18 9 2.45 25.64 1.4244 42.4423
44. 100.0 10 10 1 7.97 11.24 1.1242 12.4237
4 Mathematical Analysis of Transcendental
MATLAB program and simulation effective length due to fringing effect:
%%Dimensions of RDRA
%%length
d=[14.3,14.0,15.24,10.2,10.16,18,18,10];
%%width
w=[25.4,8,3.1,10.2,10.2,18,18,10];
%%height
h=[26.1,8,7.62,7.89,7.11,6,9,1];
%%Mode
m=1;
n=1;
p=1;
c=3e8;
cons=[10.0,10.0,10,20,20,35,35,100];
syms y real
for i=drange(1:8)
kx(i)=pi/d(i);
kz(i)=pi/2/h(i);
ko=sqrt((kx(i).^2+y.^2+kz(i).^2)/cons(i));
f=real(y.*tan(y*w(i)/2)-sqrt((cons(i)-1)*ko.^2-y.^2));
ky(i)=fzero(inline(f),[0,(pi/w(i))-0.01]);
%%Resonant frequency
fre(i)=c/2/pi*sqrt((kx(i).^2+ky(i).^2+kz(i).^2)/cons(i))*1e3;
Effwidth(i)=pi/ky(i);
factor(i)=Effwidth(i)./w(i);
perchangwidth(i)=((Effwidth(i)-w(i))/w(i))*100;
end
Effective increased length computations due to fringing effect:
Program 1
%%Dimensions of DRA
%%length
d=[14.3,14.0,15.24,10.2,10.16,18,18,10];
%%width
w=[25.4,8,3.1,10.2,10.2,18,18,10];
%%height
h=[26.1,8,7.62,7.89,7.11,6,9,1];
%%Mode
m=1;
n=1;
p=1;
c=3e8;
cons=[10.0,10.0,10,20,20,35,35,100];
syms y real
for i=drange(1:8)
kx(i)=pi/d(i);
kz(i)=pi/2/h(i);
end
Results:
Program 2
m=1;
n=1;
p=1;
E_r=10;
a=15.24e-03;
b=3.1e-03;
d=7.62e-03;
c=3e+08;
kx=m*pi/a;
ky=n*pi/b;
kz=p*(pi/d)/2;
ko=sqrt(kx^2+ky^2+kz^2)/sqrt(E_r);
fo=(c*ko/pi)/2;
foghz=fo/(1e+09);
Results:
Program 3
MATLAB programs taking parameters a, b, d same and comparing
frequency using:
Program 1: Characteristic equation
m=1
n=1
p=1
E_r=10
a=14.3e-03
b=25.4e-03
d=26.1e-03
c=3e+08
k_x=m*pi/a
k_y=n*pi/b
k_z=p*(pi/d)/2
k_o=sqrt(k_x^2+k_y^2+k_z^2)/sqrt(E_r)
f_o=(c*k_o/pi)/2
f_oGHz=f_o/1e+09
Output:
Program 4
Transcendental equation for same dimensions:
m=1;
n=1;
p=1;
E_r=10;
a=14.3e-03;
b=25.4e-03;
d=26.1e-03;
c=3e+08;
syms y real
kx=pi/a;
kz=pi/d/2;
ko=sqrt(kx^2+y^2+kz^2)/sqrt(E_r);
f=real(y*tan(y*b/2)-sqrt((E_r-1)*ko^2-y^2));
ky=fzero(inline(f),[0,(pi/b)-0.01]);
fre=c/2/pi*sqrt((kx^2+ky^2+kz^2)/E_r)*1e3;
effwidth=pi/ky;
factor=effwidth/b;
perchangwidth=((effwidth-b)/b)*100;
Output:
Program 5
MATLAB programs taking parameters a, b, d same and comparing
frequency using:
Characteristic equation
Where a=17mm
b=25mm
c=10mm
m=1;
n=1;
p=1;
E_r=10;
a=17e-03;
b=25e-03;
d=10e-03;
c=3e+08;
k_x=m*pi/a;
k_y=n*pi/b;
k_z=p*(pi/d)/2;
k_o=sqrt(k_x^2+k_y^2+k_z^2)/sqrt(E_r);
f_o=(c*k_o/pi)/2;
f_oGHz=f_o/1e+09;
Output:
Program 6
Transcendental equation
m=1;
n=1;
p=1;
E_r=10;
a=17e-03;
b=25e-03;
d=10e-03;
c=3e+08;
syms y real
kx=pi/a;
kz=pi/d/2;
effwidth=pi/ky;
factor=effwidth/b;
Program 7
MATLAB programs taking parameters a,b,d same and comparing

frequency using:
Characteristic equation
m=1
n=1
p=1
E_r=10
a=14.3e-03
b=25.4e-03
d=26.1e-03
c=3e+08
k_x=m*pi/a
k_y=n*pi/b
k_z=p*(pi/d)/2
f_o=(c*k_o/pi)/2
f_oGHz=f_o/1e+09
Output:
Program 8
Transcendental equation
m=1;
n=1;
p=1;
E_r=10;
a=14.3e-03;
b=25.4e-03;
d=26.1e-03;
c=3e+08;
syms y real
kx=pi/a;
kz=pi/d/2;
effwidth=pi/ky;
factor=effwidth/b;
Output:
Program 9
Program 10
Q:No: 1 Develop transcendental equation for moat-shaped RDRA.

Q:No: 2 Compute propagation constants in x-, y-, and z-directed propagated
RDRAs, when feed probe is given. Compute its resonant frequency
when RDRA dimensions are 5 5 3 mm3 and dielectric constant used
is 20.
Chapter 5
Mathematical Analysis of RDRA
Amplitude Coefcients
Abstract Mathematical analysis of amplitude coefcients in rectangular DRA

(RDRA) have been evaluated. Rigorous theoretical analysis has been developed for
different resonant modes inside RDRA. The resonance phenomenon and its
potential use as radiator have been described. The dielectric polarization P is equal
to the total dipole moment induced in the volume of the material by the electric
eld. The discontinuity of the relative permittivity at the resonator surface allows a
standing electromagnetic wave to be supported in its interior at a particular resonant
frequency, thereby leading to maximum connement of energy within the reso-
nator. Certain eld distributions or modes will satisfy Maxwells equations and
boundary condition. Mathematical solution to get amplitude coefcients Cmnp along
with its phase coefcients has been obtained. These are also known as eigenvector.

Keywords Amplitude coefcients Resonant modes Radiation lobes Fourier

transform Discrete solution PMC (perfect magnetic conducting) PEC (perfect

electrical conducting) Dominant mode Higher-order modes
5.1 Introduction
Rigorous theoretical analysis has been developed for resonant modes in rectangular
DRA (RDRA). RDRA resonance phenomenon and its potential, as a radiator have
been long back described. Accordingly, external electric elds bring the charges of
the molecules of the dielectric into a certain ordered arrangement in space. The
dielectric polarization P is equal to the total dipole moment induced in the volume
of the material by the electric eld. The discontinuity of the relative permittivity at
the resonator surface allows a standing electromagnetic wave to be supported in its
interior at a particular resonant frequency, thereby leading to maximum conne-
ment of energy within the resonator. Certain eld distributions or modes will satisfy
Maxwells equations and boundary conditions. Resonant modes are eld structures
that can exist inside the RDRA. The RDRA prototype is shown in Fig. 5.1.

104 5 Mathematical Analysis of RDRA Amplitude Coefcients
Fig. 5.1 Homogenous

dielectric RDRA on ground
plane
5.2 Amplitude Coefcients Cmnp
Time domain elds can be written as follows: Ez x; y; z; t

P jxmnpt
mnp ReCmnp e umnp x; y; z, using orthonormality.
In discrete form,
X
jCmnp jumnp x; y; z cos x xmnp t Wmnp
m;n;p
The probe current can be expressed as:

Z
jxlIdlx2 y2 jxtxp
x2 y2 d2

Ez x; y; d; t Gx; y e c I xejxt dx
2 3=2
4p x y d
2 2
where G(x,y) are the constant terms associated with current.

r
X 2 ppd
Resonator current Cmnp sin cosxmnpt Wmnpum;n x; y;
p
d d
Z
jxlIdlx2 y2
Probe current Gx; y 3=2 I xe dx
jkt
4p x2 y2 d2
p
jxtxc x2 y2 d2 wmnp
e um;n x; ydx dy ;
5.2 Amplitude Coefcients Cmnp 105
The probe current must be equal to the resonator current due to principle of
orthonormality.
X
1 n o
E x; y; z; t Re Cmnpejxmnpt wEmnp x; y; z
mnp1
X
1 n o
E x; y; z
mnp1
X
1 n o
H x; y; z; t Re Cmnpejxmnpt wH
mnp
x; y; z
mnp1
X
1 n o
H x; y; z
mnp1
c jxl
E? 2
r? Ez 2 r? Hz x ^z
h h
From duality
c jx
H? r Hz 2 r? Ez x ^z
h2 ? h
From above two equations, we obtain Ex and Ey elds as given below:

mpx 0 mpx npy np mpx npy
Ex a mnp cos sin b0 mnp cos sin ;
a a b b a b
npy 0 mpx npy mp 0 mpx npy
Ey a mnp sin cos b mnp cos sin ;
b a b a a b
and
r
X 2 ppd
jxmnpt
Ez ReCmnpe sin um;n x; y
m;n;p
d d
Z
jxlIdlx2 y2 jxp
x2 y2 d2

Ez e c : I xejkt dx
2 3=2
4p x y d
2 2
Here, I x is the Fourier transform of source current, i.e., I(t) is the probe current
1X

Ix jI mnpj dx xmnpej mnp ej mnpdx xmnp
2 mnp
p !!
ldlx2 y2 x2 y2 d2
Ez x; y; z; t 3=2 xmnpjI mnpj sin xmnp t /mnp
4p x2 y2 d2 c
r
2 ppd

Cmnp jumn x; y cosxmnpt wmnp sin
d d
s
bmnp2 amnp2
Hence, Cmnp h p i2 ; amplitude coefcient
p2 sin
ppd
d d
" #
1
amnp cos/mnp bmnp sin/mnp
wmnp tan ; Phase
amnp sin/mnp bmnp cos/mnp
This completely solves the problem of RDRA resonant modes coefcients in

homogeneous medium.
5.3 RDRA Maxwells Equation-Based Solution
Maxwells equations with J electric and M magnetic sources:
r E jxlH M; a
r H J jxE; b
qv qm
rE ; rH ;
e l
where qv is the electric charge density, and qm is the magnetic charge density.
For consistency; jxer H r M 0;

r J jxer E 0;
i:e: r M jxqm 0; r J jxq 0;
namely conservation of electric and magnetic charge:
rr E jxlr H r M;
taking curl on both sides
or rr E r2 E jxl J jxE r M;
rq
or r2 k2 E jxlJ r H s electric source; c
e
5.3 RDRA Maxwells Equation-Based Solution 107
i.e., E satises the Helmholtz equation with source.
Likewise; r r H r J jxer E
or rr H r2 H r J jxejxlH M
rqm
or r2 k 2 H jxeM r J f
l
magnetic source due to probe; d
Hence, H also satises Helmholtz equation with source. Rectangular cavity reso-
nator sidewalls are the perfect magnetic conductors (PMC) and top and bottom
surfaces are the perfect electric conductors (PEC). Applying these boundary con-
ditions, we get the following equation:
Hz 0; where x 0; a or y 0; b
So,
X
Hz x; y; z /mn zumn x; yja; b 5:1
m;n 1
where
2 mpx npy
umn x; yja; b p sin sin 5:2
ab a b
as we know,
2 mpx npy ppz

Hz0 x; y; z Cmn p sin sin sin
ab a b d
Let
X
fz x; y; z fzmn zumn x; yja; b 5:3
mn 1
X
r2 k2 Hz /00mn z k 2 h2 m; nja; b/mn zumn x; yja; b
mn
5:4
fz ) /00mn z k2 h2 m; nja; b/mn z
fzmn z

m 2 n 2
where k2 x2 l; and h2 x; yja; b p2 a b ;
Hz 0; for z 0; d; completely determines mn z
from (1),
Taking Laplace Transform of (5.4);
mn s ^fzmn s
S2 /mn s S/mn 0 /0mn 0 c2z m; n b
So,
^fzmn s Smn 0 0mn

b
mn s 2 5:5
s2 cz m; n
2 S c2z m; n
where c2z m; n k2 h2 m; nja; b:

Thus,
Zz
1
b
mn z sin cz m; nz n fzmn ndn
cz m; n
0

C1 sin cz m; nz C2 coscz m; nz
mn 0 mn d 0 ) C2 0;
Zd
1
C1 sincz m; nd nfzmn ndn
cz m; n sincz m; nd
0
So,
Zz
1
mn z sin cz m; nd sin cz m; nz n fzmn ndn
cz m; n sin cz m; nd
0
Zd

sin cz m; nz sin cz m; nd n fzmn ndn
0
2
m 2 n 2 p 2
In the limit k 2 ! p2 a b d ; we have, c2z m; n ! pp
d and we get;
8 z 9
Z R d pp
d < pp ppz sin d n fzmn n dn =
mn z ! sin z n fzmn ndn sin limpp 0 d
pp : d d k! d sinkd ;
0
5:6
The limit in showing resonance, when

m 2 n 2 p 2
k2 p2
a b d
Let r
X
1
2 rpt
fmn z fzmnr sin sin 5:7
r1
d d
Then
Zd pp
sin d n fzmn ndn
d
0
5:8
r Zd rpn
2X pp
fzmnr sin d n sin dn
d r d d
0
r Zd
p1 2X rpn
1 fzmnr sinkn sin dn
d r d
0
r Zd
p1 2X 1 rp rp 5:9
1 fzmnr cos k n cos k n dn
d r 2 d d
0
r " #
1 2 X sin k rp d sin k rp d
1 p1
fzmnr d d
2 d r k drp
k d
rp
Here k propagation parameter kz pp

d
Thus,
Zd
1
sinkd n fzmn ndn
sinkd
0
r " #
1 2X 1 sinkd 1r sinkd 1r
1p1 fzmnr
2 d r sinkd k rp
d k rp
d
" # 5:10
1 X 1 1
p 1 p1
1 fzmnr
r

2d r k rp
d k rp
d
2 3
1p1 X 2rp
p 1r fzmnr 4 d
2 5
2d r k rp
2
d
Writing k pp
d d d ! 0; we get;
Zd
1
sinkd n fzmn ndn
sinkd
0
1p1 fzmnp
p 1p
2d d
5:11
1
p fzmnp Dominant term
d 2d
Hence,
8 z r 9
< Z X =
d 2 rpn ppz 1
mn z sinkz n fzmnr sin dn sin p fzmnp
pp : r1
d d d d 2d ;
0
5:12
Now
Zz Zz
rpn 1 h rp rp i
sinkz n sin dn cos kz k n cos kz k n dn
d 2 d d
0 0
" rpz #
1 sin rpz sinkz sin d sinkz

d
2 k rpd k rp
d
2 ppz rpz 3
d sin d dz
1 4 sin rpz sin d sin ppz

dz
5
ppr
d
2
d d ppr
d d
5:13
There is no dominant term here, i.e., if O1d, where O-order.

2
2 2
Hence, for k 2 p2 ma nb pp d d
ppz
d 1
mn z p fzmnp sin
pp d 2d d
r 5:14
1 d ppz
fzmnp sin
ppd 2 d
Likewise, propagation in x direction can be taken as:

r2 k 2 Hx fx
X
Hx x; y; z xmn ~umn y; zjb; d
m;n
X
Let, fx ~
x; y; z fxmn x~umn y; zjb; d
m;n
mpy npy
where ~umn y; zjb; d p2
bd
sin b cos b ; orthogonal 2D half wave Fourier basis
function.
Then,

00xmn x k2 h2 m; njb; d xmn x fxmn x
Hence, general solution can be given as follows:
Zx
1
xmn x sincx m; nx nfxmn n
cx m; n
0
C1 coscx m; nxC2 sincx m; nx 5:15
Likewise,
X
Hy x; y; z ymn yumn x; zja; d
m;n
X
fy x; y; z fymn yumn x; zja; d
m;n
with
00ymn y k2 h2 m; nja; d uymn y fymn y
and with the boundary conditions:
Ex 0 where x 0; a or z 0; d;
Ey 0 where y 0; b or z 0; d;
The general solution for ymn y is given as follows:
Zy
1
ymn y sin cy m; nz n fymn ndn
cy m; n 5:16
0
D1 coscy m; ny D2 sincy m; ny
Here, cx m; n k2 h2 m; njb; d1=2
cy m; n k 2 h2 m; nja; d1=2
The equation
r H J jxE
gives
jxEx Hz;y Hy;z J

jxEy Hx;z Hz;x J
We assume that J on the walls is zero. Then, the boundary conditions yields
Hz;y Hy;z 0; where x 0; a;

Hx;z Hz;x 0; where y 0; b
Recall that Hz has been completely determined.
5.4 RDRA Inhomogeneous Permittivity and Permeability

e e0 1 dp Xe x; y 5:17

l l0 1 dp Xm x; y 5:18
At some known frequency x and dp as perturbation parameter, the solution has

been worked out using perturbation techniques to determine shift in the frequency.
As per Maxwells equation,
r E jxlH
r H jxeE
where boundary conditions are given as follows:
0 x aW
0 y bL
0 z dh
5.4 RDRA Inhomogeneous Permittivity and Permeability 113
Due to duality E ! H; H ! E; and l $ e:

Sidewalls have been taken as PMC (magnetic conductor walls) and top and
bottom as PEC (perfect electrical conductor).
Htan 0; on side walls

Hx ; Hy 0; when y 0; b
Hy ; Hx 0; when x 0; d
Ex ; Ey 0; when z 0; d
E E x; yecz 5:19
H H x; yecz 5:20
Propagation constant is given as:

h20 c2 x2 l0 0 ; c2 k 2 ; when k2 x2 l0 0
c jxl
EX Ez0 x 2 Hz0 y 5:21
h2 h
c jxl
EY Ez0 y 2 Hz0 x 5:22
h2 h
c jx
HX Hz0 x 2 Ez0 y 5:23
h2 h
c jx
HY Hz0 y 2 Ez0 x 5:24
h2 h
Top and bottom walls are perfect electric conditions so that
Ex ; Ey 0; when z 0; d
E E x; yexpcz; H H x; yexpcz
E E ? Ez^z; H H ? Hz^z:
@
r r? ^z r? c^z
@z
r? Ez ^z c^z E? jxlH ? 5:25
E E ? Ez^z; H H ? Hz^z: 5:26
r? Hz ^z c^z H ? jxE? ; 5:27

r? H ? j w Ez^z 5:28
Taking ^z of (5.25) gives
r? Ez cE? jxl^z H? 5:29
Equations (5.11) and (5.13) can be changed as:

jx c E? r? Hz ^z
5:30
c jxl Z^ H ? r? Ez

jxl c r? Hz Z^

E? c jx r? Ez

^z H ? x2 l c2
c jxl
E? r Ez 2 r? Hz ^z 5:31
h2 ? h
c jxl
^z H ? r Hz ^z 2 r? Ez 5:32
h2 ? h
h2 h2 x; y c2 x2 lx; yex; y h20 k02 dvx; y;
where h20 c2 x2 l0 e0 ; k2 x2 l0 e0 c2 k 2
vx; y ve x; y vm x; y d ve x; yvm x; y
Taking ^z of (5.32) gives

c jxe
r? Hz 2 ^z r? Ez H ? 5:33
h2 h
from Eqs. (5.31) and (5.33),
c jxl
EX Ez;x 2 Hz;y
h2 h
c jxl
EY Ez;y 2 Hz;x
h2 h
c jx
HX Hz;x 2 Ez;y
h2 h
c jx
Hy Hz;y 2 Ez;x
h2 h
From Eqs. (5.25) and (5.26),

c jxl ^
r? r E
? z r H
? z Z jxlHz^z 0 5:34
h2 h2
c jxl

or ^z r? 2 ; r? Ez r? ; 2 r? Hz jxlHz 0
h h
or

jxl h2 h2 ^ c
r2? Hz h H z r?
2
; r ? Hz Z r? ; r ? E z
h2 jxl jxl h2
0
or
l
ck2
r2? h20 Hz d k 2 XHz d1 log 2 r? ; r? Hz r X r E
jxlh2 ?
1 ? z
h
0
5:35
Now, we retain only 0 d terms.
v ve vm
vm vm
2
h2 h0
ck 2 ck2

jxlh 2 jxlh20
and (5.35) becomes
2 r? v
r2? h20 Hz d k 2 v Hz r? vm ; r? H z
h20
5:36
ck 2
r ; vr? Ez g 0
jxl0 h20 ?
By duality
E ! H; H ! E; ve $ vm
0 $ l0 ; v$v
we get from (5.36)

k2 k2
r2? h20 Ez d k2 vEz r? ve r ? v; E r ? v r v; Ez
h20 ?
z e
h20
5:37
ck 2
r v; r? Hz 0
jx0 h20 ?
Boundary conditions are given as:
Hz 0; x 0; a and Y 0; b Hz 0; Z 0; d
HX 0; Y 0; b HY 0; x 0; a EX 0; x 0; a
EX EY 0; Z 0; d EY 0 y 0; b
Equations (5.28) and (5.29) are the own fundamental equations, let h20 k:
Let

k km;n
0
d k1 0 d2

Ez Ez0 dEz1 0 d2 5:38

Hz Hz0 dHz1 0 d2 5:39
if there is non-homogeneity !
2
m n2
kn;m
0
p 2

a2 b2

0
r2? kn;m Ez0 0

0
r2? kn;m Hz0 0 5:40

By Eqs. (5.36) and (5.37) 0 d0
c 0 jx0
Hx0 Hz;x 0 0
Ez;y
k0m;n km;n
Since
0 0
HZ 0; when Y 0; b Hz;x 0; when y 0; b
Then
0
Hx0 0; when y 0; b ! Ez;Y 0; when Y 0; b
Likewise
Ez;0X 0; when X 0; a
Thus,
mpx npy 2
Hz0 C mn sin sin p 5:41
a b ab
mpx npy 2
Ez0 Dmn cos cos p 5:42
a b ab
If z-dependent is taken into account, then Hz0 ; Ez0 must be multiplied by exp
cz
according to Eq. (5.34),
c jxl0
Ex0 Hz;0Y
k0m;n k0m;n
and Ex0 0; when z 0; d; and Hz0 0; when z 0; d.

We get Ez;0X 0; when z 0; d then,
2 mpx npy ppd

Ez0 x; y; z Dmn p cos cos sin 5:43
ab a b d
jpp
c ; p 1; 2; 3:
d
Since Hz0 0; when z 0; d;
2 mpx npy ppz

Hz0 x; y; z Cmn p sin sin sin 5:44
ab a b d
Frequency of oscillations:
x xmnp
0
c2 x2 l0 0 kmn
1=2
p m 2 n 2 p 2
or x p :
l0 0 a a d
5.5 RDRA with Probe Current Excitation
The rectangular cavity has dimensions a, b, and d as shown in Fig. 5.2. Sidewalls
are taken as magnetic conductors (PMC), and top and bottom surfaces are as PEC;
theoretical elds (modes) solution has been worked under boundary conditions with
a square-type feed probe for excitation.
EX ; Ey 0; top and bottom plane being electric walls.
EX ; Ey 0; sidewalls being magnetic walls.
X 2 mpx npy ppz
Hz x; y; z; t Cm; n; p p sin sin sin fcosxm; n; pt
mnp ab a b d
/m; n; p
where m, n, and p are the integers (half wave variations in particular direction, i.e.,
x, y, z directions, respectively); a, b, and d are the dimensions (width, length, and
height) of the RDRA, C m; n; p and /m; n; p are the magnitude and phase
coefcients of Hz and Dm; n; p and wm; n; p for Ez. npy
Let, orthogonal 2D half wave Fourier basis function p2 sin mpx
ab a sin b
umn x; y for convenience.
X 2 mpx npy ppz
Ez x; y; z; t d m; n; p p cos cos cos
mnp ab a b d
fcos xm; n; pt wm; n; pg
npy
Let, p2
ab
cos mpx
a cos b vmn x; y for convenience = orthogonal 2D half wave
Fourier basis function.

From Lorentz Gauge conditions, Ez jxAz @@z/
Therefore, the magnetic vector potential can be given as below in discrete form
after taking Fourier transform of Az.
Fig. 5.2 RDRA with square

feed probe inserted in
a b plane
d
a
b
5.5 RDRA with Probe Current Excitation 119
^ z x; y; z; x l Îx dlej c ; where dl probe length

xr
A 4p r
Div A ^ @ A^ Z ; need to be computed.
@z
Now, if we insert this probe at the location dened below into the cavity to nd
the elds pattern, we get:
l2 a a l2 b b
\jx j\ ; \jy j\
2 2 2 2 2 2
Then, the magnetic vector potential will be
^ z lÎdl
@A @ ejkr lîdl cos h jk cos h jkr jx ^
e 2/
@z 4p @z r 4p r2 r c
and scalar potential will be

^
lÎdljc2 1 jk jkr lîdlc2 z jkz jkr
/ cos h 2 e e
4 px r r 4 px r 3 r 2
^ w.r.t. z

Differentiating /
^

@/ lîdlc2 1 3z2 jk 2jkz2 z jkz jkz
5 2 4 3 2 ejkr 5:45
@z 4px r 3 r r r r r r
^
when; E ^ z @ /^ ; substituting
^ z jxA @/ ^z,
in E
@z @z

^ jxlÎdl lÎdlc2 1 3z2 jk 2jkz2 jkz2 k 2 z2
Ez 5 2 4 4 3 ejkr 5:46
4pr 4px r 3 r r r r r
If we take h p2 ; z 0
p
If we take h ; z 0
2

^
^ z xclIdl jk 1 3 d jk 2jkd jkd k d
2 2 2 2 2 2
E
4pk r r3 r5 r2 r4 r4 r3
Also,
q
r x 2 y 2 d2
k2 r2
r 1r , for r k 2p=k:
Coulomb component of electric eld is dominant in this inductive zone r 2 d2

given that d r:
Minimum of r l2 and Maximum of r a; b;
kr 1
Hence,
^ lc2 dl ^
^ z lcIdl
E and I x;
4pkr 3 4pr 3 jx
Zt
dl Qtdl
Ez t; x; yd p 3 Isds 5:47
4pr 3
4p x2 y2 0
Rt
Charge flowing through the resonator is Qt 0 Isds or equivalently
^
^ x;
I x Q
jx
Here,
X
Qtdl ppd
3 Dm; n; pvmn x; y sin cosxm; n; pt wm; n; p
4px2 y2 2 mnp
d
and
Qt X pp
3 C m; n; pvmn x; y cosxm; n; pt cm; n; p
4px2 y2 2
mnp
d
For complete solution, we need to compute Dm; n; p and wm; n; p coefcients

for Hz elds and Cm; n; p and cm; n; p for Ez elds. The Dm; n; p and
C m; n; p are the desired resonant modes. For region,
l2 a a
\jx j\
2 2 2
l2 b b
\jy j\
2 2 2
Z Z X Dm; n; ppp
Q t vmn x; y
3 dxdy cosxm; n; pt wm; n; p
4p x2 y2 2
p
d
5:48

a l2 a l2 b l2 b l2
\x\a; 0\x\ \ \y\b [ 0\y
2 2 2 2
5.5 RDRA with Probe Current Excitation 121
1
hcosxt cosxti
2
hsinxt cosxti 0
Z
Dm; n; ppp 1 vmn x; y
coswm; n; p hQt cosxm; n; pti 3 dxdy
2d 4p x2 y2 2
Z
Dm; n; ppp 1 vmn x; ydxdy
sinwm; n; p hQt sinxm; n; pti 3
2d 4p x2 y2 2
Hence,
Z !
1 vmn x; ydxdy
Dm; n; p 2d=pp sinwm; n; p hQt sinxm; n; pti 3 :
4p x2 y2 2
5:49
5.6 RDRA Resonant Modes Coefcients in Homogeneous

Medium
The basic Maxwells theory can be applied with boundary conditions to express
RDRA resonant elds as superposition of these characteristics frequencies. RDRA is
shown in Fig. 5.3. umn depends on input excitation = orthogonal Fourier basis
function, hmn resonant mode (cut off frequency), k propagation constant. The gen-
eration of modes or characteristics frequencies xmnp due to electromagnetic elds
oscillations inside the cavity resonator has been described. Orthogonal Fourier basis
npy
function um;n x; y p2
ab
sin mpx
a sin b ; xmnp is the characteristic frequency
and wmnp is the phase of current applied. The rectangular cavity resonator is
excited at the center with an antenna probe carrying current i(t) of some known
Fig. 5.3 RDRA with ground

plane
frequency xmnp. This generates the eld Ez inside the cavity of the form given
below:
k2 c2mn h2mn
hence,
p 2 p2
k2 h2mn
d2
X Z
Ez x; y; z; t Re Cmnp ejxmnpt umnp x; y; z;
m;n;p
P
or Cmnp umnp x; y; z cosxmnpt wmnp;
m;n;p
Z
jxlIdlx2 y2 jxtxp
x2 y2 d2

Ez x; y; d; t Gx; y 3=2
e c I xejxt dx
4p x2 y2 d2
where G(x,y) are the constant terms associated with the current.
Equating RDRA probe current elds with the antenna-radiated current elds at
z d;
Radiated currents:
r
X 2 ppd
Cmnp sin cos xmnpt /mnp um;n x; y ;
p
d d
Due to orthonormality, probe currents will be equal to radiated elds.

Probe currents:
Z p
jxlIdlx2 y2 jxtxc x2 y2 d2 wmnp
Gx; y 3=2 I xe dx e
jkt
um;n x; y dxdy
4p x2 y2 d2
It is clear that these two expressions have to be equal due to energy conservation.
The probe current can be dened as:
1X h i
Ix jI mnpj dx xmnpejmnp ejmnp dx xmnp
2 mnp
The antenna probe current must contain only the resonator characteristics fre-
quencies xmnp. The radiated and input currents are equated as:
5.6 RDRA Resonant Modes Coefcients in Homogeneous Medium 123
r
X 2 ppd
Cmnp sin cosxmnpt /mnpum;n x; y
p
d d
Z p
jxlIdlx2 y2 jxtxc x2 y2 d2 wmnp
Gx; y 3=2
:I xe jkt
dxe um;n x; ydxdy;
4p x2 y2 d2
probe current = radiated current; thus Cmnp can be completely determined.

Hence, we can conclude that modes generation is due to the dipole moment in
cavity resonator, mostly depend on size, dimensions of device, excitation type,
coupling, and point of excitation.
5.7 RDRA Modes with Different Feed Position
Let us take z d, i.e., very small

probe length inserted into RDRA resonator at
point of insertion a=2; b=2; d or x a=2; y b=2; d ; where dlength of
insertion.
Hx ; Hy ; Ex ; Ey , transverse elds; (Ez and Hz) longitudinal elds
X
Ez umnp x; y; zRe Cmnp ejxmnp t
mnp
npy ppz
where umnp x; y; z p2 sin mpx
3=2
abd a sin b sin d Ez ; (when top and bottom

walls are PMC, rest all four walls are PEC).
Appling boundary conditions on transparent sidewalls (on all four sides of
RDRA or resonator) and top and bottom planes as electrical walls, we get Hz 0,
for magnetic walls; and Ez 0, for electrical walls; elds to be computed are
Ez ; Hz longitudinal elds;
X Z
Ez x; y; z; t Re Cmnp ejxmnpt umnp x; y; z
mnp
At z 0; Ez ; Ex ; Ey all will be zero
P
q
Ez mnp Re Cmnp ejxmnpt d2 sin ppd
d umn x; y; this is the Ez eld in the reso-
nator at z d. It must be equated to the corresponding eld generated by the
antenna probe, i.e., for the above two expressions to be equal, the antenna probe
currents must contain frequencies
onlynpfrom the set fxmnpg:
Where umn x; y p2
sin mp
ab a sin b
p
R jk 2 2 2
x y d
Ez will exist little above from z 0 plane; Ez p
jxlIdle
I xdx; where
4p x2 y2 d2
I x is the Fourier transform of i(t)
@/
Ez jxAz
@z
lIdl
divA jk cos hejkr 5:50
4p
kc2 jx
2/
x c
Hence, scalar potential / lI4pr

cos jkr
e
@/ lI cos h
jk cos hejkr
@z 4pr 5:51
jklI cos 2h jkr
e
4pr
jxlIdl jkr jxclIdl cos 2h jkr
Ez e e
4pr 4pr
Hence,
jxlIdl sin2 h jkr

Ez will be e
4pr
where
d
cos h p2
x y2 d
2
x2 y2
sin2 h
x 2 y 2 d2
r
X 2 ppd
jxmnpt
Ez jz d ReCmnpe sin um;n x; y 5:52
mnp
d d
Z
jxlIdlx2 y2 jxp
x2 y2 d2

Ez 3=2
e c I xejkt dx 5:53
4p x2 y2 d2
5.7 RDRA Modes with Different Feed Position 125
Here, Ix is the Fourier transform of source current, i.e., It probe current
1X
I x jI mnpjdx xmnpej mnp
2 mnq

ej mnpdx xmnp
1X
Ix jI mnpjdx xmnpej mnp
2 mnq

ej mnpdx xmnp
Z
Ix cosxmnptejxt dt
q
2
When xmnp p2 ma2 nb2 dp2 , probe current magnitude and phase Ix
2 2
P
m;np jI mnpj cosxmnpt /mnp mnp is the phase of current at fre-
quency xmnp.
Ez x; y; z; t
p !!
ldlx2 y2 x2 y2 d 2
3=2 xmnpjI mnpj sin xmnp t /mnp
4p x2 y2 d2 c
r
2 ppd
jCmn jumn x; y cos xmnpt wmnp sin :
d d
5:54
5.8 R, L, C Circuits and Resonant Modes
The information contained in eigenvalue or eigenvector of modes can impart the

knowledge of antenna radiation behavior, surface current distribution, input
impedance, and its feeding point location. Combinations of feeding conguration
and dimensions can generate or excite various modes. Thus, modes can be effec-
tively used in design control of an antenna. Surface current and geometry of an
antenna give eigenfunctions or eigenvectors. Closed-loop currents of eigenvectors
that present inductive nature are the magnetic elds. Horizontal and vertical
eigenvectors are noninductive are electric elds. These electric elds are produced
by supplied probe currents. Number of lobes in radiation pattern gets increased if
mode number or order of mode is increased and vice versa. The modal excitation
coefcients shall depend on position, magnitude, and phase of the applied probe
current. The effective current is superposition of all modes excited. The eigenvalue
is most important because its magnitude tells effectiveness of radiation or reactive
power and modes are the solution of characteristics equation. Smaller magnitude of
Fig. 5.4 a RLC circuit, b resonance higher modes, c magnetic dipoles
Fig. 5.5 Higher-order even and odd modes
eigenvalue is more efcient. Positive eigenvalue is the magnetic energy storing

mode, and if modes are negative, it stores electric energy. The eigenvalue variation
versus frequency gives information about resonance and radiation nature.
Excitation angle can have impact on antenna quality factor. The excited mode will
adjust the phase of the reflected currents. Orthogonality of modes can be used to
produce circular polarization in the RDRA. Figure 5.4 represents the equivalent
RLC circuit of RDRA, resonant modes excited, and corresponding magnetic
dipoles. Figure 5.5 depicts the even and odd modes generation. Figure 5.6 presents
RDRA HFF model along with its equivalent RLC circuit. Figures 5.7 and 5.8 are
RLC circuits which are used for derivation of resonant frequency and impedance.
Fig. 5.6 a RDRA model and b equivalent RLC circuit

5.8 R, L, C Circuits and Resonant Modes 127
Fig. 5.7 RLC circuit
Fig. 5.8 Series RLC circuit
R, L, C equivalent circuit: An antenna can be represented as R, L, C circuitry

with
natural frequencies xc and forced resonance due to excitation eigen-valued xmnp
has been determined along with eigenvector Jmnp . Separation of all frequencies will
be the out come of modes. The second-order differential equation is the general
solution of equivalent antenna (R, L, C) circuit. Fourier solution will provide a
discrete solution of resonance. Je is excitation current or probe current and c is an
propagation constant c a jb. L, C circuit will introduce non-homogeneous or
inhomogeneous matter, x2 will be replaced in this case by x2 l c is replaced by
~c introducing decay. Hzf represents forced resonance mode.
:: q
Lq Rq_ vs t
C
:: 2
where q ddtq2
1
XL jxL; XC
jxC
Taking Fourier Transform

2 1
jxL jxR Qx Vs x
C
Vs x
Q x 2
jxL 1c jxR
Z1
Qxejxt dx
q t
2p
1
R0
1
xL2
xc
1
x p
LC
J se x; y; z Jsx x; ydz d0 ^x Jsy x; ydz d0 ^y
where Js is the current surface density, and Je is the electron current

Z
Js dz Jsx x; y; x^x Jsy x; y; x
From Maxwells equation,
r H J e r jxE
r E jxlH
r2 E jxlr jxE J e
r2 Ez c2 xEz
When
p
c x jxlr jx ax jbx
Similarly, we can compute
r2 Hz c2 xHz
Boundary conditions are applied
Hz 0; x 0; a; or y 0; b; z 0; d
Ex 0; x 0; a; when z 0; d; Ey 0; y 0; b; z 0; d;
5.8 R, L, C Circuits and Resonant Modes 129
Fields propagating is Hz for TE mode

p npx mpy ppd
X 2 2
Hz x; y; z; x p sin sin sin Re C mnp expjxnmpt
abd a b d

r2 c2 x Ez 0
c and ~c are two propagation constants
r2 c2 xEz 0
r2 c2 xE ? J e
r2 H r J e r jxjxljH
2
r c2 x H r J e
r2 c2 xE ? J e
2
r ~c2 Hx Jsy d0 z d0
2
r ~c2 Hy Jsx d0 z d0
2
n m2 p2
~c2 x p2 2 2 2 0
a b d
2
2 n m2 p2
jxlr jx p 0
a2 b2 d 2
2
2 n m2 p2
~c xmnp x l jxlr p
2 2

a2 b2 d 2
xmnp xReal mnp jxImg mnp
ejxmnpt ejxReal mnpt exImg mnpt
x2 l ! x2 l jxlr ~c2 x
X
Jsxy x; y x Jsyx x; y xdz d0 Jz n; m; p; xunmp x; y; z
Za Zb p npx mpy d pp

2 2 0
J n; m; p; x Jsxy x; y; x Jsyx x; y; x p sin sin sin dxdy
abd a b d
0 0
5:55
X
r2 ~c2 x Hz Jz mnp; xumnp x; y; z
mnp
X
Hz f Hz mnp; xumnp x; y; z;
mnp
When Hz f is the forced resonant mode, then

n2 m2 p2
Jz nmp; x p2 ~c2 xHz nmp; x
a2 b2 d 2
X Jz nmp; xumnp x; y
Hz f x; y; z x
mnp
~c2 x ~c2 xnmp
where
p npx mpy ppz
2 2
vmnp p cos cos cos Hz
abd a b d
X
Jsy x; y; xd0 z d0 Jy mnp; xvmnp x; y; z
Hence, current density
Z a Z b Zd
Jy n; m; p; x JSy x; y; xd0 z d0 vmnp x; y; zdxdydz
0 0 0
This completes the general solution of R, L, C circuit.
5.9 Resonant Modes Based on R, L, C Circuits
r? Hz ^zc ^z H ? Je r jxE? 5:56
r? Ez ^zc ^z E ? jxlH ? 5:57
^z r? Ez ^zc ^z E? ^z jxl H ?
r? Ez cE? jxl ^z H ? 5:58
Eliminate ^z H ? from Eqs. (5.56) and (5.58)
jxl
r? Ez cE? r? Hz ^z Je r jxeE? :
c
2
c ~c2 x E? jxl r? Hz ^z jxl Je c r? Ez
5.9 Resonant Modes Based on R, L, C Circuits 131
Hence,
jxl jxl c r Ez
E? r Hz ^z 2 Je 2 ?2
c2 ~c2 x ? c ~c2 x c ~c x
Parallel RLC Circuits solution:

Z
di 1
Ii L i dt V
dt c
On differentiating
di2 di 1
L R 0
d2 t dt c
Second-order linear, homogeneous differential equation dividing by L both sides

gives the following:
di2 R di 1
2
0
d t L dt LC
Taking Laplace transform
R 1
S2 S 0
L LC
q
R2
R L 4 1
LC
S
L 2
r
R2 2 2
p
R L LC
S1
L 2
r
R2 2 2
p
R L LC
S2
L 2
Series RLC circuit

Let
R
k
s2L

2 2
R 1
x2
2L Lc
S1 k x 2
S2 k x 1
Hence, solution of differential equation can be written as:
I A1 es1 t Az es2 t
Here, A1 and A2 are the magnitude of currents

Now
2
R 1
Care 1 [
2L Lc
2
R 1
Care 2 \
2L Lc

R 2 1
Care 3
2L Lc
Z
Ldi 1
V Ri i dt
dt C
Taking Laplace transformation
V I s
RI s LsI s
s sC
V
I s

s R Ls sC 1
V
I s
Rs Ls2 C1
5.9 Resonant Modes Based on R, L, C Circuits 133
Since,
q

R 2
R=L
L 4=LC
s1;2
2
V V 1 1
I s
2 Rs 1
L s L LC L s 1 s 2 s s 1 s s 2
Taking Laplace inverse of equation
1 1
It q es1 t q es2 t
L L2 LC
R 4
L L2 LC
R 4
Example 5.1 Series RLC circuit solution
Zt
dit 1
v R it L it dt
dt C
0
Taking Laplace transform on both the sides gives
v 1
Is R Ls I s i0 I s i0 0
s Cs

v 1
Is R L s
s Cs

1
v Is R s L s2
C
Roots of the equation are as follows:
L s2 R s C1 0
s
2
R R 1
S

2L 2L LC
s
2
R R 1
Let s1 s
2L 2L LC
s
2
R R 1
and s2
2L 2L LC
R 1
Now; s1 s2 and s1 s2
L LC
v Is Ls2 s s1 s2 s1 s2
v Is Lss s1 s2 s s1
v Is Ls s1 s s2
v 1
Is
L s s1 s s2
Using partial fraction solution, we get

v 1 1
Is
Ls1 s2 s s1 s s2
Taking inverse Laplace transform on both the sides

v
it es1 t es2 t
Ls1 s2
r
R 4L
s1 s2 1
L C
v
it q es1 t es2 t
R 1 4L C
v
Let; A1 A2 q
R 1 4L C
it A1 es1 t A2 es2 t
Chapter 6
Mathematical Analysis of Radiation
Pattern of RDRA
Abstract In this chapter, detailed study using mathematical analysis for radiation
pattern of RDRA has been described. RF excitation with proper impedance match
can generate J-current density into surfaces of RDRA, which leads to produce
A-magnetic vector potential and nally E-electric intensity or H-magnetic eld
intensity. Acceleration or deceleration of charge carriers causing current is
mandatory phenomenon for radiations. Wave can only propagate if wave vector
k > kc, where kc is cutoff frequency. The lowest resonance can be termed as
dominant mode and second and third resonances are higher-order modes.
Propagation constant kx np=a, and propagation takes place if kx [ np=a, while
no propagation takes place if kx \np=a. Thus, standing waves inside the resonator
are formed and energy storing will take place. Hence, mode spectrum will result
into corresponding resonant frequency generation. Wave propagation can be well
dened by Helmholtz equation. The Maxwells equations describe the behavior of
electromagnetic elds and form the basis of all EM classical phenomena. Prad
(power radiated) can be evaluated using Parsevals power theorem. The radiated
power is produced by oscillating dipole moments. The current varying in time can
be analyzed by Fourier analysis. If medium is inhomogeneous, wave possesses
exponential growth or decay in some direction. Thus, Poynting vector S shall
give the magnitude and phase of the radiated elds in particular direction.

Keywords Impedance match Current density Magnetic vector potential

Power radiated Poynting vector Persvals power theorem Moat-shaped DRA
6.1 Introduction
RF excitation with proper impedance match can generate J-current density into
surfaces of RDRA, which leads to produce A-magnetic vector potential and nally
E-electric intensity. Acceleration or deceleration of charge carriers causing current
is mandatory phenomenon for radiations. Wave can only propagate if wave vector

136 6 Mathematical Analysis of Radiation Pattern of RDRA
k > kc, where kc is cutoff frequency and the lowest resonance can be termed as
dominant mode and second and third resonances are higher-order modes.
Propagation constant kx np=a. Propagation takes place if kx [ np=a, while no
propagation takes place kx \np=a. Standing waves inside the resonator are formed
and energy storing will take place. Hence, mode spectrum will result into corre-
sponding resonant frequency generation due to equivalent RLC circuit formation.
Wave propagation can be well dened by Helmholtz equation. The Maxwells
equations describe the behavior of electromagnetic elds and form the basis of all
EM classical phenomenon. Prad (power radiated) can be evaluated using Parsevals
power theorem. The radiated power is produced by oscillating dipole moments. The
current varying in time can be analyzed by Fourier analysis. If medium is inho-
mogeneous, wave possesses exponential growth or decay in some direction. Thus,
Poynting vector S shall give the magnitude and phase of the radiated elds in
particular direction.
Finally, the radiation pattern produced by the surface electric and magnetic
current densities on the RDRA surfaces is computed. PEC walls, the surface electric
current density is Js ^n E.
Then, the far-eld magnetic vector and electric vector potentials are determined
by the usual reactance potential formulae as follows:
Z
l ejkr
Ax; r J s x; r 0 expjk^r r 0 dsr 0 ; 6:1a
4p r
s
and
Z
ejkr
F x; r M s x; r 0 expjk^r r 0 dsr 0 : 6:1b
4p r
s
Lorentz force conditions are applied to determine the far-eld electric scalar and
magnetic scalar potentials as follows:
j
/e x; r div Ax; r
xl
6:2a
k
^r ; Ax; r
xl
j
/m x; r div F x; r
xl
6:2b
k
^r ; F x; r
xl
The far-eld electric and magnetic elds (i.e., up to Order r 1 ) are then
determined as follows:
6.1 Introduction 137
1
E r/e jxA r F; 6:3

1
H r A r/m jxF;
l
jk 2 jk
^r ^r ; A jxA/ ^r F; 6:4
xl
jk
jxA? /m ^r F;

where
A? Ah ^h A/ /;
^
jk
H ? ^r A jxF 1 ;
l
Finally, we derive expression for the Poynting vector as follows:
1
S RefE H g:
2

Up to order r12 i.e., value 1/r2 is taken into account from where, the RDRA
radiation resistance is evaluated:
Z
1 2
I Rr lim S ^r r 2 dX;
2 r!1
when I is the input current to the RDRA, Rr or Rr x is radiation resistance and

depends on the frequency.
6.2 Radiation Pattern of RDRA Due to Probe Current

i(t) and Probe Length dl
!
lI dl ejkr !
A; where A is magnetic vector potential 6:5
4pr

From Helmholtz equation ~
A
E jx ~
A
Radiated power can be given as follows:

2 r
jE j2 x2 ~
A l
; g characteristic impedance:
2g 2g
Z
J r 0 ; xejkjrr j 3 0
0
~ l
A d r ; at source: 6:6
4p jr r 0 j
Volume
We know that radiation pattern can be dened by the electrical eld intensity
Eh ; E/ :
Eh jxAh and Ah ^h A
Antenna surface current density can be expressed as follows:

X
J r 0 ; x J s mnp; r 0 ejxmnpt ; where; r x; y; z 6:7
mnp
The magnetic vector potential in terms of J can be written as follows:

jrr0 j
Z
l X J s mnp; r 0 ejxmnp t c
A dsr 0 ; where; ds is surface of RDRA
4p mnp jr r 0 j
Z
l ejkn X 0
J s mnp; r 0 ejxmnp^rr dsr 0
4p jr r 0 j mnp
s
6:8

H/ Eh =g; Hh E/ g:
Hence radiated power can be given as:
1 2
Prad j E h j 2 E /
2g

^h ^x cos u cos h ^y sin u cos h ^z sin h
^ ^x sin u ^y cos u
/
l XZ n o
Eh 2
Re Jsx mnp; r0 cos u cos h Jsymnp;r0 sin / cos h Jszmnp;r0 sin h
4pr mnp
s
mnp 0
expjx gx cos / sin h y0 sin / sin h z0 cos h dsr0 ejxmnpt ;
c
6:9a
6.2 Radiation Pattern of RDRA Due to Probe 139
XZ n o xmnp
Eu Re Jsx mnp; r 0 sin / Jsymnp;r0 cos / ej c
mnp
s 6:9b
0 0 0 0 jxmnpt 0
x cos u sin h y sin / sin h z cos h dsr e dsr :
Radiated power Prad, x, y, z component wise, can thus be dened as follows:

Z
xmnp^r r0
Px ^r jmnp Jsx mnp; r 0 ej c dsr 0 6:10a
s
Z
xmnp^r r0
Py ^r jmnp Jsy mnp; r 0 ej c dsr 0 6:10b
s
Z
xmnprr 0
Pz ^r jmnp Jsz mnp; r 0 ej c dsr 0 6:10c
s
^r h; / ^x cos / sin h ^y sin / sin h ^z cos h:
Let s = mnp for convenience then

X

Eh Re Px ^r js cos / cos h Py ^r js sin / cos h Pz ^r js sin h ejxst
s
X 6:11
Re Esh ejxst
s
3 2
000
where s mnp 4 001 5 and so on till s 111, similarly
010
X
E/ Re Px ^r js sin / Py ^r jscos/ejxst 6:12
s
6.2.1 Radiation Pattern
Now, power radiation pattern can be dened as follows:

2 ( )
jEh j2 E/ 1 X
Esh ejxst Es/ ejxst
2g 2 s
( ) 6:13
1 X jxst
X
jxst
Hsh e Hs/ e
2 s s
!
1 X X
Es Hm ejxs xm t Es Hm ejxm xs t
4 s s
1X

Esh Hs Es Hsh
4 s
1 X

Re Esh Hs 6:14
2 s
1 ^ E Es ^

^ sh ^
Esh h E / h
2 g g
X jEsh j2 jEs j2
^r ^r ;
s
2g 2g
6.3 Poynting Vector
Poynting vector is dened as radiated power flux per unit solid angle or power
radiated in particular direction in specied angular zone.
H rA
E r ddAt ; scalar and magnetic vector potential from Lorentz gauge
conditions.
S E H ; S is Poynting vector (energy flow or flux).
Prad
Z Input impedance
jIj2
1 Xh
S ^r xs2 jPx ^r js cos / cos h Py ^r js sin / cos h Pz ^r js sin h2
2g mnp
i
xs2 jPx ^r js sin / Py ^r js cos /2 6:15
1 X
S ^r r; h; / xmnp2 fjPx h; /jmnp cos / cos h
2g mnp
6:16
Py h; /jmnp sin / cos h Pz h; /jmnp sin h2

jPx h; /jmnp sin / Py h; /jmnp cos /2
6.4 Moat-Shaped RDRA Radiation Pattern 141
6.4 Moat-Shaped RDRA Radiation Pattern
Moat-shaped RDRA is shown in Fig. 6.1a with x, y, and z coordinates, and feed is
given at a/2 position.
In Fig. 6.1b, rectangular moat-shaped RDRA is covered with r copper plate to
reduce resonant frequency.
E(t, x, y, z) is electric eld intensity of RDRA to be computed in time domain
and E(, x, y, z) in frequency domain having a, b, and d dimensions, excited with
feed probe at a2 ; a2 ; 0 point by I0 cos xt RF current.
A Az^z (due to RF excitation current I0 cos xt along length d inserted into the
RDRA).
Hence, magnetic vector potential can be written as follows:
Zd
lI0 ejkjra=2^xb=2^yn^zj
Az x; x; y; z dn; 6:17
4p jr a=2^x b=2^y n^zj
0
Fig. 6.1 a Moat-shaped RDRA. b RDRA moat cover with rectangular copper plate to reduce
resonant frequency
Let C 4p
l
; k x=c and n = variable probe length.
9
Zd expfjk x a= 2 y a= 2 z n2 > =
2 2
Az CI0 1=2 >dn 6:18
x a=22 y a=22 z n2 ;
0
Far-eld approximation can be determined as follows:
CI0 ejkr
Az Ph0 ; 0 ; 6:19
r
where Ph0 ; 0 is radiation pattern.

Here, it is assumed that probe is very small as compared to RDRA.
x a=22 y a=22 z n2
2
x a=22 y a=2 z2 2zn
2
x a=22 y a=2 z2 d 2
1=2
r x a=22 y a=22 z2
where
r = distance from the points (x, y, z) in the center of the feed probe
a= ; a= ; 0
2 2
1=2
x a=22 y a=22 z2

r2 2zn1=2 r1 zn r0 2 r zn=r0 :
Hence, magnetic vector potential due to source inside RDRA can be computed
as follows:
Zd
CI0 ejkr0 jkzn
Az exp dn; where I0 probe RF current:
r0 r0
0

jkr0 exp jkzn=r

CI0 e 0 n d
n 0 i.e., variable probe length:
r0 jkz=
r0
6:20
jkzd= 1
CI0 jkr0 exp r0
e
r0 jkz=
r0

exp jkzd= 2j sin kzd=
CI0 jkr0 2r0 2r0
e
r0 jkz=
r0
6.4 Moat-Shaped RDRA Radiation Pattern 143

n o sin kzd=2r
Az 2CI0 exp jk r0 zd=2r0
0
kz
where, z r cos h:
jkd cos h0

cos h0 sin kd
Az x; x; y; z CI 0 expjkr0 exp 2 2
6:21
kr0 cos h0

here, (r, , ) are spherical polar coordinates of (x, y, z) so as to relate a=2; a=2; 0 ,
the probe insertion point. Hence, magnetic vector potential can be expressed as
follows:
x 2 2 a
Et; x; y; z 3
jPh0 jsinxt kr0 Wh0 xa=2 y a=2 x z^x
r0 2

x y =2 za
jPh0 jsinxt kr0 Wh0 ^y
r0 3
6:22

k x a2
Bt; x; y; z jPh0 jsinxt kr0 Wh0 ^y
r2
0 a
k y2 6:23
jPh0 jsinxt kr0 Wh0 ^x
r0 2
Ph0 a a
k sin xt kr 0 W h 0 y ^
x x ^y
r0 2 2 2
Finally, we derive expression for the Poynting vector as follows:
1
S RefE H g
2
1
Up to O r2 from where the radiator resistance is evaluated as
Z
1 2
I Rr lim S ^r r 2 dX
2 r!1
where I is the input current to the RDRA. Rr or Rr x depends on the frequency.

Hence, this completes the solution for radiation pattern of RDRA.
6.5 Quality Factor of RDRA
The quality factor Q of the RDRA can be evaluated by comparing the power
radiated Prad 12 I 2 Rr with the average electromagnetic energy (W) stored with the
RDRA as follows:
Z
1
W x E; E lH; H dx dy dz 6:24
4
0;a0;b0;c
The average energy stored per unit cycle with the RDRA is
W x x
Px W x 6:25
2p= 2p
x
The quality eld factor of the RDRA is thus
2xW x
Qx ;
jI xj2 Rr x
where x corresponds to resonant frequency.

The quality factor of a resonant mode measures how sharp its resonance is. As
per conservation of energy,
Z Z
jEj2 dv jH j2 dv
(time) average magnetic energy will be equal to electric energy inside the resonator.
The time-averaged energy dissipated in the walls of RDRA in unit time can be
calculated as of energy into walls from the electromagnetic elds in the cavity
normal component of energy based on the boundary conditions as energy flux
density as follows:

C
S ReE H 6:27
8p
Hence, total energy dissipated is given by

I
c
RejH j2 df
8p
Change in resonant frequency due to dielectric material used in RDRA:

p
The resonant frequency is reduced by l
p
If x ! x l
6.5 Quality Factor of RDRA 145
xa; xb are orthogonal frequencies, and Ea and Eb are orthogonal elds.

x0 0 00
2jx00 j = quality factor (Q), x is real frequency, and x is imaginary frequency.
0 00
Complex freq x x jx
Z Z
Ea Eb dv Ha Hb dv 0
Resonator lled with non-absorbing dielectric, for which and l differ from
p
unity by replacing x by x l and E by E, and H by lH.
The (time) average energy flux through surface is
c
S Re Et Ht 6:28
8p
c
where S E H .
4p
If Q of heat evolved per unit time and volumes
x 00 2
Q E l00 H 2 6:29
4p
Bar denotes time-average exciting frequency, must be exactly equal to the chosen
resonance frequency, and is required to establish eld conguration inside res-
onator. This results in dissipation of energy in the cavity walls and dielectric lling
of the cavity resonator. A measure of the sharpen of response of the cavity to
external excitation is quality of the cavity. This is dened as 2p times the ratio of
the time-averaged energy stored in the cavity to the energy dissipated.
stored energyfWxg
Q x0 6:30
power lossI I Rr
Fig. 6.2 Rectangular RDRA

moat
where x0 Resonant frequency oscillations of elds are damped and time depen-
dent. Change in frequency Dx to occur based on superposition of frequencies:
x x0 Dx
1 1 Z
E t p E xejxt dx
2p 1
Q. No. 1 Compute resonant frequency and propagation constant of given RDRA

shown in Fig. 6.2 and also compute quality factor of a RDRA having dimensions
10 10 10 mm3 with dielectric constant 10 and probe current 10 mA.
Chapter 7
Rectangular DRA Higher-Order Modes
and Experimentations
Abstract In this chapter, rectangular DRA higher-order modes have been realized
by mathematical modeling. Resonant modes are seen with experimentations in
anechoic chamber. These resonant modes impart physical insight into the radiating
phenomenon of the antenna. Knowledge of modes can be boon to the antenna
designer. If antenna resonant modes are known, radiation parameters can be steered.
There are two types of modes and they are dominant and higher-order modes. The
dominant mode corresponds to the lowest resonant frequency. These higher-order
modes can be generated either by increasing electrical length of RDRA or by
applying higher excitation frequency. The resonant frequencies of the modes are
represented by eigenvalues and currents by eigenvectors. Radiating behavior of the
antenna can be predicted by modes. They can also help to determine input exci-
tation point. Moreover, having in mind the current distribution of the modes, the
geometry of the antenna can be modied. The aspect ratio is the important
parameter in RDRA. Devising control on aspect ratio can alter resonant frequency,
gain, and bandwidth. RDRA has two fold design flexibility because of two aspect
ratios. The resonant modes of RDRA can be described with an equivalent
sequential RLC circuit having different sequential LC values. Thus, they form many
series-tuned resonant circuits. The superposition of these modes generally give rise
to resulting or weighted resonant frequency. The top-loading RDRA has been
completely modeled. Antenna gain and bandwidth enhancement techniques
have been worked out with examples.

Keywords Aspect ratio RLC circuit Tuned cavity Weighted sum Resonant

frequency Eigen frequency Design flexibility Top-loading RDRA Gain and
bandwidth enhancement

148 7 Rectangular DRA Higher-Order Modes and Experimentations
7.1 Introduction to Higher Modes
Resonant modes impart physical insight into the radiating phenomenon of the
antenna. Knowledge of modes can be boon to the antenna designer. If antenna
resonant modes are known, radiation parameters can be steered. Any of the
antennas have two types of modes. They are dominant and higher-order modes. The
dominant mode corresponds to the lowest resonant frequency. Other than dominant
frequency, all higher resonant frequencies are higher-order modes. These
higher-order modes can be generated either by increasing electrical length of RDRA
or by applying higher excitation frequency.
The resonant frequencies of the modes are represented by eigenvalues and
currents by eigenvectors. Radiating behavior of the antenna can be predicted by
modes. They can also help to determine input excitation point. Moreover, having in
mind the current distribution of the modes, the geometry of the antenna can be
modied. The aspect ratio is the important parameter in RDRA. Devising control on
aspect ratio can alter resonant frequency, gain, and bandwidth. RDRA has two fold
design flexibility because of two aspect ratios.
The resonant modes of RDRA can be described with an equivalent sequential
RLC circuit having different sequential LC values. Thus, they form many
series-tuned resonant circuits. The superposition of these modes generally gives rise
to resulting or weighted resonant frequency. The modes are dened as E and H elds
pattern inside a device, whose EM wave propagation is governed by Maxwells
equations under certain boundary conditions. RF input excitation currents get dis-
tributed on RDRA surfaces. Thus, weighted sum of eigen currents or superposition
of all these currents inside device is the resultant mode at any instant of time.
The resonant modes are state of excited elds at any instant inside the device,
generally classied as transverse electric (TE), transverse magnetic (TM) and
hybrid electromagnetic (HEM), dominant modes and higher modes. TE modes will
have only Hz component as propagating elds. TM modes will have only Ez
component as propagating elds. These propagating elds are longitudinal elds.
HEM has hybrid mode and will have both Ez and Hz components simultaneously as
propagating elds at any instant of time. These eld perturbations form a particular
excited resonant mode in the device.
In the literature, stacking of the RDRA has been used for enhancement in the
directivity of the antenna by Petosa [1]. This can be achieved by devising proper
control on higher-order modes. Higher modes correspond to higher resonant fre-
quency and higher antenna gain. RDRA higher-order modes and hybrid modes are
useful and provide design space to antenna designers, but congure complex elds
structure. The generation of higher modes mainly depends on RF excitation, device
dimensions, dielectric material, perturbation, and coupling used in RDRA.
An aperture-coupled microstrip slot feed RDRA is discussed in this chapter. This
has the advantage of isolating the feeding network from the radiating element. Aspect
ratio can be changed by changing the RDRA dimensions a, b, and d. This will have
impact on resonant modes, and thus, change in resonant frequency will take place.
7.1 Introduction to Higher Modes 149
Fig. 7.1 a RDRA higher modes. b RDRA mode generated. c RDRA mode control due to dipole
moment. d S11 of RDRA with mode merging. e Higher-order modes eld conguration
The number of half-wave variations corresponding to x, y, and z directions can alter

the operating mode. Once the dimensions of RDRA are xed, modes can also be
altered by excitation frequency. Operating frequency of RDRA has inverse rela-
tionship with permittivity of the material. The device size can be minimized by using
higher permittivity material. Figure 7.1 shows the resonant mode structure.
RDRA shown in Fig. 7.1a has been made with alternate layers of RT-Duroid and
FR-4 dielectric materials having permittivity 10.2 and 4.4. These dielectric mate-
rials are easily available. The fabrication is also simple. The dimensions of these
sheets are 6 6 10 mm3 and 6 6 0.8 mm3, respectively. Figure 7.1b shows
resonant modes. Figure 7.1c shows how these dipoles are broken by introducing air
between these stacking layers. Figure 7.1d presents mode merging due to proper
dipole control, i.e., merging these modes by removing one stacking layer and
creating airspace between these dipoles. This has enhanced the antenna gain.
Figure 7.1e shows higher-order modes in RDRA. Figure 7.2 presents prototype
RDRA with VNA feed probe. Figure 7.3 shows modes or eld pattern. Figures 7.4
and 7.5 show odd and even modes. Figure 7.6 represents structure, and Fig. 7.7
represents gain plot. Figure 7.8 shows microstripline used in RDRA feed.
Figure 7.9 shows RDRA with top loading. Figure 7.10 shows H and E elds
Fig. 7.2 RDRA excitation by aperture-coupled slot
Fig. 7.3 EH elds pattern or resonant modes
Fig. 7.4 Odd number of modes
Fig. 7.5 Even number of modes

Fig. 7.6 Structure of RDRA
Fig. 7.7 Slot position versus gain
Length of strip s length of stub
Lumped port
W (slot)
Fig. 7.8 Slot and microstrip feed
Fig. 7.9 Top-loaded RDRA with a, b, and d = h dimensions
Fig. 7.10 a H eld distribution inside RDRA and b E eld distribution inside RDRA
(a) -2.00 (b) 10.00 m1 m2
-4.00 0.00
dB(GainTotal)
-6.00
dB(S(1,1))
-10.00
Name X Y
-8.00 m1 -30.0000 5.0095

m2 30.0000 5.1616
-20.00 Curve Info
-10.00 dB(GainTotal)
Setup1 : LastAdaptive
Freq='12GHz' Phi='0deg'
-12.00
-14.00 -40.00
8.00 10.50 13.00 15.50 18.00 -200.00 -100.00 0.00 100.00 200.00
Freq [GHz] Theta [deg]
Fig. 7.11 a Return loss at f = 11 GHz and b radiation pattern showing 5.16 dB gain of the antenna
at f = 12 GHz
(a) (b)
-2.00
(c) (d) 10.00
m2
-4.00 5.00
m1
dB(GainTotal)
-6.00
-0.00
dB(S(1,1))
-8.00
-5.00
-10.00 Name X Y
-10.00 m1 -30.0000 5.0674

-12.00 m2 30.0000 5.3275
Curve Info
-14.00 Setup1 : LastAdaptive
-16.00 -20.00 dB(GainTotal)
-18.00 -25.00
8.00 10.50 13.00 15.50 18.00 20.00 -200.00 -100.00 0.00 100.00 200.00
Fig. 7.12 TE111 mode at frequency 10 GHz. a H eld distribution, b E eld distribution, c return
loss, d gain
distributions inside RDRA. Figure 7.11 shows the plot of return loss at f = 11 GHz
and radiation pattern. Figure 7.12 shows TE111 mode at a frequency of 10 GHz.
Figures 7.13, 7.14, 7.15 and 7.16 show resonant modes and Fig. 7.17 TE113 mode
at a frequency of 10 GHz. Figure 7.18 shows TE115 mode at a frequency of 12 GHz.
Figure 7.19 depicted TE117 resonant mode at frequency 15 GHz. Figures 7.20, 7.21,
(a) (b)
(c) -2.50 (d) 10.00 m1
-5.00
5.00
dB(GainTotal)
dB(S(1,1))
-7.50
0.00
-10.00
-5.00
Name X Y
-12.50 m1 0.0000 8.9392

-10.00 Curve Info
dB(GainTotal)
-15.00 -15.00 Freq='12GHz' Phi='0deg'
dB(GainTotal)
-17.50 -20.00
8.00 10.00 12.00 14.00 16.00 18.00 20.00 -200.00 -100.00 0.00 100.00 200.00
loss, d gain
(a) (b)
(c) 10.00
m1
m2
5.00
0.00
dB(GainTotal)
-5.00
-10.00
Name X Y
m1 -20.0000 6.0721
-15.00 m2 20.0000 5.3982
Curve Info
-30.00
-200.00 -100.00 0.00 100.00 200.00
Theta [deg]
Fig. 7.14 TE115 mode at frequency 15 GHz. a H eld distribution, b E eld distribution, c gain
Fig. 7.15 Gain versus 10.00

frequency plot
5.00
0.00
dB(GainTotal)
-5.00
-10.00
Curve Info
dB(GainTotal)
-15.00 Setup1 : Sweep
Phi='0deg' Theta='0deg'
-20.00
8.00 10.00 12.00 14.00 16.00 18.00 20.00
Freq [GHz]
7.22, 7.23, 7.24, 7.25, 7.26, 7.27, 7.28, 7.29, 7.30, 7.31 and 7.32 show parameter
measured in anechoic chamber and HFSS simulated results. Working is mentioned
below each gures.
Resonant modes take the real orthogonal basis for currents on the antenna
surfaces.
In this chapter, mechanism for mode generation and their possible control in
RDRA are discussed. These are validated with simulated and experimental results
using prototype models. Figure 7.2 shows the prototype of RDRA. VNA probes are
connected in order to take measurements. Top-loading RDRA is used for genera-
tion of higher-order modes. These higher-order modes are of even and odd types. If
RDRA design is isolated even as well as odd modes will be available, i.e., both
even as well as odd modes will be present in isolated DRA case. RDRA once
Fig. 7.16 Spacing adjustment between short magnetic dipoles placed at the center of each mode
(a) (b)
(c) 0.00 (d) 10.00 m1
0.00
-10.00
dB(GainTotal)
dB(S(1,1))
-10.00
-20.00
Name X Y
-20.00 m1 0.0000 8.0869

Curve Info
dB(GainTotal)
-30.00 Freq='10GHz' Phi='0deg'
dB(GainTotal)
-40.00 -40.00
8.00 10.50 13.00 15.50 18.00 -200.00 -100.00 0.00 100.00 200.00
loss, d gain
extended with ground plane, only odd modes will be generated, because even
modes get grounded. Thus, ground plane canceled out even modes. The E and
H elds patterns are shown in Fig. 7.3.
Higher-order even and odd modes are shown in Figs. 7.4 and 7.5. These modes
can further be identied as TE/TM/HEM.
If H eld is propagating, then it is TE mode. By contrary, if E eld is propa-
gating, then it is TM mode. Also, when both types of elds, E and H, are excited
simultaneously, then it is HEM mode. HEM modes are most advantageous but have
complex structure. The detailed analysis of hybrid modes is described later in
Chap. 10. Resonant modes can be shifted, merged, and independently controlled by
different techniques. Increasing RDRA electrical length and input excitation fre-
quency can generate higher-order modes into RDRA.
(a) (b)
(c) 5.00 m1 m2
0.00
dB(GainTotal)
-5.00
Name X Y
m1 -40.0000 4.2889
-10.00 m2 40.0000 4.0314
Curve Info
dB(GainTotal)
-15.00 Freq='15GHz' Phi='0deg'
dB(GainTotal)
-20.00
-200.00 -100.00 0.00 100.00 200.00
Theta [deg]
7.2 Resonant Frequency and RDRA Structure
Table 7.1 consists of RDRA specications for prototype.

The structure of the antenna is shown in Fig. 7.6. Slot and microstrip are shown
in Figs. 7.7 and 7.8. The feed is aperture-coupled. The substrate rectangular plane
of 50 50 mm with a thickness of 0.6 mm was used. FR4 was used as RDRA
substrate, and RDRA with a dielectric constant (permittivity) of 10.2 was placed on
top of the substrate. The width of the microstrip used was 1.15 mm. Slot dimen-
sions were 6 mm in length and 1 mm in width. RDRA dimensions were
6 6 5 mm3.
7.2 Resonant Frequency and RDRA Structure 157
(a) (b)
(c) 10.00 Curve Info

dB(GainTotal)
m1
m2 Setup1 : LastAdaptive
5.00 dB(GainTotal)
dB(GainTotal)
0.00
-5.00
-10.00
-15.00 Name X Y
m1 -20.0000 6.1242
m2 20.0000 5.7701
-20.00
-200.00 -100.00 0.00 100.00 200.00
Theta [deg]
Fig. 7.20 Gain versus 10.00

frequency plot at h = 15 mm
7.50
5.00
2.50
dB(GainTotal)
0.00
-2.50
-5.00
Curve Info
-7.50
dB(GainTotal)
Setup1 : Sweep
-10.00 Phi='0deg' Theta='0deg'
-12.50
8.00 10.50 13.00 15.50 18.00 20.00
Freq [GHz]
(a) (b)
m1
(c) 0.00 (d) 10.00
5.00
-10.00 0.00
dB(GainTotal)
dB(S(1,1))
-5.00
-20.00 -10.00
Name X Y
-15.00 m1 0.0000 9.5037

Curve Info
dB(GainTotal)
-30.00 -20.00 Setup1 : LastAdaptive
Freq='9.5GHz' Phi='0deg'
-40.00 -30.00
8.00 10.00 12.00 14.00 16.00 18.00 20.00 -200.00 -100.00 0.00 100.00 200.00
Fig. 7.21 TE115 mode at frequency 9.5 GHz. a H eld distribution, b E eld distribution, c return
loss, d gain
7.2.1 Fields in Rectangular DRA
TE11 resonant mode in rectangular DRA the elds can be dened using dielectric
waveguide model depending upon given boundry conditions
kx2 ky2
Hx coskx x cosky y coskz z 7:1
jxl0
ky kx
Hy sinkx x sinky y coskz z 7:2
jxl0
kz kx
Hz sinkx x cosky y sinkz z 7:3
jxl0
Ex 0
7:4
Ey kz coskx x cosky y sinkz z
Ez ky coskx x sinky y coskz z 7:5
So by solving these equations, we get transcendental equation as follows:

(a) (b)
m1
(c) 10.00
5.00
0.00
dB(GainTotal)
-5.00
-10.00
Name X Y
-15.00 m1 0.0000 8.6772

Curve Info
dB(GainTotal)
-30.00
-200.00 -100.00 0.00 100.00 200.00
Theta [deg]
Fig. 7.22 TE117 mode at frequency 10.5 GHz. a H eld distribution, b E eld distribution, c gain
q
kx d
kx tan er 1k02 kx2 7:6
2
The resonant frequency and propagation constant can be determined from the
transcendental equation.
The characteristic equation is as follows:
kx2 ky2 kz2 er k02 7:7
So, the resonant frequency can be obtained for grounded RDRA as follows:
r
m2 n2 p 2
c
fmnp p : 7:8
2 er a b 2d
Frequency for isolated RDRA is given as follows:

(a) (b)
(c) 10.00 Name X Y
m1 m2 m1 -30.0000 5.7498
5.00 m2 30.0000 6.1212
-0.00
dB(GainTotal)
-5.00
-10.00
Curve Info
-25.00
-200.00 -100.00 0.00 100.00 200.00
Theta [deg]
Fig. 7.23 TE1,1,11 mode at frequency 15 GHz. a H eld distribution, b E eld distribution, c gain
Fig. 7.24 Gain versus RDRA 10.00

height
0.00
Gain [dB]
-10.00
-20.00
-30.00
-40.00
-200.00 -100.00 0.00 100.00 200.00
Theta [deg]
Fig. 7.25 Prototype development of RDRA
Fig. 7.26 Radiation pattern H eld measurements under anechoic chamber
r
m2 n2 p2
c
fmnp p ; 7:9
2 er a b d
kx, ky, and kz can be determined by using characteristic equation. Propagation

constant k0 can be determined in terms of kx. Transcendental equation can be solved
for kz0 using kx. This solution can be obtained using MATLAB for xed value of n,
p, and d. kx will now contain a in place of a. a is the extended electrical length due
to fringing effects. Hence, kx is the complete solution of transcendental equation.
Example 1 Let us determine the resonant frequency for dominant and higher-order
modes of RDRA with given dimensions and dielectric constant:
Fig. 7.27 Radiation pattern E eld measurements under anechoic chamber
Fig. 7.28 Azimuthal radiation pattern measurements inside anechoic chamber

Fig. 7.29 Elevation radiation pattern
Fig. 7.30 Ready for measurement

Fig. 7.31 RDRA mounted with wooden block for measurement
Fig. 7.32 RDRA inside anechoic chamber for measurement

Table 7.1 Specications of S. No. Element Dimension (mm)

RDRA and design dimensions
1 Ground plane 20 30
2 Substrate 20 30 0.8
3 RDRA 4.6 9 10.8
4 Width of microstrip 2.4
5 Length of stub and microstrip 18.693
6 Ground slot (l w) 3.743 0.404
r
m2 n2 p2
c=2
fr p
er a b d
s

3 108 1 1 1
p 106
2 10 10 10 10
r p p
3 108 3 3 3 1011 3 3
p 10 3
1010
2 10 10 2 10 2
fr 2:598 1010 Hz 25:98 GHz
s

3 108 1 1 9
Higher-order mode: fr113 p 106
2 10 10 10 10
p
3 108 103 p 3 1011 p 3 11
p 11 11 1010
2 10 2 10 2
fr 4:9749 1010 Hz 49:749 GHz
7.3 Modes (Resonant) Mathematical Solution
Rectangular dielectric resonator antennas (RDRAs) have received lots of attention

in the last two decades due to several attractive characteristics, such as design
flexibility, high gain, and wide bandwidth. RDRA has two different aspect ratios (b/
a, d/a), high radiation efciency, light weight, and low prole. In contrast, patch
antenna has low gain, less bandwidth, and design flexibility.
The resonant modes are E and H elds pattern. They can be expressed as
follows:
X
Ez umnp x; y; z Re C mnp ejxmnp t ; where; Cmnp are amplitude coefficients:
mnp
7:10
where
23=2 mpx npy ppz

umnp x; y; z p cos cos sin ; orthogonal Fourier basis function:
abd a b d
7:11
v
u
ubmnp2 amnp2
Cmnp u
t hp ppdi2 ; amplitude coefficient: 7:12
p2
sin d
d
" #
amnp cos/mnp bmnp sin/mnp
1
wmnp tan ; Phase: 7:13
amnp sin/mnp bmnp cos/mnp
7.4 Top-Loading RDRA
This chapter is developed based on new approach using a top-loading excited

RDRA as shown in Fig. 7.9, for generating higher-order modes. The even as well as
odd modes can be generated even with ground plane. It has also been studied that
even modes (in z-direction) were short-circuited, when RDRA was placed on a
ground plane. Short magnetic dipoles are basis for generation of these resonant
modes. Nearly, /3, /2, and heights of the dielectric resonator generated TE11,
TE13 and TE15 (0 < 1) modes. The gain is found to be increasing in higher
modes. This is also evident from the ndings that gain of RDRA starts decreasing or
reducing even in higher modes, when magnetic dipoles start overlapping. This
overlapping of dipoles can be seen when the wavelength used is very small.
Top loading excited both even and odd modes. Simulations have shown that 1st
and 3rd resonances got shifted toward 2nd resonance, when the space s between
top and bottom RDRAs varies. Merging of neighboring resonance modes could be
done using this method. This is an excellent phenomenon, which can be used for
bandwidth enhancement. This merging of bands helped to increase the antenna
bandwidth. Thus, existing patch antenna gain and bandwidth can also be increased
by using the concept of higher-order modes. Blocking or shifting of any modes has
become possible in RDRA.
E and H elds perturbations in RDRA can be introduced by carrying out small
changes in the structure, or this can be obtained by input excitation currents. This
perturbation gets converted into eigenvector or eigenvalues. The perturbations are
proportional to eigenvector and resonant mode. The number of modes is directly
related to number of lobes occuring in radiation patterns. There are two ways in which
the number of modes can be increased in RDRA: One is by increasing the RDRA
dimensions a, b, and d, and the other is by increasing the input excitation frequency.
7.5 Simulated HFSS Results 167
7.5 Simulated HFSS Results
From Fig. 7.10, it is evident that single resonant mode, as one half-wave variation
take place in x-direction.
The gain of the antenna got enhanced due to increase of RDRA height. TE111
mode resonant frequency is 10 GHz, TE113 mode frequency is 12 GHz and TE115
frequency is 15 GHz. Figure 7.12 shows the magnetic and electric eld distributions
inside RDRA. The return loss and radiation pattern are shown in that gure.
It has been seen from Fig. 7.13 that gain has been enhanced. Figure 7.14 shows
TE115 mode, which has higher resonant frequency. Figure 7.15 shows gain versus
frequency plot. Figure 7.16 depicts the spacing between short magnetic dipoles.
The above results obtained from the analysis of RDRA carried out revealed that
higher-order modes offer high gain until dipole overlapping does not take place.
The decrease in gain due to overlapping of short magnetic dipoles takes place. This
will happen when there is a less spacing between two short magnetic dipoles.
Hence, minimum spacing between short magnetic dipoles must be equal to 0.4. If
the spacing between short magnetic dipoles is less than this limit, then the gain will
be reduced. This is depicted by simulations in Fig. 7.14. TE113 gain has been
reduced even at TE115 as shown in Fig. 7.18. Now, if we obtain TE115 mode with
increase in RDRA height, then more gain can be obtained. This is the reason why
gain at TE115 is less than TE113 as shown in Figs. 7.14 and 7.15.
At h k=2
When h = 15 mm, three modes got generated, i.e., TE113, TE115, and TE117 cor-
responding to 10, 12, and 15 GHz, respectively. It is clear that gain has been
decreased at higher modes due to the reason explained earlier that spacing between
short dipoles placed at the center of the eld is less than 0.4 . At frequency 10 GHz
inside RDRA, there is proper spacing between these dipoles; hence, the gain is
maximum. Various excited modes are shown in Figs. 7.16, 7.17, 7.18, 7.19, 7.20,
7.21, 7.22 and 7.23 are excited, i.e., resonant modes in RDRA. Their results have
also been taken on S11 plots.
hk
Here, the height of RDRA has been chosen as h = 30 mm, and mode was
operating at 10.5 GHz. The highest gain was due to the same concept of spacing of
short magnetic dipoles. From Figs. 7.17, 7.18, 7.19 and 7.20, very important fact is
noticed that when spacing between short magnetic dipoles was reduced, then the
order of mode becomes high, while the power of main lobe was distributed to side
lobes. Thus, the gain of the antenna was reduced at higher mode if the spacing is
less than 0.4 .
In the above gures, generation of higher modes, limitation, and their effect on
antenna gain have been clearly shown.
7.6 Modes at Varying Heights of RDRA
The comparison between three RDRAs of different heights have been made. It was
noticed that RDRA having less height and operating at lower mode offers less gain
but higher bandwidth. On increasing the height of RDRA, the gain of the antenna is
found to be higher along with directivity, thus narrowing the beam width.
Figure 7.24 shows that gain is increasing, when the height of RDRA is increased.
7.7 Distortions Due to Overlap of Dipole Moment
RDRA of height /3, /2, and operating around 1115 GHz consists of TE111,
TE113, and TE115 modes. This fullls the requirement of separation of magnetic
elds by spacing 0.4 . But when RDRA with same height operates at higher
frequency, then the spacing between dipoles is reduced. The power of main lobe is
distributed to the side lobes, which creates the loss of antenna power and gain.
Hence, any desired resonant modes inside the device can be excited for desired
radiation pattern at known resonant frequency. The higher modes amplitude coef-
cients equation has been developed. Modes can be used to visualize corresponding
radiation pattern and polarization of the antenna. Modes give physical insight into
eigenvalue for determining resonant frequency and feeding point for 50
impedance.
7.8 Prototype and Anechoic Chamber Experimentations
Prototype RDRA was made, and it was tested inside the anechoic chamber using
VNA. Results for radiation pattern and other antenna parameters have been taken
and are shown in Figs. 7.26, 7.27, 7.28, 7.29, 7.30, 7.31 and 7.32. Each gure is
captioned below for the results.
It was seen in RDRA of particular height, more number of higher-order mode
can be excited by applying another excitation on the top loaded RDRA as shown in
Fig 7.3. The reduced spacing s between top and bottom RDRAs, merged even
modes, thus increased bandwidth of the antenna. The RDRA under top loading
converted few odd modes to nearest even mode. Thus, both even and odd modes
were made available due to top loading. Thus, spacing s seems to control band-
width of RDRA. High gain, miniaturization, high bandwidth, directive antenna can
be designed by having proper control or maneuvering resonant modes.
The design of this antenna offers wide scope of achieving wide bandwidth along
with high gain. The application of this antenna includes satellite tracking, air trafc
control Wi-, Wi-max, and mobile communication.
7.8 Prototype and Anechoic Chamber Experimentations 169
By developing control on modes, we can control beam width of antenna and can
restrict the reception of signal to a particular area and hence it can be used for
military applications. Presently, we face the problem of TV signal reception during
rainy season, due to the absorption of signal by rain drops due to signal being single
polarized either TE or TM. This can be minimized by application of dual polarized
or hybrid antenna. The other application could be miniaturization of antenna. By
keeping the dimensions of antenna xed, the mode of antenna can be changed by
changing the permittivity of RDRA and thereby changing the frequency. To
automate the mode generation, microcontroller-based lookup table can generate
possible combinations of bandwidth, gain, and frequency.
7.9 Adjacent Modes Combination for Broadband

Applications
The merging of adjacent bands or neighboring modes of RDRA can be used for
enhancement of the bandwidth. By varying the aspect ratio, three resonant bands
can be obtained for useful operation as shown in Fig. 7.33.
The dimensions of RDRA are given in Table 7.2. Figure 7.33b shows the return loss
of the antenna with three bands resonating at 2.89 GHz at dominant mode, i.e., TE111
mode and at 3.61 GHz for TE121 mode and at 4.6 GHz for TE131. Figure 7.34 shows H
and E elds distributions inside RDRA. The direction of the electric eld is indicated
by arrow.
Figure 7.35 depicted that the lower gain at lower mode and high gain at higher
modes.
7.10 Effect of Air Gap Between RDRA and Ground Plane
The effect of the air gap between ground plane and RDRA is shown in Fig. 7.36.
Table 7.3 shows the variation in the resonant bands of the antenna. Effect of
introducing the gap between RDRA and ground plane is depicted in Fig. 7.36.
Results obtained by simulation along with the results obtained by calculations
clearly indicate the effect of air gap. The modes are spreading as the frequency of
the modes is shifted in forward direction with respect to increase in the gap.
Tables 7.1, 7.2, 7.3, 7.4, 7.5 and 7.6 are the results tabulated for various
simulations.
0.00
(a) (b)
-2.50
-5.00
dB(S(1,1))
-7.50
-10.00
-12.50 m1 m2
-15.00
Name X Y
m1 2.8963 -13.4462
-17.50 m2 3.6187 -13.2545 m3
m3 4.6120 -19.2167
-20.00
2.50 3.00 3.50 4.00 4.50 5.00
Freq [GHz]
Fig. 7.33 a Top view of the DR. b Return loss
Table 7.2 Simulated and Mode/gap Results G=0 G = 0.01 G = 0.02

calculated resonant
frequencies of the modes in TE111 Simulated 2.89 2.83 2.88
DR Calculated 2.84 2.85 2.87
TE112 Simulated 3.67 3.83 Omitted
Calculated 3.69 3.71 3.74
TE113 Simulated 4.64 4.72 4.81
Calculated 4.62 4.69 4.73
7.11 Effect of Asymmetrical Wells Inside RDRA
When two asymmetrical wells are inserted, then these modes start to merge together
and all the bands are shifted. This is shown by return loss graph in Fig. 7.38.
Comparison between asymmetrical wells and without asymmetrical wells has been
made. This is to note that by adding wells, higher frequency bands get shifted more
as compared to lower frequency bands as shown in Fig. 7.40. Dimensions of the
structure are shown in Table 7.4. The results were taken for various values of width
of well as 0.5, 1.1, 2.0 mm respectively are shown in Fig. 7.38.
7.12 Effect of Moat Insertion Inside RDRA
In this way of merging of modes takes place when air gap is inserted in the structure
of RDRA as shown in Fig. 7.41 then we get broader bandwidth. Plot for reflection
coefcient in Fig. 3.9 shows clearly the effect of moat in the structure. All the
dimensions of the structure are shown in Table 7.3.
7.13 Effect of a/b and d/b Aspect Ratio 171
Fig. 7.34 Magnetic and electric eld distributions inside DR. a H eld TE111 mode, b E eld
TE111 mode, c H eld TE112 mode, d E eld TE112 mode, e H eld TE113 mode, f E eld TE113
mode
7.13 Effect of a/b and d/b Aspect Ratio
The effect of length and width of RDRA is such that if we increase the dimensions
then there can be large number of modes generated. The effect of a/b and d/b ratio
has been speculated in the manner such that when the ratio a/b is increased, the
modes come closer to each other and merged, and when the ratio d/b is increased,
resonant frequencies of all modes are diverged. Further, if the ratio d/b is reduced,
then the modes are merged. So here, we increased the a/b ratio and reduced the d/
b ratio, and then we pointed aristocratically that intensied changes like modes have
been merged to increase the bandwidth of the device. The important thing to note is
that mode TE112 has been merged, and there are the resonant frequencies of modes
TE111 and TE113 only. Figure 7.31 reflects the effect of the overall process.
Fig. 7.35 Showing the gain 10.00 m2

of the antenna for frequency
m1
range 5.00
0.00
dB(GainTotal)
-5.00
-10.00
-15.00
-20.00
Name X Y
-25.00
m1 2.8930 4.9158
m2 4.6572 8.2124
-30.00
Fig. 7.36 Effect of air gap 0.00

m3
between RDRA and ground
plane m2
-10.00
m1
-20.00
S11 [dB]
-30.00
-40.00
-50.00
2.50 3.00 3.50 4.00 4.50 5.00
Freq [GHz]
Table 7.3 Modes and their Mode Resonant frequency (GHz) Gain (dB)
resonant frequencies
TE111 4.56 5.2
TE112 4.96
TE113 5.56 8.78
Figures 7.35, 7.36, 7.37, 7.38, 7.39, 7.40, 7.41, 7.42, 7.43 and 7.44 presented effect
on change in aspect ratios. Tables 7.3 and 7.4 indicated design parameters.
Tables 7.5 and 7.6 show resonant modes.
Table 7.4 Dimensions of all the structures

Dimension
Parameter/structure Aperture-coupled DR with two DR after increasing in DR with moat
DR asymmetrical a/b and decreasing in
wells d/b ratio
Ground plane and 70 70 and FR_4 70 70 and 70 70 and FR_4 70 70 and
substrate (mm) and epoxy (4,4) FR_4 epoxy epoxy (4,4) FR_4 epoxy
permittivity (4,4) (4,4)
DR (mm) 28 9 10 28 9 10 30 19 4 30 19 4
Inner
dimensions:
16.5 10 4
Moat gap
(mm):
Gi = 0.1,
G2 = 5.3,
G3 = G4 = 0.2
Microstrip (mm) 35 1.15 38 1.15 37 1.15 40 1.15
Slot (mm) 10 2 10 2 71 13.5 2
Ds (mm) 8 7 12 13
Ls (mm) 6 6 4 5
Table 7.5 Comparison among all the structures

Comparison between all the structures
Parameter/structure Aperture-coupled DR with two DR after DR with Moat
DR Asymmetrical Increasing in
Wells a/b and
decreasing in
d/b ratio
Modes and TE111(2.92), TE111(2.92), Modes are Modes are
frequency (GHz) TE112(3.70), TE112(3.70), merged with merged with
TE113(4.64) TE113(4.64) each other, each other, gain
gain is is enhanced, and
enhanced, and the bandwidth is
the bandwidth increased
is increased drastically
Gain (dB) 1.7 2.1 3 4.8
Bandwidth (GHz) Less Less Bandwidth is Bandwidth is
enhanced by enhanced and is
merging the larger, and
modes modes have the
closer resonant
frequency to
each other
Table 7.6 RDRA simulated resonant modes parameters

Height Excitation at top of Generated Frequency Gain Bandwidth
h (mm) the DR Mode f (GHz) (dB) (GHz)
5 No TExd11 11.3 5.6 10.112.0
Yes TExd11 11.8 4.8 9.517.7
TExd12 15 4.5
10 No TExd11 10 5 9.2512.1
5 12.8114.85
TExd15 13.7 7 16.517.1
Yes TExd14 13 6 9.416.8
15 No TExd13 10 9.1 9.4811.4
TExd17 16 5.6 15.816.4
6 18.418.6
Yes TExd14 11.7 5 9.317.6
TExd16 13.7 4.3
TExd18 16.7 4.5
30 No TExd;1;5 9.8 8.2 9.3011.85
TExd;1;11 12.56 6 12.4512.70
TExd;1;15 15.95 5.8 15.7016.20
TExd;1;17 17 5.9 16.9017.25
6 18.2018.50
Yes TExd;1;6 10.83 5 9.3517.1
TExd;1;12 13.58 5.1
TExd;1;14 14.88 4.7
TExd;1;16 15.7 5.1
Fig. 7.37 Structure after insertion of two asymmetrical wells
Fig. 7.38 Showing the effect 0.00

of addition of wells
-5.00
m2
-10.00
m3
-15.00
S11 [dB]
-20.00 m1
-25.00
-30.00 Without wells
-35.00 With wells Without

wells
-40.00
2.50 3.00 3.50 4.00 4.50 5.00 5.50
Freq [GHz]
Fig. 7.39 Merging of second band in rst and eld distributions. a Hx eld at f = 2.9 GHz, b E
eld at f = 2.9 GHz, c Hx eld at 3.6 GHz, d E eld at 3.6 GHz
Fig. 7.40 Gain after insertion 10.00 m2

of wells m1
5.00
0.00
dB(GainTotal)
-5.00
-10.00
-15.00
Name X Y
-20.00
m1 2.9314 5.4961
m2 5.0686 7.8804
-25.00
-30.00
2.50 3.00 3.50 4.00 4.50 5.00 5.50
Freq [GHz]
Fig. 7.41 Structure of the antenna after insertion of moat inside RDRA and eld distribution at
frequency 4.56 GHz
(a) 0.00 (b) 3.75 m1

-2.50 2.50
-5.00
dB(S(1,1))_1
0.50
dB(GainTotal)
-7.50
-10.00 -1.50
-12.50
-3.50
-15.00
-17.50
-5.50 Name X Y
-20.00 m1 4.7559 2.9000
-22.50 -7.50
4.00 4.50 5.00 5.50 6.00 6.50 7.00 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00
Freq [GHz] Freq [GHz]
Fig. 7.42 a Frequency response showing return loss. b Gain of the antenna over frequency
-2.00
0.00
-4.00
-5.00
-6.00
dB(S(1,1))
S11 [dB]
-8.00 -10.00
-10.00
-15.00
-12.00
-20.00
-14.00
- - - - with moat
-16.00 -25.00
4.00 4.50 5.00 5.50 6.00 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00
Freq [GHz] Freq [GHz]
Fig. 7.43 Return loss and comparison with the structure with moat
Fig. 7.44 Gain after 10.00 m1

increasing a/b ratio and
decreasing d/b ratio m2
5.00 m3
dB(GainTotal)
0.00
- 5.00
Name X Y
m1 5.4983 8.7898
-10.00 m2 4.5686 5.2020
m3 4.9699 4.0093
-15.00
4.00 4.50 5.00 5.50 6.00
Freq [GHz]
It is found that merging of modes can be very advantageous for broadband

applications. The aspect ratio plays very important role in this phenomenon.
Excitation applied at top of the RDRA converted odd modes into nearby even
modes.
Reference
1. Mongia RK, lttipiboon A, Cuhaci M, Roscoe D (1994) Radiation Q-facter of rectangular

dielectric resonator antennas theory and experiment. In: IEEE international symposium on
Antennas Propagation Society, Seattle, WA, pp 764767, June 1994
Chapter 8
RDRA Angular Excitation Mathematical
Model and Resonant Modes
Abstract This chapter narrates angular excitation mathematical modeling of

RDRA. The shift in radiation pattern and resonant modes have been realized based
of angular shift in input. Slot is the source of input voltage to RDRA. Slot size and
orientation effects loading of RDRA. The resonant characteristics of a RDRA are
dependent on shape, DRA volume and excitation. The excitation current can be
dened in terms of magnetic vector potential A based on applied current densities
J. This A can be expressed in terms of E and H elds or as S Poynting vector.

Keywords Slot Angular variation Change in radiation pattern Resonating

modes Power flux HFSS VNA Hardware model Anechoic chamber
8.1 Introduction
Slot is the source of input to RDRA. Slot size and orientation is responsible for
loading of RDRA. The angular orientation of slot has been investigated in this
chapter with simulations and experimentation. The resonant characteristics of a
RDRA depend upon the shape and size of the (volume) dielectric material along
with feeding style. It is to be appreciated that in a RDRA, it is the dielectric material
that resonates when excited by the feed. This phenomenon takes place due to
displacement currents generated in the dielectric material. The excitation current
can be dened in terms of magnetic vector potential A based on the current
densities J inside the resonator, at any far-eld point. This A can be expressed
in terms of E and H elds. Later, this is expressed as S Poynting vector. Now the
flux described can be treated with boundary conditions to nd Radiated power Prad
into space. Figure 8.1 presented RDRA excited at slot angle. Figures 8.1 and 8.2 are
HFSS model of RDRA. In Fig. 8.3, slot is shifted with certain amount of angle. If
two slots are placed at 90, circular polarization will take place. If one slot area is
larger than the other, then LHCP (left-hand circular polarization) or RHCP
(right-hand circular polarization) will take place. Figure 8.4 RDRA is excited at 45
angle. Figures 8.5, 8.6 and 8.7 presented radiation pattern at slot angles. Using two

182 8 RDRA Angular Excitation Mathematical Model
Fig. 8.1 RDRA with slot at an angle (0 ; /0 ), a, b, and d are dimensions
Fig. 8.2 Let rectangular DRA excited by slot at an angle hi ; /i
Fig. 8.3 Slot at an angle (i ; /i ) shifted to left
cross slots circular polarization can be integrated. If two slot of different lengths are
used then due to differential signal LHCP and RHCP can be generated. This
indicates that a mechnism for polarization cantrol can become possible if these
slots are arranged in a particular manner.
Fig. 8.4 RDRA angular excitation left side (3040)
Radiation Pattern 6 HFSSDesign1 Radiation Pattern 6 HFSSDesign1

0 Curve Info 0 Curve Info
rETotal rETotal
-30 30 Setup1 : Sweep -30 30 Setup1 : Sweep
12.80 Freq='8.451903808GHz' Phi='0deg' 9.60 Freq='8.451903808GHz' Phi='0deg'
rETotal rETotal
Setup1 : Sweep Setup1 : Sweep
9.60 7.20
Freq='8.451903808GHz' Phi='90deg' Freq='8.451903808GHz' Phi='90deg'
-60 60 -60 60
6.40 4.80
3.20 2.40
-90 90 -90 90
-120 120 -120 120
-150 150 -150 150

-180 -180

rETotal rETotal
rETotal rETotal
9.00 9.00
-60 60 -60 60
6.00 6.00
3.00 3.00
-90 90 -90 90
-120 120 -120 120
-150 150 -150 150
-180 -180
HFSSDesign1 HFSSDesign1
Radiation Pattern 8 Curve Info Radiation Pattern 8 Curve Info
0 rETotal 0 rETotal
-30 30 Freq='16.64328657GHz' Phi='0deg' -30 30 Freq='16.64328657GHz' Phi='0deg'
14.00 rETotal 14.00 rETotal
10.50 10.50
-60 60 -60 60
7.00 7.00
3.50 3.50
-90 90 -90 90
-120 120 -120 120
-150 150 -150 150

-180 -180
Fig. 8.5 Radiation pattern at an angle (30 40 to the left)
The detailed description of radiation phenomenon is given below. The resonator

RDRA radiates from the fringing elds. The resonator acts as tuned sequential RLC
circuits having different values of LC or resonant cavity with an electric eld
perpendicular to the resonator, that is, along the Z-direction. The magnetic eld has
Fig. 8.6 Angular excitations in RDRA to the right side

rETotal rETotal
rETotal rETotal
8.40 8.40
-60 60 -60 60
5.60 5.60
2.80 2.80
-90 90 -90 90
-120 120 -120 120
-150 150 -150 150

-180 -180

rETotal rETotal
rETotal rETotal
9.00 9.00
-60 60 -60 60
6.00 6.00
3.00 3.00
-90 90 -90 90
-120 120 -120 120
-150 150 -150 150

-180 -180

rETotal rETotal
rETotal rETotal
9.00 10.50
-60 60 -60 60
6.00 7.00
3.00 3.50
-90 90 -90 90
-120 120 -120 120
-150 150 -150 150

-180 -180
Fig. 8.7 Radiation pattern when slot is shifted to the right
vanishing tangential components at the four side walls. The four extended edge
surfaces around RDRA serve as the effective radiating apertures. These fringing
elds extend over a small distance around the side walls and can be replicated as
elds Ex that are tangential to the substrate surface. The only tangential aperture
eld on these walls is E = Ez, because the tangential magnetic elds vanish by the
boundary conditions. The ground plane can be eliminated using the image theory,
resulting in doubling the aperture magnetic currents, that is, J = n E. Hence, the
effective tangential elds can be expressed in terms of the eld Ez. Now, radiated
power pattern can be compared with the modes generated inside the resonator. The
surface current density can be the main source of EH elds pattern when applied
with boundary conditions inside the resonator. This can be correlated with far-eld
pattern. The physics of this radiation is based on the fringing effect due to dipole
moments. First derivative is velocity elds, and then, the second derivative on
dipole moments can be termed as acceleration, which is main source of radiations.
Hence, steering of the resonant modes mainly depends on excitation. Ez, Hz, or both
Ez and Hz elds at any instant of time can dene TM, TE and HEM modes.
8.2 Angular Shift in Excitation
Let aperture-coupled microstrip with slot and stub (feed) is situated in xy plane of
RDRA at bottom part and slot placed at an angle hi ; /i as shown in Fig. 8.1. The
resonator modes and radiation pattern generated have been investigated as follows:
1. Hz ; Ez elds are longitudinal. These have been expressed in terms of
orthonormality with signals umnp x; y; z and vmnp x; y; z at frequency xmnp
based on the Maxwells equations with given boundary conditions of RDRA.
2. At
z = 0; surface (x, y)excitation is applied with slot and surface current density
Jsx x; y; t; Jsy x; y; t is developed into RDRA.
3. The surface electric current density is equated with generated magnetic elds
into RDRA:

Js x; y; d Jsx ; Jsy ^n H Hy ; Hx ;
at z = 0; amplitude coefcients are obtained on expansion of Hz , Ez in terms of

Cmnp and Dmnp .
4. Equating tangential component of Ez at boundary, i.e., Ey jz0 to zero, the
amplitude coefcients Dmnp for Hz and Cmnp of Ez are expressed.
5. Feed position in xy plane can be dened as follows:
x0 ; y 0 / 0 ; h0
8 9
<1 x0 x dllength =
6. f x; y j yj W width
: ;
0 otherwise
7. Excitation current in time domain can be expressed as:
Js x; y; z; t
where is the
angle of variation in excitation.
Here, we apply excitation through slot dl at some specic angle. Later, shift in the
position of slot is provided. Change in radiation pattern or resonant modes is inves-
tigated with mathematical equation, simulations and experimentations on RDRA.
!
lI dl ejkr ~
A; where A is magnetic vector potential 8:1
4pr
Z
A V E dl
E jx ~
A
Radiated power
2 r
jE j2 x2 ~
A l
; g impedance: 8:2
2g 2g
Z
J r 0;x ejkjrr j 3 0
0
~ l
A d r ; at source: 8:3
4p jr r 0 j
Volume
We know that radiation pattern can be dened by Eh ; E/
Eh jxAh and Ah ^h A
Antenna current density can be expressed as follows:

X
J r 0 ; x J s mnp; r 0 ejxmnpt 8:4
mnp
The magnetic vector potential in terms of J can be written as follows:

jrr0 j
Z
l X J s mnp; r 0 ejxmnp t c
A dsr 0 ; ds is surface
4p mnp jr r 0 j
Z 8:5
l ejkn X 0
0
J s mnp; r 0 ejxmnp^rr dsr 0
4p jr r j mnp
s

Eh ; E/ ; H/ Eh =g; Hh E/ g:
8.2 Angular Shift in Excitation 187
Radiated power can be given as follows:
1 2
Prad jEh j2 E/ 8:6

2g
(
^h ^x cos u cos h ^y sin u cos h ^z sin h
/^ X ^ sin u Y^ cos u
l XZ
Eh Re fJsx mnp; r 0 cos u cos h
4pr 2 mnp
s
o
Jsymnp; r0 sin / cos h Jszmnp; r0 sin h

mnp
exp jx X 0 cos / sin h Y 0 sin / sin h Z 0 cos hdsr 0 ejxmnpt
c
8:7
XZ n o
Eu Re Jsx mnp; r 0 sin / Jsymnp; r0 cos /
mnp
xmnp
ej c X 0 cos u sin h Y 0 sin / sin h z0 cos hdsr 0 ejxmnpt dsr 0 8:8
Radiated power Prad can thus be dened as:

Z
xmnp^r r 0
Px ^r j mnp Jsx mnp; r 0 ej c dsr 0 8:9a
s
Z
xmnp^r :r 0
Py ^r j mnp Jsy mnp; r 0 ej c dsr 0 8:9b
s
Z
xmnp^r:r 0
Pz ^r j mnp Jsz mnp; r 0 ej c dsr 0 8:9c
s
^ cos / sin h Y^ sin / sin h Z^ cos h:

^r h; / X
Let
s mnp for convenience; then
Xn o
Eh Re Px ^rjs cos / cos h Py ^r js sin / cos h Pz ^rjs sin hgejxst
X
s
8:10
Re Esh ejxst
s
3 2
000
where s mnp 4 001 5 and so on till s 111: Similarly
010
X
E/ Re Px ^r js sin u Py ^r js cos /ejxst 8:11
s
Hence, Radiation Pattern of RDRA: Power flux per unit solid angle will describe
the pattern. Power radiation pattern can be dened as follows:
2 ( !)
jEh j2 E/ 1 X
jxst jxst
Esh e Es/ e
2g 2 s
( !)
1 X X
jxst jxst
Hsh e Hs/ e
2 s s
!
1 X X
Es Hm ejxs xm t Es Hm ejxm xs t
4 s s
1X
8:12
Esh Hs Es Hsh
4 s
1 X

Re Esh Hs
2 s
1 ^
E E

Esh h E ^ sh
^ s ^h
u
2 g g
X jEsh j2 jEs j 2
^r ^r
s
2g 2g
Poynting vector
1 Xh
S ^r xs2 jPx ^r js cos / cos h Py ^r js sin u cos h Pz ^r js sin hj2
2g
2 i
xs2 jPx ^rjs sin u Py ^r js cos u 8:13
Radiated power per unit solid angle or Poynting vector
1 X
S ^r r; h; / xmnp2 fjPx h; /jmnp) cos / cos h
2g mnp
8:14
Py h; /jmnp sin / cos h Pz h; /jmnp sin h2

jPx h; /jmnp sin / Py h; /jmnp cos /2
8.3 Radiation Pattern Based on Angle 0 ; /0 Variation in xy Plane 189
8.3 Radiation Pattern Based on Angle 0 ; /0 Variation

in xy Plane
^
~ mnp pp Dmnp ejwmnp
Let D
d h2mn
~ mnp ej/mnpp=2 C
C ^ mnp
h2mn
" #
X X n o
H? ~ mnp ejxmnpt gr? ~umnp x; y; z
Re fD Re C~ mnp ejxmnpt r? ~
umnp x; y; z
mnp mnp
8:15
Probe orientation
nh0 ; /0 ^x cos /0 sin h0 ^y sin /0 sin h0 ^z cos /0

^
^n H Js
h2mn c2 k2 nx^x ny^y nz^z
jpp
c ; for all wave guide and
d
d d
let c ; jx ; for all cavity resonator
@z dt
dl cos h0 d; Probe length
Matrix-based computations are as follows:

Ez E?
Hz H?
Hx Jsy
Hy Jsx
H? Hx^x Hy^y
X
Ez Re C mnp umnp x; y; z expjxmnp t 8:16
mnp
c jx
H? r? Hz r? Ez ^z 8:17
h2mnp h2mnp

jppz
Hz umn x; y exp 8:18
d
Hz umn x; yCmnpejxmnpt d
jppz
n o
Hz Re Cmnp umn x; ye jxmnpt d
jppz
n jppz
o
Hz Re C1 mnpumn x; yejxmnpt d
n jppz
o 8:19
Re C2 mnpumn x; yejxmnpt d
" #
1 d @
H? 2 r? Hz x; y; z; t 2 r? Ez x; y; z; t 8:20
hmn dz hmn @t
Hence
X
Ez x; yz; t Cmnp umnp x; y; z cosxmnpt umnp 8:21a
mnp
X
Hz x; yz; t Dmnp umnp x; y; z cosxmnpt wmnp 8:21b
mnp
8.4 Replacing Probe with Slot of Finite Dimensions

(Ls, Ws) at an Angle h0 ; /0
We replace excitation probe to slot umnp and vmnp by ~umnp and ~vmnp
p mpx
npy
qpz
2 2
umnp x; y; z p sin
~ sin sin 8:22a
abd a b d
p mpx
npy
qpz
2 2
~vmnp x; y; z p cos cos cos 8:22b
abd a b d
0
Imnp eg xmnpt
Js x; y; d ; where g0 is an angle
x
X n o
H ? jzd Re D~ mnp ejxmnpt r?~vmnp x; y; z
mnp
X n o 8:23
Re C~ mnp ejxmnpt r? ~umnp x; y; z
mnp
8.4 Replacing Probe with Slot of Finite Dimensions (Ls, Ws) at an Angle h0 ; /0 191
^n H
Ex ; Ey ; Hx ; Hy are the elds in terms of surface current density due to applied probe
current at an angle ; Jsx ; Jsy can be expressed as current density using Fourier
coefcients; Cmnp and Dmnp Hy Hx elds can be computed from Ex Ey fields; prop-
agation terms h2mn c2 k 2
Imnp jgmnp
H? e f^x sin/0 cos/0 ^yggx; y
xwidth
p n mpx
npy
o
X
~ jxmnpt 2 2
H? RefDmnp e p r? cos cos : 8:24
mnp abd a b
H ? jz0 HX x; y; 0^x HY x; y; 0^y

Jsy x; y; t Jsx x; y; t
X n o mp
2p2 mpx
npy
Jsy x; y; t ~ mnp e
Re D jxmnpt
p sin cos
a abd a b
Za Zb mpx
npy
2
Jsy x; y; t p sin cos dxdy
ab a b
0 0
n o mp
r
2 mpx
~
Re Dmnp e jxmnpt
sin
a d a
Hence,
n o
Jsy x; y; t Re fy x; yejxmnpt
Za Zb r
mpx
npy
2 2 mp ~
fy x; y p ; Dmnp sin cos dxdy 8:25
ab d a a b
0 0
Now
r Za Zb mpx
npy
~ mnp 2 a 2
D fy x; y p sin cos dxdy 8:26
d mp ab a b
0 0
Hy Jsx
Jsx x; y; t H y jz0
X n o
2p2 mpx
npy
Re D ~ mnp ejxmnpt np p cos cos

mnp
b abd a b
8:27
n o
Jsx x; y; t Re fx x; yejxmnpt
r Za Zb mpx
npy
~ mnp 2 2 b
D fx x; y cos cos dxdy
ab d np a b
0 0
X
Jsx x; y; t Reffx fx; yjmnpgejxmnpt 8:28
mnp
Similarly; compute D ~ mnp :

X
Jsy x; y; t Reffy fx; yjmnpgejxmnpt 8:29
mnp
8.5 HFSS Computed Radiation Pattern with Shifted

hi ; /i Slot Positions
Angular excitation at 45 and 30 left side.

Results of angular excitation on radiation pattern have been evaluated on HFSS
and shown in Fig. 8.5.
Angular shifts in excitation at 45 and 30 right side and radiation pattern are
shown in Fig. 8.7 and summarized are placed in Table 8.1.
Results of Radiation pattern when angular excitation is given to the right side
have been shown in Fig. 8.7.
Table 8.1 Summarized results

Frequency S11 in S11 in dB S11 in dB S11 in dB Gain Gain Gain Gain
in GHz dB slot slot position cross-slot in dB in dB in dB (cross-slot)
position 45o right at left right in dB
at 45o left 45o 45o
8.45 11.2 11.6 10.8 11.5 1.5 3.75 2.7 1.5
12.81 13.3 12.4 13.4 12.2 1.1 1.0 1.1 1.2
16.64 14.7 18.1 18.0 15.9 1.5 1.8 1.6 3.0
8.5 HFSS Computed Radiation Pattern with Shifted hi ; /i 193
Table 8.1 described results of antenna parameters in tabular form. It has been
observed that angular variation in excitation has direct impact on the radiation
pattern as well as number of modes generated. This has been veried by the plots
given above. We have taken measurements of radiation pattern on varying slot of
feed at 30 and 45 to left and right from its original position. These plots have been
veried at two different frequencies. This completes the solution.
8.6 Experimentations
Figures 8.8, 8.9, 8.10, 8.11, 8.12, 8.13, 8.14, 8.15 and 8.16 present the experimental
results of RDRA. Their signicance is placed below each gure. The RDRA made
from acrylic glass sheets having dimensions of 9, 6 and 3 cm. The silicon oil having
e 2:2 was used as RDRA dielectric material. The resonant frequency of RDRA
was measured to 4.55 GHz. The measurements were taken at various angular
positions of the slot. Aperture-coupled feed RDRA is shown in Fig. 8.1. The feed
position was shifted to investigate RDRA S11 using VNA 40 GHz. The results are
shown in Figs. 8.2, 8.3, 8.4 and 8.5.
Fig. 8.8 RDRA under measurements with VNA

Fig. 8.9 28.23 dB measured S11 of RDRA at 4.63 GHz
Fig. 8.10 S11 RDRA with shifted slot resonant frequency 3.67
8.6 Experimentations 195
Fig. 8.11 Smith chart showing proper Z11 of RDRA
Fig. 8.12 S11 at shifted slot frequency 3.57 GHz

Fig. 8.13 Measurements of RDRA with aspect ratio changed
Fig. 8.14 RDRA aspect ratio changed

8.6 Experimentations 197
Fig. 8.15 Shifted frequency observed (3.60 GHz)
Fig. 8.16 Aperture-coupled feed showing slot

The results obtained with VNA have clearly shown shift in resonant frequency
due to feed orientation. It indicated that resonant modes are changing based on the
slot orientation. Hence, it is clearly evident that radiation pattern can be steered with
slot position in RDRA. If these results can be placed in look up table, then
microcontroller-based orientation can result into automated antenna. This auto-
mated antenna can be very useful for military applications. These cross slot can be
arranged in such a manner that circular polarization becomes possible in RDRA.
Also by varying lengths of cross slots, left hand or right hand polarization can be
achieved. The circular ploarization makes the signal robust and help to reduce
electromagnetic polution.
Chapter 9
Sensitivity Analysis of Rectangular DRA
Abstract Sensitivity analysis of rectangular DRA depending on dielectric material,

and a, b, and d dimensions. These dimensions decides resonant frequency of
RDRA. The resonant modes are formed when realized with excitation. The resonant
frequency solution is worked with MATLAB and HFSS software. When these
dimensions are changed, resonant frequency of RDRA also changes. Variance
method has been tried out to evaluate error.
Keywords Isolated RDRA

Ground plane RDRA Resonant frequency

Sensitivity analysis Variance Error minimization
Rectangular DRAs of dielectric material having a, b, and d dimensional length have

been analyzed for frequency and resonant modes. RDRA is shown in Fig. 9.1.
These have been solved based on MATLAB and HFSS. Figure 9.2 presented
rectangular DRA with a, b, and d dimensions. Table 9.1 has shown RDRA
dimensions and their corresponding resonant frequencies. Figure 9.3 indicated
resonant modes with RDRA height. Plot of frequency versus length a variation is
shown in Figs. 9.4, 9.5 and 9.6. HFSS simulated modes in RDRA with S11
parameters are shown in Figs. 9.8 and 9.9 (Fig. 9.7).
da; db; dd
are (small change in length) random variables, and computed func-
tions are f dmnp and xmnp . The variance functions are ra ; rb ; rd . These are mainly
dependent on a, b, and d. Taylors expansion is restricted to second-order variable.
Hence, da; db; dd are mapped in terms of ra ; rb ; rd using diagonal matrix. Cmnp ;
Dmnp ; are amplitude coefcients which depend on the RDRA a, b, or d.
Frequency relationship can be determined based on a, b and d length variation
as given below:
d d dxmnpja; b; d
; ;
dd db da

200 9 Sensitivity Analysis of Rectangular DRA
Fig. 9.1 Isolated RDRA
b
a
Fig. 9.2 Rectangular DRA with a, b, and d dimensions
Table 9.1 RDRA dimensions

x (mm) y (mm) z (mm) r Material used
RDRA a=7 b=7 d = 10 10 Sapphire
a=6 b=6 d = 15
a=5 b=5 d = 30
Substrate 20 30 0.5 3.38 Arlon25N(tm)
Ground plane 20 30
Microstrip feed line 15 1.11
DRA dimensions (mm) Resonant frequencies (GHz) simulated
a b d
7 7 10 f1 = 13.46
6 6 15 f2 = 13.85
5 5 30 f3 = 14.21
9 Sensitivity Analysis of Rectangular DRA 201
Fig. 9.3 Resonant mode and RDRA height relationship
Fig. 9.4 Plot of frequency versus length a variation
Fig. 9.5 Plot of frequency versus length b variation

Fig. 9.6 Plot of frequency versus length d variation
Fig. 9.7 RDRA having three different heights for increasing resonant modes
Fig. 9.8 Modes pattern
Hz and Ez elds are expressed based on principle of orthogonality as:

X
Ez x; y; z; t Re ejxmnpt Cmnp umnp x; y; z 9:1a
mnp
X
Hz x; y; z; t Re ejxmnpt Cmnp vmnp x; y; z 9:1b
mnp
At z = 0, Ez eld
r
X 2
jxmnpt
Ez t; x; y; 0 Re e Cmnp umn xy 9:2
mnp
d
and
Hx ; Hy Jsy ; Jsx
Cmnp ; Dmnp are amplitude coefcients.

Name X Y XY Plot 1 HFSSDesign1

0.00 9.5741 -23.9108
m1 Curve Info
dB(S(1,1))
Setup1 : Sweep
-5.00
-10.00
dB(S(1,1))
-15.00
-20.00
m1
-25.00
7.50 8.75 10.00 11.25 12.50 13.75 15.00
Freq [GHz]
Name X Y
0.00 9.4238 -21.4417
XY Plot 1 HFSSDesign1
m1 Curve Info
m2 10.0251 -20.5547 dB(S(1,1))
m-2.50
3 14.5190 -17.9202 Setup1 : Sweep
-5.00
-7.50
-10.00
dB(S(1,1))
-12.50
-15.00
-17.50 m3
-20.00 m2
m1
-22.50
7.50 8.75 10.00 11.25 12.50 13.75 15.00
Freq [GHz]
Name X Y
m0.00
1 9.5741 -23.0345
Curve Info
m2 14.0080 -12.2318
dB(S(1,1))
m3 16.0170 -16.3397 Setup1 : Sweep
m4 17.5701 -15.7949
m-5.00
5 17.5952 -15.7757
-10.00
m2
dB(S(1,1))
-15.00 m3 m4
-20.00
m1
-25.00
7.50 10.00 12.50 15.00 17.50 20.00
Freq [GHz]
Fig. 9.9 Resonant frequency based on RDRA height

We have computed
p Z mpx npy
2 2
Cmnp p Jsx X; Y sin sin dxdy 9:3
abd a b
Now if
a ! a da a is increased to a da
b ! b db b is increased to b db
d ! d dd d is increased to d dd
We need to compute
Cmnpja da; b db; d dd
and similarly
xmnpja da; b db; d dd
This can be approximated by mean variance method and Taylors expansion

N normal distribution with mean zero
variance function
To compute error
2 3 0 2 2 31
da ra 0 0
4 db 5 N @0; 4 0 r2b 0 5A
dd 0 0 r2d
(By Taylors expansion)

dCmnp dCmnp dCmnp
Cmnpja da; b db; d dd Cmnpja;b;d da db dd
da db dd
2
1 d Cmnp 2 d2 Cmnp 2 d2 Cmnp 2
da db dd
2 da2 db2 dd 2

d2 Cmnp d2 Cmnp d2 Cmnp
2 dadb 2 dadd 2 dbdb
dadb dadd dbdd
9:4
Hence, variance or error can be written as

dCmnp 2 D E
hCmnpja da; b db; d dd C2mnpja; b; d i jdaj2
da
E 2 D
dC mnp 2 D E
jdbj2 dCmnp jdd j2
db dd
2 2
dCmnp 2
r2a r2 dCmnp r2 dC mnp
da b
db d
dd
9:5
Similarly, we can compute:
hxmnpjada;bdb;ddd xmnpja;b;d i
Error value
D E
xmnp dxmnp jdxj2
Error or variance:

dxmnp 2 2 dxmnp 2 2 dxmnp 2 2
r r r
da a db b dd d
r
m2 n2 p2
xp
a2 b2 d 2
The change in frequency based on RDRA change in dimension in x, y, and

z direction is given below:
m p 2
@x
q
a3
9:6a
@a p2
a2 b2 d 2
m 2 n2
n p 2
@x
q
b3
9:6b
@b p2
a2 b2 d 2
m2 n2
p2 p
@x
q
d3
9:6c
@d p2
a2 b2 d 2
m 2 n2
This gives the complete solution of RDRA sensitivity analysis. The higher-order
modes of a rectangular DRA were used to produce radiation patterns with enhanced
gain. The advantage of this approach is for enhancing gain. The maximum
achievable gain on mode m = 1, n = 7 to increase Directivity to 13.7 dBi.
Such DRA designed at 11 GHz with height 35 mm, this investigation focused on
rectangular DRAs, for excitation of the appropriate higher-order modes in RDRAs.
9.1 MATLAB Simulation
Matlab Program for Sensitivity Analysis
clear all;
clc;
close all;
c=3*10^8;
m=1;
n=1;
p=1;
E=10;
a=5*10^-3:.1*10^-3:30*10^-3;
b=10*10^-3;
d=15*10^-3;
for i=1:length(a)
f(i)=c/(2*pi)*sqrt(E)*sqrt((m*pi/a(i))^2+(n*pi/b)^2+(pp*pi/(2*d)^2));
end
a1=15*10^-3;
b1=10*10^-3:.1*10^-3:40*10^-3;
d1=20*10^-3;
9.1 MATLAB Simulation 207
for k=1:length(b1)
f1(k)=c/(2*pi)*sqrt(E)*sqrt((m*pi/a1)^2+(n*pi/b1(k))^2+(pp*pi/(2*d1)^2));
end
a2=10*10^-3;
b2=5*10^-3;
d2=10*10^-3:.1*10^-3:50*10^-3;
for t=1:length(d2)
f2(t)=c/(2*pi)*sqrt(E)*sqrt((m*pi/a2)^2+(n*pi/b2)^2+(pp*pi/(2*d2(t))^2));
end
subplot(3,1,1);plot(a,f);title('plot a vs f when a is varying');
subplot(3,1,2);plot(b1,f1);title('plot b vs f when b is varying');
subplot(3,1,3);plot(d2,f2);title('plot d vs f when d is varying');
9.2 HFSS Simulations
Now using HFSS software, we shall verify.

This gives the complete solution for RDRA sensitivity analysis.
It has been observed that resonant modes have been increasing based on increase
in dipole moment, i.e., modes are proportional to the height of RDRA.
9.2.1 HFSS Result
See below Figs. 9.10, 9.11, 9.12, 9.13 and Table 9.2.
Fig. 9.10 HFSS models of RDRA
Fig. 9.11 Return loss versus frequency with dimensions a = 5 mm, b = 5 mm, d = 30 mm shows
Return loss 18 dBi at f = 10.95 GHz
9.2 HFSS Simulations 209
Fig. 9.12 Return loss versus frequency with dimensions a = 6 mm, b = 6 mm, d = 15 mm shows
return loss 24 dBi at f = 10.95 GHz
Fig. 9.13 Gain = 5.5 at f = 11 GHz for DRA 1
Table 9.2 Dimensions table

x (mm) y (mm) z (mm) r Material used
RDRA a=7 b=7 d = 10 10 TMM10i
a=6 b=6 d = 15
a=5 b=5 d = 30
Substrate 20 30 0.5 3.38 Arlon25N(tm)
Ground plane 20 30
Microstrip feed line 19.2 1.1672
Lumped element 1.1672 0.5
9.3 Radiation Pattern
See below Figs. 9.14 and 9.15.

Chapter 10
Hybrid Modes in RDRA
Abstract In this chapter, new kind of resonant mode, i.e., hybrid mode in RDRA
(rectangular dielectric resonator antenna), is described using mathematical model-
ing. RDRA is excited by inserting RF feed probe or microstripline having nite
dimensions, carrying electric and magnetic currents at a given frequency. The
charge conservation equations then imply the presence of electric charge densities
and magnetic charge densities within the resonator. From Maxwells equations, we
derive vector Helmholtz equations for the electromagnetic elds. The vector
sources provide electric charge which gets converted into magnetic charge. In one
of the models, sidewalls of the resonator are perfect magnetic conductors, and top
and bottom surfaces are perfect electric conductors. Thus, the boundary conditions
on the elds are such that the tangential components of the magnetic eld vanish on
sidewalls and the normal components of the magnetic eld vanish at top and bottom
surfaces. The normal components of the electric eld vanish on sidewalls. Hz can
therefore be expanded as linear combinations of sin functions in xy direction along
with z-component of the source. For the Hz, Helmholtz equation can be expanded
in terms of sin functions (assuming that these sources vanish on boundary), with
z-dependent coefcients.
Keywords Hybrid modes

Mathematical model
Normal component

Tangential component Conservation equation Magnetic energy Electrical

energy Field diversity Fourier basis function
10.1 Introduction
RDRA (rectangular dielectric resonator antenna) is excited by inserting RF feed

probe or microstripline having nite dimensions, carrying electric and magnetic
currents at a given frequency. The charge conservation equations then imply
the presence of electric charge densities and magnetic charge densities within the
resonator at that particular frequency. To completely solve for the elds with the

212 10 Hybrid Modes in RDRA
resonator, we therefore set up four Maxwells equations taking into account mag-
netic and electric currents and charge densities. From these equations, we derive
vector Helmholtz equations for the electromagnetic elds with vector sources
determined from gradient and curl of the electric charge, magnetic charge, and
current densities. The sidewalls of the resonator are perfect magnetic conductors,
and top and bottom surfaces are perfect electric conductors. Thus, the boundary
conditions on the elds are such that the tangential components of the magnetic
eld vanish on sidewalls and the normal components of the magnetic eld vanish at
top and bottom surfaces. The normal components of the electric eld vanish on
sidewalls. Hz can therefore be expanded as linear combinations of sin functions in
xy direction along with z-component of the source. For the Hz, Helmholtz equation
can be expanded in terms of sin functions (assuming that these sources vanish on
boundary), with z-dependent coefcients.
On substitution of these expressions into the Helmholtz equation for Hz, source
then gives us a second-order linear differential equation for the coefcient functions
of z in Hz with a source term. This is solved, and the solution consists of a
superposition of a source (particular solution or inhomogeneous solution) term and
a homogeneous term (i.e., general solution of the homogeneous part). In hybrid
modes, total solution is developed, i.e., homogeneous and inhomogeneous. Two
constants in the homogeneous part are determined by applying the vanishing
boundary conditions on Hz at top and bottom surfaces, i.e., at z = 0, d. Likewise
applying boundary conditions on the vanishing of the normal component of the
E eld on the sidewalls, expressions are determined for Hx and Hy. Then, resonance
is seen, i.e., the electromagnetic eld inside the resonator is proportional to 1d, where
d is the frequency perturbations determined from Dirac delta functions. This
completely solves the problem of RDRA modes.
Hybrid modes can be generated by superposition of TE and TM modes inside
RDRA. In this case of RDRA, hybrid modes have been generated by using a probe
of nite dimension (d) is inserted into z direction and excitation of this probe
(d length) current is given to rectangular copper plane (x, y) as shown in Fig. 10.1
The current density can be determined based on KAM (KolmogorovArnold
Moser) time-averaging method and using dDirac delta function. The principle of
orthogonality is nally applied to determine Cmnp and Dmnp amplitude coefcients
of hybrid modes with elds in homogeneous and particular case. Here, particular
case will have inhomogeneous medium with source applied. Hz and Ez elds have
been computed simultaneously to generate hybrid modes and their coefcients.
Fig. 10.1 RDRA with copper x, y rectangular plane and feed d length
Fig. 10.2 Rectangular DRA
Figure 10.1 indicates RDRA for generating hybrid modes. Figure 10.2 shows
wall conguration of rectangular DRA required for hybrid modes. The hybrid modes
offer high efciency and polarization diversity. Frequency bandwidth can be con-
served by using polarization diversity. Maxwells equation is applied to nd solution
of RDRA. The eigen functions are obtained by solving Helmholtz equation. The
transverse components of Ex , Ey , Hx , and Hy have been expressed in terms of
longitudinal components Ez and Hz . The RDRA has been excited with a RF feed
probe of nite radius and small length inserted through ground plane into RDRA
along z-axis. Surface current density
on walls of resonator is produced due to
excitation given at feed point a2 ; b2 ; z of rectangular resonator. The azimuthal
component of magnetic elds inside the resonator is introduced, which is also
equivalent to z-component of surface current density. The modal longitudinal
coefcients are Ez and Hz . The radiation pattern or power distribution among these
different eigen modes is controlled by current distribution inside the resonator. The
inner product or reaction term of eigen function will be equal to corresponding eigen
mode. It is because magnetic currents are equal to electric currents in an antenna, due
to orthonormality principle or conservation of energy methods. Some of these power
coefcients can be made zero by canceling a particular resonant mode or blocking a
particular eigen function. This is possible for TE and TM modes. The same fre-
quency is introduced inside the guide with phase opposite to each other. Extracting a
particular resonant mode is also possible if surface current density of that mode is
made large enough by input excitation. When we apply input excitation frequency
matching to the desired mode, weighted magnitude of that particular mode coef-
cient becomes large and corresponding mode gets excited in the RDRA. The mode
merging can also be made possible by introducing shift in more than two modes
toward a common desired point. Equal weightage of TE and TM at same fre-
quency with opposite phase can cancel the mode. Higher-order modes can also be
generated in RDRA as shown in Fig. 10.1. Higher modes can provide higher gain
and high directivity to prevent EM pollution in microwave devices.
Figure 10.3 shows circular polarization of EM waves. The transverse components
Ex , Ey , Hx , Hy are the components determined in terms of longitudinal components
Ez and Hz . These transverse elds satisfy Helmholtz equations, are expressible in
terms of umnp ejxmnpt and vmnp ejxmnpt i.e., fourier basis functions. Ex , Ey , Hx ,
Hy elds are also expressible in terms partial derivatives of umnp, vmnp and
Fig. 10.3 Circular

polarization
hence if Cmnp, Dmnp denotes the linear combinational coefcients of umnp;

vmnp for Ez , Hz , then the same coefcients appear in Ex ; Ey ; Hx ; Hy . These
amplitude coefcients C mnp, Dmnp can be determined by matching Hx , Hy at
z = 0, to the surface current density of RDRA when feed is at z = 0. If the surface
excitation at z = 0 has frequency component other than xmnp, say x, then the eld
amplitude components corresponding to this excitation are determined from given
below terms. The source frequency, which is other than xmnp, introduces decay in
the resonator due to nite conductivity of the medium. Using the KAM theory of
averaging, the resonator extracts out only xmnp frequencies with amplitude.
Equivalently, if source contents are switched on for a nite duration and then
switched off, the only the dominant xmnp frequencies will be present in the
resonator. This situation is analogous to connecting a voltage source to an LC
oscillator for a nite duration and then switching it off. If, however, the source is not
switched off, then the other (non-dominant frequencies) will also be present and
these can be computed based on above-mentioned methods. The composite structure
having combination of PMC and PEC walls can generate hybrid modes (HEM).
The HEM can be further classied HE as odd hybrid modes and EH as even hybrid
modes. The applications for higher modes generation, mode shifting, mode merging,
and mode control can be made in antenna design. They can impart wide design space
in the eld of antenna. These designs can be used in beam control and regulation.
10.2 Mathematical Model
Maxwells equations:
For magnetic elds:
1 0
0 1
Hx r J x x; y; z; x
r2 k @ Hy A @ r Jy x; y; z; x A;
2
10:1
Hz r Jz x; y; z; x
10.2 Mathematical Model 215
Similarly, for electric elds:

J current density
wn function
a n, b n amplitude coefcients
umnp, vmnp Fourier basis function
hmn cut off frequency
0 1 2P P 3 2 3 2 3
Ex an wn bn un wn un
B C 6P P X X
0 0 7 6 07 6 07
@ Ey A 4 an wn bn un 5 an 4 wn 5 bn 4 u n 5
P P
Ez an w00n bn u00n w00n u00n
0 1 10:2
Ex
B C 1 @ 2 umnp
@ Ey A C mnp ;
hmn2 @z@x
Ez
0 P Cmnp @ 2 umnp P lDmnp @ 2 vmnp
1
hmn2 @x@z
hmn2
jxm; n; p @y@z
B mnp C
B C
E B P 1 C @ 2 umnp P l jxm; n; pD @ 2 vmnp Cejxmnpt ; 10:3
@ hmn2 mnp @y@z hmn2 mnp @x@z A
mnp
Cmnp umnp
E x; y; z; t electric field component

0 E 1
wmnpx x; y; z
X B E C jxmnpt
C mnpB C
@ wmnpy x; y; z A e
mnp
wEmnp x; y; z 10:4
0 Ez 1
umnpx x; y; z
X B E C jxmnpt
DmnpB C
@ umnpy x; y; z A e ;
mnp
umnpz x; y; z
E
with duality:
H x; y; z; t magnetic field component

0 H 1
wmnpx x; y; z
X B H C jxmnpt
CmnpB C
@ wmnpy x; y; z Ae
mnp
wHmnpz x; y; z 10:5
0 H 1
/mnpx x; y; z
X B H C jxmnpt
DmnpB C
@ /mnpy x; y; z Ae ;
mnp
/Hmnpz x; y; z
^
n H Js surface current density on walls;
Hsx x; y; 0 Jsy x; y
Hsy x; y; 0 Jsx x; y:
Hence,
Jsy x; y; dt current density into RDRA

X
jxmnpt
CmnpwH mnpx x; y; 0e
mnp 10:6
X
jxmnpt
DmnpuH
mnpx x; y; 0e ;
mnp
X
jxmnpt
Jsx x; y; dt CmnpwH
mnpy x; y; 0e
mnp
X 10:7
jxmnpt
DmnpuH
mnpy x; y; 0e :
mnp
If d is the length of probe inserted into RDRA, wH ; wE ; /H ; /E equations, we get

from Linear combinations of sine and cosine terms given below:
mnpx x; y; 0 / cos sin

wH a
mnpy x; y; 0 / sin cos

wH b
mnpx x; y; 0 / sin cos

uH c
mnpy x; y; 0 / cos sin

uH d
Amplitude
D coefcients can E be determined from principle of orthonormality:
C mnp wmnpy x; y; 0; wmnpy inner product or reaction terms can be written as
H H
follows:
D E D E
mnpx ; wmnpx Dmnp /mnpx ; wmnpx
C mnp /H H H H
1 T
Z
D E 10:8
lim Jsy x; y; t; wH
mnpx x; y; 0 ejxmnpt dt;
T!1 2T
T
and
D E D E
C mnp /H
mnpy ; wmnpy Dmnp /mnpy ; wmnpy
H H H
1 T
Z
10:9
jxmnpt
lim hJsx x; y; t; wH
mnpy x; y; 0i e dt:
T!1 2T
T
These are the solutions of amplitude coefcients Dmnp and Cmnp using
time-averaging KAM method (KolmogorovArnoldMoser).
These are the solutions of hybrid modes.
Helmholtz Equations
r2 c12 @t@2 wx; y; z; t 0; Helmholtz equation in time domain

2
r2 x2 w ^ x; y; z; x 0; Helmholtz equation in frequency domain
c
^ x; y; z; x X xY yZ z; separation of variables will be used
w
Solutions for RDRA on application of boundary conditions are as follows:
r2 k2 Ez 0; TM mode
r2 k2 Hz 0; TE mode
Hz 0; on all walls
Ex ; Ey 0; z 0; d;
Ex 0; x 0; a;
Ey 0; x 0; b;

1 @ l @
E ? 2 r? Ez 2 r? Hz ^z ; Hybrid mode electric fields; 10:10
h @z h @t

1 @ l @
H ? 2 ; r? Hz 2 r? Ez ^z ; Hybrid mode electric fields: 10:11
h @z h @t
Transverse components are as follows:
1 @ 2 Ez l @ 2 Hz
Ex ;
h2 @x@t h2 @y@t
1 @ 2 Ez l @ 2 Hz
Ey 2 2 ;
h @y@z h @x@t
Boundary conditions are as follows:
@ 2 Ez
0; x 0; a;
@x@z
@Ez
0; x 0; a;
@x
@Ez
0; y 0; b:
@y
There are three types of resonant modes at any known frequency xmnp:
TE; TM; HEM modes ! xmnpresonant mode:
@ 2 Ez
0; z 0; d;
@x@z
@Ez
0; z 0; d:
@z
For homogeneous medium without source terms:

X
Ez C mnpejxmnpt umnpx; y; z; t;
X
Hz Dmnpejxmnpt vmnpx; y; z; t:
10.3 Modes in Homogeneous Medium with Source Terms
For homogeneous medium case

X X
Ehom x; y; z; t C mnpejxmnpt wEmnp r Dmnpejxmnpt wEmnp r ;
10:12
where r x; y; z
X X
H hom x; y; z; t C mnpejxmnpt wH
mnp
r Dmnpejxmnpt wH
mnp
r;
10:13
Js x; y Jsx x; y^x Jsy x; y^y

J ! Js x; ydz
Js x; y; x
10.3 Modes in Homogeneous Medium with Source Terms 219
Maxwells equation:
r H J jxE
r E jxlH
E or H can be calculated
r2 H r J x2 lH
r2 k2 H r J with source;
H Hhom Hpart
J s x; y; z J sy x; ydz d Jy;z
r J Jz;y Jy;z
r2 k2 Hx Jsy x; yd0 z d
r2 k2 Hy Jsx x; yd0 z d
2
r k2 Hz Jsyx Jsxy d0 z d
Hence, we can compute
Hx ; Hy ; Hz x; y; z; x
vmnp sinx siny sinz
umnp cosx cosy cosz
Hz from E ? ; H ? equations:
X 10:14
Dsource m; n; p; xvmnp r Csource m; n; p; xumnp r ;
!
X xmnp2 x2
r k Hz
2 2
2 Czs m; n; p; xvmnpr : 10:15
mnp
c2 c
Hybrid modes can be generated by introducing non-resonant terms. Set innity

magnitude of coefcients for non-resonant frequency, and x and fxmnpg: x are
non-resonant terms.
10.4 Current Density in RDRA

X
Jsx x; y x2 xmnp2 C s mnp; xvmnpx; y; d ;
mnp
x is non-resonant frequency, and xmnp is resonant frequency

c2 Jsy;x x; y Jsx;y x; y
Hence, integrating and multiplying dz d
x2 xmnp2 Csource mnp; x

Z0 Z0
10:16
c2 Jsy;x x; y; x Jsx;y x; y; xvmnpx; y; d dx dy;
a b
Z
1
Czs mnpx Jsy;x x; y; x Jsx;y x; y; xvmnpx; y; d dx dy:
x2 xmnp2
10:17
Similarly, we can compute Cxs Cys terms which are the desired solutions of
hybrid resonant modes.
10.5 E and H Fields
E, H, elds:
r E jxlH
r H jxE J
r2 E jxljx
rHJ
E
Jx
2 R
hence, Czs mnp x2 xc2 mnp Jsy;x x; y; x Jsy;y x; y; xvmnp x; y; d dx dy is the
solution of hybrid modes. Similarly, other hybrid mode coefcients can be worked
out.
Cxs mnp dxs mnp

Cys mnp dys mnp
Czs mnp dzs mnp
Thus, complete solution of hybrid modes has been obtained.

10.6 Mathematical Modeling of Hybrid Modes 221
10.6 Mathematical Modeling of Hybrid Modes
n H Js on each wall
n E Ms on each wall
The mathematical derivation of Hybrid modes is given below. This is purely

based on solution of Maxwells equation.
First, we develop solution of rectangular waveguide and switch to resonator. The
waveguide solution is very simple. Figure 10.4 indicates eld conguration inside
RDRA. These waveguide equations will have both the elds Hz and Ez as given below:
r? c^z Ez^z E? jxlHz^z H? 10:18
r? c^z Hz^z H ? jxEz^z E ? 10:19
r? Ez ^z c^z E? jxlH? 10:20
r? Hz ^z c^z H ? jxE ? 10:21
r? Ez cE? jxl^z H? 10:22
jxl
r? Ez cE ? r? Hz ^z jxE ? 10:23
c
Fig. 10.4 Resonating modes in RDRA

2
jxl x l
r? E z r? Hz ^z c E?
c c
c2 x2 l h2
Waveguide equations based on Helmholtz equations are as follows:
c jxl
E? r? Ez 2 r? Hz ^z 10:24
h2 h
c jxl
H? 2
r? Hz 2 r? Ez ^z 10:25
h h
h i
Resonator equations are obtained by simply replacing c by ddz :
1 d l d
E? r? Ez 2 r? Hz ^z 10:26
h2 dz h dt
1 d l d
H? 2
r? Hz 2 r? Ez ^z 10:27
h dz h dt

2 Ez
r h2 0; Helmholtz equation
Hz
Boundary conditions in RDRA have been taken as, top and bottom walls of
resonator are PEC other four sides walls are PMC.
Hz 0; at x 0; a; and y 0; d; z 0; d;
Ex Ey 0; at z 0; d;
Hx 0; at y 0; b;
Hy 0; at x 0; a:
Standard elds are as follows:

X
Hz Re Dmnpejxmnpt vmnp r 10:28
X
Ez Re Cmnpejxmnpt umnp r 10:29
Standard orthogonal elds are as follows:

p mpx mpy mpz
2 2
vmnp p sin sin sin 10:30
abd a b d
10.6 Mathematical Modeling of Hybrid Modes 223
p mpx mpy mpz

2 2
umnp p cos cos cos 10:31
abd a b d
Equations (10.30) and (10.31) have been obtained from expansion of Helmholtz
equation by separation of variables method:

r2 h2 Hz 0

m2 2
Hence, h2 h2mn p2 a2 nb2 ; this gives the resonant frequency of RDRA.
Tensor product of linear combination can appear as given below:
n mpx npx npx mpx
Hz L cos cos ; cos sin ;
a a
mpx a a o
npx npx mpx
sin cos ; sin sin
a a a a
where L denotes linear combinations. It turns out that depending on the nature of
wall or surface (PEC or PMC), four possible linear combinations can appear
cos sin; sin cos; and sin sin; cos cos. Also,
x2 l c2 h2mn
Hence,
Hz 0; when x 0; cos terms are ruled out from x:

Hz 0; when y 0; against cos terms are ruled out from y:
mpx npy
Hz sin sin C1 ecmnz C2 ecmnz
a b
Hz 0; when z 0; d
C1 C2 0
ecmnd ecmnd 0; b
C1 C2 ; sincmnd 0
cmn jbmn
bmnd pp
Hence,
pp
bmn
d
mpx npy ppz
Hz sin sin sin 10:32
a b d
pp2
x2 l d h2mn ; hence, resonant frequency can be determined as follows:
s
p m 2 n2 p2
x l p ;
a2 b2 d 2
Here, we note that resonant frequency in hybrid mode is same for TE and TM
modes.
Now
1 @ @Hz l @ @Ez
Hx 2 ;
h2 @z @x h @t @y
Hx 0; at y 0; b;
dHx
0; at y 0; b;
dx
dEz
0; at y 0; b;
dy
Ex 0;
Ey 0; z 0; d;
Hence,
1 @ @ l @ @Hz
Ex Ez 2 ;
h2 @z @x h @t @y
@ 2 Ez
0 at z 0; d
@x@z
@Ez
0 at z d
@t

z-dependence of Ez is cos ppz
d ; Ex 0; when x 0; a;
mpx npy ppz
Ez cos cos cos 10:33
a b d
This is the way of getting Ez and Hz longitudinal components by the method of

separation of variables.
10.7 General Solution of Hybrid Modes (HEM) 225
10.7 General Solution of Hybrid Modes (HEM)
The investigations are based on rst applying waveguide theory, it models

regarding the electromagnetic elds to vary with z-axis, i.e., these are exploited into
the Maxwell curl equations, then manipulating them to express the transverse
components of the elds in terms of partial derivatives of the longitudinal com-
ponents of the elds w.r.t. x and y axis (i.e., the transverse coordinates). Waveguide
models of four different rectangular DRAs with specied boundary conditions lled
with homogeneous material having linear permittivity have been mathematically
developed and realized to determine TE and TM modes propagating elds. These
have resulted in different sinecosine combinations. Propagation of these elds
have been split as inside and outside of the RDRA with an interfacing surface
having two different permittivity. The solution is transcendental equation which
purely characterizes rectangular DRA resonant frequency and propagating elds.
The amplitude coefcient of these elds Cmnp and Dmnp inside the DRA can be
determined by comparing time-averaged magnetic energies equal to time-averaged
electrical energies by KAM method based on principle of orthonormality. The
transverse components Ex , Ey , Hx , Hy are the components determined in terms of
longitudinal components Ez , Hz . These transverse elds satisfy Helmholtz equa-
tions, are expressible in terms of umnp ejxmnpt and vmnp ejxmnpt Fourier basis
function. Ex , Ey , Hx , Hy elds are also expressible in terms partial derivatives of
umnp, vmnp, and hence if C mnp, Dmnp denotes the linear combinational
coefcients of umnp; vmnp for Ez , Hz , then the same coefcients appear in Ex ,
Ey , Hx , Hy . These coefcients Cmnp, Dmnp can be determined by matching Hx ,
Hy at z = 0 to the surface current density of RDRA, when feed is at z = 0. If the
surface excitation at z = 0 has frequency component other than xmnp, say x, then
the eld amplitude components corresponding to this excitation are determined.
Both the elds Ez and Hz will remain excited at any instant of time in resonator, and
then, these modes can be termed as hybrid modes. Our solution is developed based
on homogeneous medium in the resonator.
X n o
Ez Re Cmnpejxmnpt umnp r 10:34
X n o
Hz Re Dmnpejxmnpt vmnp r 10:35
X 1 n o@
jxmnpt
E? Re Cmnp e r? umnp r
h2mn @z
n o 10:36
l
2 Re jxmnpDmnpejxmnpt r? vmnp r ^z
hmn
E? ^x Ex ^y Ey ;
H? ^x Hx ^y Hy ;
@ @
r? ^x ^y ;
@x @y
0 1
Ex X X
E @ Ey A Cmnp ejxmnpt WEmnp r Dmnpejxmnpt /Emnp r 10:37
Ez
0 1
Ex Xh i
H @ Ey A C mnp ejxmnpt WH
mnp r Dmnpe
jxmnpt H
/mnp r 10:38
Ez
0 1
1 @2
2 @x@z umnp
B h1mn C
WEmnp r @ 2 @2 A
hmn @x@z umnp
umnp
where WEmnp is the electric eld vector coming from the z-components of electric
eld, i.e., TM mode.
And /Emnp is the electric eld vector coming from the z-components of magnetic
eld, i.e., TE mode.
Similarly, magnetic eld vectors /H
mnp
r and WH
mnp r can be solved.
0 1
1 @2
h2mn @x@z vmnp
B C
/H r B
@ 1 @2 C
A
mnp h2mn @x@z vmnp
vmnp
0 jlxmnp 1
@
h2mn @y vmnp
B C
/Emnp r B jlxmnp
@ 2 @ C
A
hmn @x vmnp
0
and,
0 jlxmnp
1
@
@y umnp
2
B hmn C
mnp r
WH @ jlxmnp A
@
@x umnp
h2mn
0
Hence, general hybrid equations can be written as follows:

Xh i
E CmnpWEmnp r Dmnp/Emnp r ejxmnpt 10:39
Xh i
H C mnp WH
mnp r D mnp /H
mnp
r ejxmnpt 10:40
Solution of the RDRA can be developed by using these above two equations.
For this, we insert a probe of d length having R radius into rectangular DRA. This is
pointing toward z-axis.
Figure 10.5 gives a clear picture of RDRA with feed associated, and Fig. 10.6
shows the structure of cylindrical probe.
a
x R cos /;
2
b
y R sin /;
2
z 0:
Fig. 10.5 RDRA with feed
Fig. 10.6 Feed probe

It is expressed based on Cartesian to cylindrical coordinates.

^
n H J s : This is based on boundary conditions inside the RDRA.

^ J
^ Hz^z H/ /
q s
^ J
H/^z Hz / s
and

a b
H/ R cos /; R sin /; z Jsz /; z 0\z\d; 0\/\2p
2 2

a b
Hz R cos /; R sin /; z Js //; z
2 2
X
a b
Js //; z; t Dmnpvmnp R cos /; R sin /; z ejxmnpt
2 2

a b
H/ R cos /; R sin /; z Hx sin / Hy cos /
2 2
Jsz /; z 0\z\d; 0\/\2p
Js/ /; z
X jlxmnp @vmnp a

b

Jsz /; z; t sin / cmnp R cos /; R sin /; z
h2mn @y 2 2

dmnp @ 2 vmnp a b
2 R cos /; R sin /; z ejxmnpt
hmn @z@x 2 2
X
jlxmnp @

a b

cos / Cmnp umnp R cos /; R sin /; z
h2mn @x 2 2

Dmnp @ vmnp a
2
b
2 R cos /; R sin /; z ejxmnpt ;
hmn @z@x 2 2
10:41
X
Js/ /; z; t Cmnp ejxmnpt Xmnp /; z;
mnp
where,

a b
Xmnp /; z vmnp R cos /; R sin /; z ; 10:42
2 2
X
1 2
Jsz /; z; t Cmnp gmnp /; z Dmnp gmnp /; zejxmnpt
mnp
where,

1 jxmnpt l sin / @ a b
gmnp /; z u mnp R cos /; R sin /; z
h2mn @y 2 2

2 sin / @ 2 vmnp a b e mnp /; z
gmnp /; z R cos /; R sin /; z ejxmnpt X
h2mn @x@z 2 2
Z
1 ~ mnp /; zejxmnpt dt d/ dz
Cmnp lim J s/ /; z; tX 10:43
T!1 2T
jtj\T
0\/\2p
0\z\d
Z

Xe mnp /; z2 d/ dz
Z 2 Z
1 1 2
Cmnp gmnp /; z d/ dz Dmnp gmnp /; z gmnp /; z d/ dz
10:44
Z
1 1
lim gmnp /; zJsz /; tejxmnpt d/ dz dt
T!1 2T
jtj T
Z Z
1 2 2 2
Cmnp gmnp /; z gmnp /; z d/ dz dt Dmnp jgmnp /; zj d/ dz
Z
1 2
lim Jsz /; z; t gmnp /; zejxmnpt d/ dz dt
T!1 2T
jtj\T
10:45
If we keep Js/ 0, from Eqs. (10.23) and (10.24), we get C(mnp) and D(mnp).
The study of electric and magnetic elds for maxima and minima inside RDRA
introduces us to dene mode number. By applying perturbations, higher modes can
be excited. The increase in the electrical length of the antenna on higher-order mode
causes higher antenna gain. Short and open boundaries are the basis of modes. The
half-wavelength resonant modes with odd numbers only will be excited when
ground plane is used as even modes get short-circuited due to ground plane. The
polarization of even and odd modes is opposite. The higher modes will have higher
resonant frequency. A number of higher modes also modify the radiation patterns,
i.e., the modes number will be equal to number of lobes in nal radiation pattern.
Highly directive patterns can be obtained at higher modes. Bandwidth of
higher-order modes will be decreased. HEM 1, 3, 5, 7 are odd modes that can be
written as HE. Similarly HEM 2, 4, 6, 8, are even modes or EH mode. The care
must be taken to select this hybrid number n because it has a direct relationship with
the radiation pattern of far elds or beam shape. Gain of antenna can also decrease
abruptly due to dispersion at higher modes. This is introduced when the dipole
moment starts overlapping. Based on various solutions, hybrid modes can be
memorized for any particular mode with desired radiation patterns. Automated
applications using microcontroller can generate lookup table for desired radiation
patterns or beam pattern for any desired frequency as well as gain. Thus, hybrid
modes can be used for automated RDRA recongurability.
10.8 HFSS Results
Figures 10.7, 10.8, 10.9, 10.10, 10.11, 10.12 and 10.13 are results of simulated and
experimentations of RDRA. Type of result is mentioned in each picture.
Fig. 10.7 The excitation is given by TE and TM modes at the same time
Fig. 10.8 The excitation is given by TE and TM modes at the same time
10.8 HFSS Results 231
Fig. 10.9 The HE elds in RDRA
HFSSDesign1
Name Theta Ang Mag Radiation Pattern 2
m1 360.0000 -0.0000 1.8149 Curve Info
0
m2 60.0000 60.0000 0.6570 m1 GainTotal
-30 30 Setup1 : Sweep
1.60 Freq='10.05511022GHz' Phi='0deg'
GainTotal
1.20 Setup1 : Sweep
-60 60 Freq='10.05511022GHz' Phi='90deg'
0.80
0.40 m2
-90 90
-120 120
-150 150
-180
Fig. 10.10 Orthogonal polarization due to two feeds
Name Theta Ang Mag

Radiation Pattern 8 HFSSDesign1
Curve Info
m1 360.0000 -0.0000 12.1194 0
dB(rELHCP)
20.00 Freq='15.34673367GHz' Phi='0deg'
m1 dB(rELHCP)
10.00 Setup1 : Sweep
-60 60 Freq='15.34673367GHz' Phi='90deg'
0.00
-10.00
-90 90
-120 120
-150 150
-180
Fig. 10.11 Radiation pattern

HFSSDesign1
Name X Y XY Plot 4 Curve Info
80.00
m1 70.0000 4.3036 dB(AxialRatioValue)
Setup1 : Sweep
70.00
dB(AxialRatioValue)
Setup1 : Sweep
dB (AxialRatioValue)
60.00 Freq='13.34170854GHz' Phi='90deg'
50.00
40.00
30.00
20.00
10.00 m1
0.00
0.00 125.00 250.00 375.00
Theta [deg]
Fig. 10.12 Plot of axial ratio indicating polarization
10.9 Prototype RDRA Results
See Fig. 10.13.
Fig. 10.13 Prototype double-feed RDRA S11 measurement

Chapter 11
Inhomogeneous Permittivity, Permeability,
and Conductivity Solution in Rectangular
DRA
Abstract Fields solution for inhomogeneous permittivity, permeability, and con-

ductivity in rectangular DRA was found. The solution is very complex. This is
based on the solution of rectangular waveguide lled with inhomogeneous per-
mittivity, permeability, and conductivity material. These amplitudes are called the
waveguide modes and are of the form cossin. Depending on the nature of wall
surfaces (PEC or PMC), four possible linear combinations can appear (cossin, sin
cos, sinsin, and coscos). The discrete modes enable us to visualize the resonator
as collection of L, C oscillators with different L, C values.

Keywords Inhomogeneous Permittivity Permeability Linear combinations

Sequential RLC circuits Discrete modes Complex solution
11.1 Introduction
Solution of rectangular waveguide with inhomogeneous permittivity, permeability,

and conductivity of the medium was found. These amplitudes are called the
waveguide modes and are of the form cossin and sincos which denotes linear
components. It turns out that depending on the nature of wall surfaces (PEC or PMC),
four possible linear combinations can appear (cossin, sincos, sinsin, and coscos).
In a rectangular DRA, weve got to applying in additional boundary conditions on top
and bottom surfaces to be the linear combinations as compared to waveguide. They
have two possible linear combinations of sin cos. Thus, the possible frequencies
obtained by solving them comes out to be an equivalent, but computationally simpler
way to pass on from waveguide physics to resonator physics is to just replace byin all
the waveguide formulae that express the tangential eld components in terms of the
longitudinal components. This is done after solving the full 3-D Helmholtz equations
using separation of variable in x, y, z. The discrete modes enable us to visualize the
resonator as collection of L, C oscillators with different L, C values. The outcome of all
these analyses enables us to write down the elds inside the resonator, as superposition
of four or three vector-valued basis functions.

234 11 Inhomogeneous Permittivity, Permeability, and Conductivity
11.2 Mathematical Model
x; y 0 1 d ve x; y 11:1a
lx; y l0 1 d vm x; y 11:1b
Er Ex; yexpcz
H r H x; yexpcz 11:1c
Curl Er jxlx; yH r
Curl H r r jx1 x; yE r

jrx; y 11:1d
jx1 x; y 1 E r jxx; yE r
x1 x; y
We dene

jrx; y
x; y 1 x; y 1 0 1 d ve x; y
x1 x; y
Note that ve and vm are the frequency dependent complex functions of x; x; y

and by substituting into the Maxwell equations, it gives the following equations:
Ez;y c Ey jxlHx ; 11:3a
cEx Ez;x jxlHy ; 11:3b
Ey;x Ex;y jxlHz ; 11:3c
Hz;y cHy jxEx ; 11:4a
cHx Hz;x jxEy ; 11:4b
Hy;x Hx;y jxEz 11:4c
where Ex ; Ey ; Ez ; Hx ; Hy ; Hz are the functions (complex) of X, Y, and frequency x

only. We can arrange the equations [i.e., (11.3a), (11.3b), (11.4a) and (11.4b)] as
follows:

c jxl Ey Ez;y
;
jx c Hx Hz;x

jx c Ex Hz;y
:
c jxl Hy Ez;x
Thus,

Ey 1 c jxl Ez;y
2 ;
Hx h jx c Hz;x

Ex 1 jxl c Ez;y
;
Hy h2 c jx Hz;x
h2 h2 x; y c2 x2 l
c2 x2 0 l0 1 dve 1 dvm
h20 k2 dve vm k2 d2 ve vm
where
h20 c2 x2 0 l0 c2 k2 ; k 2 x2 0 l0
Then,
1
Ex cEz;x jxlHz;y ;
h2
1
Ey 2 cEz;y jxlHz;x ;
h
1
Hx 2 cHz;x jxEz;y ;
h
1
Hy 2 cHz;y jxEz;x
h
Equations (11.3c) and (11.4c) then give the following equation:

Ez;y lHz;x Ez;x lHz;y
c 2 ;X jx ;X c 2 ;Y jx ;Y jxlHz 0 11:5a
h h2 h h2

Hz;y Ez;x Hz;x Ez;y
c ;X jx ;X c ;Y jx ;Y jxEz 0 11:5b
h2 h2 h2 h2
These can be expended as:
jxl n l l o
DH z jxlH z jx ; H
2 X z;x
; H
Y z;y
h2 h h2
11:6a
1 1
c ;Y Ez;x 2 ;X Ez;y 0
h2 h
and,
jx n o
2
DEz jxEz jx 2
;X Ez;x 2 ;Y Ez;y
h h h 11:6b
1 1
c ;X Hz;y 2 ;Y Hz;x 0
h2 h
or equivalently,
h2 n o
D h2 Ez ; X Ez;x ; Y E z;y
h2 h2
11:7a
ch2 1 1
; X H z;y ; Y H z;x 0
jx h2 h2
h2 n l l o
D h2 H z ;X H z;x ;Y Hz;y
l h2 h2
11:7b
ch2 1 1
2
;Y Ez;x 2 ;X Ez;y 0
jxl h h
or in matrix notation with
ve vm v1 x; y; ve vm v2 x; y 11:7c
this can be expended as:

Ez v1 0 Ez v2 0 Ez
D h2 k2 d k 2 d2
Hz 0 v1 Hz 0 v2 Hz

v2 0 Ez @
k 2 d2 log 2 ; x
0 v2 Hz h @x
@ c @ c @ c @
log 2 ;y ; log h2 ; x log h2 ;y log h2 ;y
h @y jx @y jx @x jxl @x
@ l @ l @ E
z
log h2 ;x ; log 2 ;x log 2 ;y 0
@y h @x h @y Hz
Note that h2 x; y h20 k2 dv1 x; y k 2 d2 v2 x; y:

p
We write c h20 h2 in (11.8) and then (11.8) can be expanded as:

Ez v 0 Ez v2 0 Ez
D h2 k2 d 1 k 2 d2
Hz 0 v1 Hz 0 v2 Hz

@ @ @ @
f11 x; y; k; d g11 x; y; k; d ; f12 x; y; k; d g12 x; y; k; d
@x @y @x @y

@ @ @ @ Ez
f21 x; y; k; d g21 x; y; k; d ; f22 x; y; k; d g22 x; y; k; d 0
@x @y @x @y Hz
where k h20 ,

f11 x; y; k; d log ; x g 11 x; y; k; d log ;y
h2 h2
c
f12 x; y; k; d log h2 ;y
jx
c
g12 x; y; k; d log h2 ;x
jx
c
f21 x; y; k; d log h2 ;y
jxl
c
g21 x; y; k; d log h2 ;x
jxl
l
f22 x; y; k; d log 2 ;x
hl
g22 x; y; k; d log 2 ;y
h
We therefore dene the matrix differential part as follows:

!
f11 @@x g11 @@y f12 @@x g12 @@y
lk; d
f21 @@x g21 @@y f22 @@x g22 @@y
where fab ; Iab are the functions of x, y, k; d: It is early to see that for small d; fkb and
gkb can be expanded in power of d with the series state from d0 .
In other words,
fab x; y; k; 0 0;
gab x; y; k; 0 0:
Writing therefore
fab x; y; k; d dfab1 x; y; k d2 fab2 x; y; k Od3
and likewise,
gab x; y; k; d dgab1 x; y; k d2 gab2 x; y; k Od3

We have
@ @
lk; d F x; y; k; d Gx; y; k; d
@x @y

@ @
d F1 x; y; k G1 x; y; k
@x @y

@ @
d2 F2 x; y; k G2 x; y; k Od3
@x @y
where

F x; y; k; d fab x; y; k; d j1 a;b 2j

Gx; y; k; d gab x; y; k; d j1 a;b 2j

F1 fab1 ; G1 fab1 ;

F2 fab2 ; G2 fab2 ;
Note that
F dF1 d2 F2 Od3
G dG1 d2 G2 Od3
Thus,
lk; d dl1 k d2 l2 k Od3
where
@ @
l1 k F1 x; y; k G1 x; y; k
@x @y
@ @
l2 k F2 x; y; k G2 x; y; k
@x @y
We dene
v1 k l1 k k 2 v1 x; yI2 ;
v2 k l2 k k 2 v2 x; yI2 ;
then

D kwx; y dv1 k d2 v2 k wx; y 0

With neglecting higher order terms of O d3 ; where

Ez x; y
wx; y
Hz x; y
The boundary conditions are Ez 0; where x = 0, a or y = 0, b, Hx = 0; where x = 0, a,

Hy 0; where y = 0, b. Using the expression for Hx Hy in terms of Ez ; Hz and the
boundary conditions on Ez , it follows that the boundary conditions on H can be
replaced by Hz;x 0, where x = 0, a and Hz;y 0, where y = 0, b.
Where d 0; (the homogeneous case), w satises D kw 0:
We denote the solution to this by wmn 0
x; y since application of the boundary
conditions leads to (after separation of variables)

Am; numn x; y
wx; y
Bm; nvmn x; y
where Am; n; Bm; n are the complex constants and
2 mpx npy
umn x; y p sin sin
ab a b
2 mpx npy
vmn x; y p cos cos
ab a b
Note that,
Za Zb
humn ; ukl i umn x; yukl x; ydxdy
0 0
dmk dnl
humn ; ukl i dmk dnl
humn ; ukl i 0
The general solution to the d 0 case (homogeneous medium) is
D kw0 0; k h20
With boundary condition, the equation can be written as:

0 Am; numn x; y
w0 x; y wx;y x; y
Bm; nvmn x; y

mp 2 np 2
k kmn0
h2 m; n
a b
where m, n = 1, 2, . The corresponding z-components of the electromagnetic eld

are given as follows:
1
X
Ez t; x; y; z Am; numn x; y
expcmn z
Hz t; x; y; z m;n1 Bm; nvmn x; y
X 1 h i
wm;n
0
x; y exp cm;n z
m;n1
The active z-component of electromagnetic elds in time at frequency x are

given as follows:
( )
Ez t; x; y; z X
0
Re wmn x; y expfjxt cmn zg
Hz t; x; y; z
mn
the x and y components of the electromagnetic elds are easily obtained for d 0:
Fields are easily obtained for d 0:
X 1
Ex x; y; z cmn Am; num;n ; xx; y
m;n h0 m; n
2

jxl0 Bm; nvm;n Xx; y expcm;n z ;
X 1
Ey x; y; z cmn Am; num;n ; xx; y
m;n h20 m; n

jxl0 Bm; nvm;n Yx; y expcm;n z ;
X 1
Hx x; y; z jx0 Am; num;n ; xx; y
m;n h20 m; n

cmn Bm; nvm;n Xx; y expcm;n z ;
X 1
Hy x; y; z jx0 Am; num;n ; xx; y
m;n h20 m; n

cmn Bm; nvm;n Yx; y expcm;n z ;
where
q
cmn h20 m; n k2
q
0
cmn kmn k2
Consider now d [ 0 and d is small. Then, we have to solve the following equation:
D kw dV1 k d2 V2 k1 w 0
Let

k k0 dk1 d2 k2 O d3 ;

w w0 dk1 d2 k2 O d3
Then, we get on equality coefcient of d0 ; d; d2 separately

O d0 D k0 w0 0;

O d1 D k0 w1 V1 k0 w0 k1 w0

O d2 D k0 w2 V1 k0 w1

V1 k0 w0 k1 V10 k0 w0 k1 w1 k2 w0 0

The solution to O d0 case has already been obtained:
For each m and n, we have two orthogonal solutions:

umn x; y 0
w01 and w02 :
mn 0 mn vmn x; y
These solutions are normalized as follows:

Z
0K 2
wmn dxdy 1; k 1; 2 where knk2 nT n;
n 2 D2
0\x\a; 0\y\b;
Both solutions have the same eigenvalues

mp 2 np 2
wmn
0

a b
We can set k0 wmn

0
in az and also

umn 0
wmn
0
A B
0 vmn
Awmn
01
Bwmn
02
To get

0
D kmn w1 k1 wmn
0 0
V1 kmn 0
wmn 0
Taking the inner product with both w0k

mn ; k 1; 2 gives the following equation:
D E
k1 A wmn
01
; V1 kmn
0
wmn
0
0;
D E
k1 B wmn
02
; V1 kmn
0
wmn
0
0:
This expands as:

0D ED E1
wmn
01
; V1 kmn
0
wmn
01
wmn
01
; V1 kmn
0
wmn
02
@D ED EA
wmn
02
; V1 kmn
0
wmn
01
wmn
02
; V1 kmn
0
wmn
02

A 1 A
x k
B B
Thus, the given unperturbed of eigenvalues kmn0

split into two eigenvalues
0 1
kmn dk .
Where k1 can be any one of the two eigenvalues of the secular matrix (2 2)
D E
wmn
0a
; V1 kmn
0
wmn
0b
1 a;b 2

Aa m; n
Let ; a 1; 2 be the corresponding normalized eigenvalues:
Ba m; n
jAa m; nj2 jBa m; nj2 1.
We denote the corresponding eigenvalues by kmn 1
a; a 1; 2.
The principal normalized eigenfunction of D corresponding to the eigenvalue
kmn
0
is split into kmn
0
dkmn
1
a; a 1; 2; and

Aa m; numn x; y
kmn
0a
; a 1; 2;
Ba m; nvmn x; y
respectively.
The radiation of an antenna that feeds a waveguide is controlled through a robot

(dynamic or moving probe).
The antenna orientation at time t is specied by the vector
^lt R/t; ht; wt^z DRt^z
where
R/; h; w Rz /; Rx h; Rz w
i.e., /; h; w are the Euler angles. The antenna is a rigid body (like a top) that
carries a current density. Jt; r in its initial conguration. So, after some time the
volume current density within the antenna body is given by

Jb t; r J t; Rt1 r ; r RtB
where B is the antenna body space at the time t = 0, we wish to control the
orientation angles /t; ht; wt, 0 t T so that the radiation pattern of the
antenna is as close as possible to a given pattern. Let us say that the pattern is
specied by the vector potential Ad t; r in space.
Then, the vector potential produced by the rotating antenna is as follows:

Z Jb t jrr0 j ; r 0
l c
Ad t; r d 3 r0
4p jr r 0 j
R3

Z Jb t jrr0 j ; R1 t jrr0 j r 0
l c c
d 3 r0
4p jr r 0 j
RtB

Z J t jrRtnj ; R1 t jrRtnj r 0 Rtn
l c c
d3 n
4p jr Rtnj
B
The far-eld vector potential pattern, i.e., jr j jnj is given by

Z
l r ^r ; Rtn 1 r
At; r J t ; R tn d 3 n
4pr c c c
B
We wish to track Ad t; r over the space-time region 0; T n where n 2 R3 is

in a region of space in the far-eld zone.
We assume that r is relatively constant in n (it is the distance of the center of n

from the origin).
Then,
Z
jAt; r Ad t; r j2 dt sin h0 dh0 d/0
0;TX0
is to be minimized where X0 is the solid angle subtended at the origin by n shifting

l
r
our time or region by c; the quantity to be minimized is taking 1
4p
2
Z Z

rAd t; r Jt R1 t^r ; n ; R1 tnd 3 n sin h0 dh0 d/0
c

0;T X0 B
where
r; h0 ; /0 r; h0 ; /0 ^r
Ad t; r Ad t; r; h; / r
constant
Note that R1 t Rz wtRx htRz ht.

Once the optimal trajectory f/d t; hd twd tg has been determined by opti-
mizing this highly nonlinear functional, we decide how to apply machine torques
fsd t; sd tsd tg; 0 t T.
The rigid body carrying current so that the time fsd t; sd tsd t; 0 t T g is
as close the desired trajectory as possible.
The kinetic energy of the top is given in terms of its principal moments of inertia
by (golds term classical mechanism)
1
T I1 x21 I2 x22 I3 x23
2
where
x1 h_ cos w h_ sin w sin h;

x2 h_ sin w h_ cos w sin h;
x3 w_ h_ cos h:
For a symmetric top (body carrying current), I1 I2 :

We get,
1 1 2
T I1 h_ 2 /_ 2 sin2 h I3 w_ /_ cos h
2 2
Then, the Lagrangian after taking machine torque into account is given as:
L T V s/ th_ sh th_ sw tw_
where V mgl cos h; l being the distance between the CM of the top and the origin.
The equation of motion is given as:
d @L @L d @L @L d @L @L
; ; give
_
dt @ / @/ dt @ h_ @h dt @ w_ @w
d _
I1 / sin2 h I3 cos h w_ h_ cos h sw t;
dt
d _
I1 h I3 /_ sin hw_ /_ cos h I1 /_ 2 sin h cos h mgl sin h sh t;
dt
d _
I3 w h_ cos h sw t
dt
We dene
Zt
F / t s/ sds;
0
Zt
F w t sw sds;
0
and then, the equations of motion are as follows:

I1 /_ 2 sin2 h I3 cos h w_ h_ cos h F/ t 11:8

I3 w_ h_ cos h Fw t 11:9

I1
h I3 / sin h w_ h_ cos h I1 /_ 2 sin h cos h mgl sin h s/ t 11:10
The total work done by the machine torque is given as:

ZT
W s/ t/_ t sh th_ t sw tw_ t dt
0
Z
F_ / t/_ t sh th_ t F_ w tw_ t dt
This must be minimized subject to the constraints robotic of the equation of motion
nal orientation dened
(11.8)(11.10) and that the by /t; ht; wt is given.
Here, we calculate F/ t; Fw t; Fh t ; 0 t T and hence, fsd t; sd t;
sd t; 0 t Tg by putting
/t /d t
h t hd t
wt wd t
This gives us an algorithm for calculating the machine torque to be supplied over
the range [0, T], so that the top antenna follows a desired trajectory that will lead to
a radiation pattern that matches a given radiation pattern as closely as possible in a
given solid angle element X0 over a given time interval [0, T]. Another way to
design the machine torque is to minimize a weighted combination of the error
energy between the desired trajectories f/d ; hd ; wd g and the actual trajec-
tory, and the total work done by the torques over the duration [0, T] is minimized as
follows:
ZT ZT 2
Sa F_ / t/_ t sh th_ t F_ w tw_ t dt b /t /d t2 ht hd t
0 0

2
wt wd t dt
where F/ t; sh tFw t are given in terms of /; h; w in above equation we can

write the equation as:
ZT
_ /;
S n /; /; h; h;
_ h; w; w;
_ w dt
0
a; b are the weights a; b [ 0. The optimal trajectories /; h; w then satisfy the

Euler trajectories equations:
d @n d2 @n @n
;
dt @ /_ dt2 @ /
@/
d @n d2 @n @n
;
dt @ h_ dt2 @ h @h
d @n d2 @n @n
:
dt @ w_ dt2 @ w
@w
Now support that the electromagnetic eld generated by an antenna falls on the
aperture z = 0, 0 < x < a, 0 < y < b, of a rectangular waveguide.
We wish to complete the elds inside the guide. Assuming the guide to have
constant permittivity and permeability, we get for the phase elds inside the guide
at a given frequency x,

Ez x; y
D h20 0
Hz x; y
h20 c2 x2 0 l0 c2 k2
We have
1
Ex 2
cEz;x jxlHz;y
h0
1
Ey 2 cEz;y jxlHz;x
h0
1
Hx 2 cHz;x jxEz;y
h0
1
Hy 2 cHz;y jxEz;x
h0
The general solution for the elds within the given satisfying boundary condi-
tions is (as weve seen earlier) given by
1
X
Ez x; y; z Am; numn x; y
ecmn z
Hz x; y; z Bm; nvmn x; y
m;n1
2 mpx npy
umn x; y p sin sin ;
ab a b
2 mpx npy
vmn x; y p cos cos
ab a b
q
cmn cmn x h20 m; n x2 l0 0

m2 2
where h20 m; n p2 a2 nb2 :
When anelectromagnetic
eld is incident on the surface z = 0 of the guide, let
E0x x; y; E0y x; y ; be the incident electric eld (tangential components to the
surface). Then by continuity of the tangential components of the electric eld, we have
Ex x; y; 0 E0x x; y; Ey x; y; 0 E0y x; y;
0 x a; 0 y b;
where
X 1
Ex x; y; 0 cmn Am; numn x; y jxl0 Bm; nvmn x; y
m;n h20 m; n
and
X 1
Ey x; y; 0 cmn Am; numn x; y jxl0 Bm; nvmn x; y
m;n h20 m; n
Thus
X c 2 mp mpx npy
E0x x; y mn
A m; n p cos sin
m;n h20 m; n ab a a b

jxl 2 np mpx npy
2 0 p Bm; n cos sin
h0 m; n ab b a b
np mpx npy
2 X cmn
E0y x; y p A m; n sin cos
ab m;n h20 m; n b a b

jxl mp mpx npy
2 0 Bm; n sin cos
h0 m; n a a b
or equivalently
!
X 2cmn mp 2jxl0 np mpx npy
E0x x; y p A m; n p B m; n cos sin
m;n h20 m; n ab h20 m; nb ab a b
and
!
X 2cmn np 2jxl0 mp mpx npy
E0y x; y p Am; n 2 p Bm; n sin cos
m;n h0 m; nb ab
2 ah0 m; n ab a b
Thus, we get
Z mpx npy
2
p E0x x; y cos sin dxdy
ab a b
0\x\a; 0\y\b

mpcmn jxl0 np
A m; n B m; n
ah20 m; n h20 m; nb
These are the two simultaneous linear equations for the two variables
Am; n and Bm; n which are easily solved. 2 2 matrix notation supports the
incident electric eld is E0 x; y; z; t: Then, we take
Z
^ 0x x; y; x
E E0x x; y; 0; tejxt dt
R
Z
^ 0y x; y; x
E E0y x; y; 0; tejxt dt
R
Dene
Z mpx npy
2
/x m; n; t p E0x x; y; 0; t cos sin dxdy
ab a b
D
Z mpx npy
2
/y m; n; t p E0y x; y; 0; t sin cos dxdy
ab a b
D
Then,
Z
^ m; n; x ,
/ /x m; n; tejxt dt
x
R
Z
2 ^ 0x x; y; x cos mpx sin npy dxdy
, p E
ab a b
D
Z
^ m; n; x ,
/ /y m; n; tejxt dt
y
R
Z
2 ^ 0y x; y; x sin mpx cos npy dxdy
, p E
ab a b
D
11.3 Applications: Hybrid Modes Generation Inside

RDRA Can Be Used for Polarization Diversity
11.3.1 RF Measurements for Antenna Parameters
S11
Gain
Radiation pattern
VSWR
Dispersion
Polarization
Permittivity and permeability
Bandwidth
S21
Isolation
Efciency
Directivity
Resonant frequency
Propagation constants
Axial ratio
Resonant mode
Dominant mode
Higher-order modes
Chapter 12
Case Studies
Abstract This chapter deals with case studies, where implemented cases have been
discussed. Various type of antennas have been fabricated. Their dominating para-
meters are shown. These cases have either been developed using HFSS simulations
or hardcore experimentations. The case study is based on the rectangular DRA
using ceramics such as eccostock-500. Nomenclature and parameters obtained have
been mentioned below the each gure. Emphasis is also given on geometry of
antennas and their experimental results. The experimental results have been
obtained under specic environmental conditions i.e. anechoic chambers.

Keywords Rectangular DRA Simulated and experimental results Isolated RDRA

and RDRA with ground plane Single feed Double feed Anechoic chamber

measurements Radiation pattern Gain S11, VSWR, Z11, and E and H elds

distribution RF absorbers Test set for measurement Prototype Azimuth and

elevation pattern Manganesemanganese material Bandwidth enhancement

Higher order resonant modes Variable DRA height Smith chart Group

delay Rectangular wells LHCP RHCP Circular polarization Phase distortion

S21 Ferrite RDRA Slot variation Permittivity variation effects Hardware

implementation VNA calibration Aperture coupled RDRA and probe fed RDRA
12.1 Structure and Hardware Experimentations
The case study is based on the rectangular DRA where the various designs of DRA
have been presented and the nomenclature is given below each gure. These
antennas have been simulated and fabricated. The results obtained have been pre-
sented graphically here. These antennas also have been placed inside anechoic
chamber to minimize external effects that come during measurements. The gures
indicate various measurement steps involved in this process. For simulated results,
Ansoft HFSS 13.0 has been used. These graphs of radiation pattern, gain, S11,
VSWR, Z11, and eld distribution have been presented and their domains are
mentioned below each gure (Figs. 12.1, 12.2, 12.3, 12.4, 12.5, 12.6 and 12.7).

252 12 Case Studies
Fig. 12.1 a Model of electronic band gap (EBG) structure cavity rectangular dielectric resonator.
b Diagrammatic representation of RDRA with top-loading DRA
Fig. 12.2 Positioning of RDRA antenna ready for test procedure setup
Fig. 12.3 Positioning of RDRA antenna under measurement setup

12.1 Structure and Hardware Experimentations 253
Fig. 12.4 RDRA antenna between RF absorbers inside chamber for gain testing
Fig. 12.5 View of RDRA antenna under test setup on sliding table
Fig. 12.6 View of RDRA antenna under test setup on sliding table
Fig. 12.7 RDRA antenna on sliding table with variation in position of RDRA
12.1.1 RDRA Antenna Results
See Figs. 12.8, 12.9, 12.10, 12.11, 12.12, 12.13, 12.14, 12.15, 12.16, 12.17, 12.18,
12.19, 12.20, 12.21, 12.22 and 12.23.
254 12 Case Studies
Fig. 12.8 E-plane radiation pattern at 13.7 GHz of simulated RDRA for electric eld distribution
Fig. 12.9 H-plane radiation pattern of simulated RDRA for magnetic eld distribution
12.1 Structure and Hardware Experimentations 255
Fig. 12.10 VSWR measurement for measurement of reflected eld strength of simulated RDRA
Fig. 12.11 H-plane radiation pattern at 16.8 GHz of simulated RDRA for magnetic eld distribution
256 12 Case Studies
Fig. 12.12 Measurement of gain at 16.8 GHz of simulated RDRA
12.2 RDRA with ManganeseManganese Material

as Dielectric
In this case, the designing of RDRA using manganesemanganese material has

been presented. This dielectric material shows the various effects on the parameters
of the developed antenna. The bandwidth enhancement techniques have been
implemented using two wells. The results obtained have been presented graphically
here. These antennas also have been placed inside anechoic chamber to minimize
external effects that come during measurements. The gures indicate various
measurement steps involved in this process. These antennas have been simulated
and fabricated. For simulated results, Ansoft HFSS 13.0 has been used. These
graphs of radiation pattern, gain, S11, VSWR, Z11, and eld distribution have been
12.2 RDRA with ManganeseManganese Material as Dielectric 257
Fig. 12.13 Radiation pattern at 13.7 GHz for radiated eld pattern of simulated RDRA
258 12 Case Studies
Fig. 12.14 Return loss S11 of higher-order modes excited inside simulated RDRA
Fig. 12.15 Gain of simulated RDRA at various heights (with the excitation of higher-order modes)
Fig. 12.16 Excitation of higher mode at RDRA 10 mm height with variable frequency of 10, 12,
15 GHz
Fig. 12.17 RDRA height 15 mm, excitation frequency variable, generated higher modes. Even
modes excitation with top excitation a TE114 at 11.7 GHz b TE116 at 13.7 GHz c TE118 at
16.7 GHz in RDRA
Fig. 12.18 Measured return loss S11 at 33.596 dB of RDRA

260 12 Case Studies
Fig. 12.19 Measured input impedance of 50.089 by Smith chart of RDRA
Fig. 12.20 Prototype RDRA under test setup for measurements with VNA
Fig. 12.21 Experimental gain of antenna of 16.624 dBi at 16.8 GHz RDRA inside microwave
anechoic chamber
Fig. 12.22 E-plane radiation pattern at 16.8 GHz dBi inside microwave anechoic chamber
amplitude versus theta
262 12 Case Studies
Fig. 12.23 H-plane radiation pattern at 16.8 GHz dBi inside microwave anechoic chamber
amplitude versus theta
presented and their domains are mentioned below each gure. The phase versus
frequency plots indicate distortions in the developed RDRA. The group delay,
forward power has also been indicated in simulated results (Figs. 12.24, 12.25,
12.26, 12.27, 12.28, 12.29, 12.30, 12.31, 12.32, 12.33, 12.34, 12.35, 12.36, 12.37
and 12.38).
Fig. 12.24 HFSS model of dual-feed RDRA with circular polarization

Fig. 12.25 HFSS model of DRA with two rectangular wells
Name X Y XY Plot 1 HFSSDesign1 ANSOFT
-4.00 15.3657 -17.2753

m1 Curve Info
dB(S(1,1))
Setup1 : Sweep
-6.00
-8.00
dB(S(1,1))
-10.00
-12.00
-14.00
-16.00
m1
-18.00
10.00 11.00 12.00 13.00 14.00 15.00 16.00 17.00
Freq [GHz]
Fig. 12.26 Return loss curve S11 of simulated DRA at 15.5 GHz
Radiation Pattern 2 HFSSDesign1 ANSOFT
0 Curve Info
rETotal
14.00 Freq='15.34673367GHz' Phi='0deg'
rETotal
Setup1 : Sweep
10.50
Freq='15.34673367GHz' Phi='90deg'
-60 60
7.00
3.50
-90 90
-120 120
-150 150
-180
Fig. 12.27 Radiation pattern 2 of simulated DRA at 15.5 GHz

264 12 Case Studies
XY Plot 2 HFSSDesign1 ANSOFT
50.00
Curve Info
dB20(Z(1,1))
Setup1 : Sweep
45.00
40.00
dB20(Z(1,1))
35.00
30.00
25.00
20.00
10.00 11.00 12.00 13.00 14.00 15.00 16.00 17.00
Freq [GHz]
Fig. 12.28 Impedance plot Z11 of simulated DRA at 15.5 GHz
Name Theta Ang Mag Radiation Pattern 1 HFSSDesign1 ANSOFT
m1 360.0000 -0.0000 0.7290 0 Curve Info

m2 56.0000 56.0000 5.0286 GainTotal
4.80 Freq='15.34673367GHz' Phi='0deg'
GainTotal
Setup1 : Sweep
3.60
Freq='15.34673367GHz' Phi='90deg'
-60 m2 60
2.40
1.20
m1
-90 90
-120 120
-150 150
-180
Fig. 12.29 Gain plot of 4.80 dBi of simulated DRA at 15.5 GHz
Ground plane 20 30 mm2

Substrate e 2:2 20 30 0:8 mm3
DRA material e 12:2 12 manganesemanganese
Name X Y
80.00
m1 70.0000 4.3036 Curve Info
dB(AxialRatioValue)
Setup1 : Sweep
70.00 Freq='13.34170854GHz' Phi='0deg'
dB(AxialRatioValue)
Setup1 : Sweep
60.00 Freq='13.34170854GHz' Phi='90deg'
dB(AxialRatioValue)
50.00
40.00
30.00
20.00
10.00
m1
0.00
0.00 125.00 250.00 375.00
Theta [deg]
Fig. 12.30 Axial ratio magnitude for polarization inside DRA
m1 360.0000 -0.0000 12.1194 0 Curve Info

dB(rELHCP)
20.00 Freq='15.34673367GHz' Phi='0deg'
m1 dB(rELHCP)
Setup1 : Sweep
10.00
Freq='15.34673367GHz' Phi='90deg'
-60 60
0.00
-10.00
-90 90
-120 120
-150 150
-180
Fig. 12.31 Left Circular polarization radiation plot of simulated DRA

266 12 Case Studies
m1 2.0000 2.0000 14.5283 0 Curve Info

dB(rERHCP)
19.00 Freq='15.34673367GHz' Phi='0deg'
m1 dB(rERHCP)
Setup1 : Sweep
13.00 Freq='15.34673367GHz' Phi='90deg'
-60 60
7.00
1.00
-90 90
-120 120
-150 150
-180
Fig. 12.32 Right Circular polarization of simulated DRA at 15.5 GHz
Fig. 12.33 E eld pattern with dual feed of simulated DRA

Fig. 12.34 E eld pattern when single feed along y-axis of simulated DRA
Fig. 12.35 E eld pattern when single feed along x-axis applied of simulated DRA
268 12 Case Studies
Fig. 12.36 Group delay measurements in RDRA dual feed of simulated DRA
Fig. 12.37 Phase versus frequency plot (phase distortion) of simulated DRA
12.3 Dual-Feed RDRA Hardware and Measurements 269
Fig. 12.38 S21 measurement plot in dual feed of simulated DRA
12.3 Dual-Feed RDRA with Measurements Results
In this case, the designing of RDRA dual-feed mechanism has been implemented
for circular polarization. Ferrite DRA has been used for bandwidth enhancement
using magnetization concept. The results obtained have been presented graphically.
These antennas also have been placed inside anechoic chamber to minimize
external effects that come during measurements. The gures indicate various
measurement steps involved in this process. These antennas have been simulated
and fabricated. For simulated results, Ansoft HFSS 13.0 has been used. These
graphs of radiation pattern, gain, S11, VSWR, Z11, eld distribution have been
presented and their domains are mentioned below each gure. The impedance
versus frequency has been presented. The hardware results using VNA for return
loss S11 have been also included (Figs. 12.39, 12.40 and 12.41; Table 12.1).
Optimization of the feed position for impedance match to have maximum gain
(Figs. 12.42, 12.43, 12.44, 12.45, 12.46, 12.47, 12.48, 12.49, 12.50, 12.51, 12.52,
12.53, 12.54, 12.55, 12.56, 12.57, 12.58, 12.59, 12.60, 12.61, 12.62 and 12.63).
270 12 Case Studies
Fig. 12.39 Double-feed RDRA HFSS model
DRA (4.6 x 9 x 10.8 mm)
Fig. 12.40 Design dimensions of RDRA under design
Feed design dimensions
30mm
20mm
Fig. 12.41 Ground plane with slot/stub/micro-strip feed in RDRA
Table 12.1 Specication/dimensions

S. No. Element Dimension (mm)
1 Ground plane 20 30
2 Substrate er 2:2 20 30 0.8
3 DRA er 12:2 4.6 9 10.8
4 Width of micro-strip 2.4
5 Length of stub and micro-strip 18.693
6 Slot (l w) 3.743 0.404
Fig. 12.42 Variation of slot Gain Vs slot Position

position for maximum gain in 10
Gain (db) ------>>

gain at dominant mode
RDRA 8
6
4
2
0
-1 -2 -3 -4 -5 1 2 3 4
slot position (mm) ---------->>
Fig. 12.43 Return loss plot S11 vs frequency

S11 for single and double 0
feed of RDRA
-5
<<<----------- s11(db)
-10
-15
-20
single feed
Double feed
-25
10.71
11.51
12.31
13.11
13.91
14.71
15.52
16.32
17.12
17.92
18.72
19.52
7.50
8.30
9.10
9.90
frequency(GHZ) -------->>
Fig. 12.44 Impedance match Impedance vs Frequency

for single and double feed in 60
single feed
RDRA double feed
50
Impedance(ohm) --- >>
40
30
20
10
0
7.50
8.13
8.75
9.38
10.01
10.63
11.26
11.88
12.51
13.14
13.76
14.39
15.02
15.64
16.27
16.89
17.52
18.15
18.77
19.40
Frequency(GHZ) ----->>
272 12 Case Studies
Fig. 12.45 Effect of variation S11 vs Frequency

in permittivity on return loss 5.00
er 9.8 er 20 er 40
S11 of RDRA 0.00
-5.00
<<<------S11(db)
-10.00
-15.00
-20.00
-25.00
-30.00
-35.00
-40.00
7.50
8.30
9.10
9.90
10.71
11.51
12.31
13.11
13.91
14.71
15.52
16.32
17.12
17.92
18.72
19.52
Frequency(GHz)------->>>
Fig. 12.46 Effect of variation S11 vs Frequency

in permittivity on return loss 0
S11 of RDRA
-5
er 12.9
-10
<<---- S11(db)
-15
er 20
-20
-25
er 30
-30
-35
-40
7.50
8.75
10.01
11.26
12.51
13.76
15.02
16.27
17.52
18.77
Frequency(GHz) ----->>
m1 360.0000 -0.0000 2.1468 Curve Info

0
m1 GainTotal
2.00 Freq='8.251503006GHz' Phi='0deg'
GainTotal
Setup1 : Sweep
1.50
Freq='8.251503006GHz' Phi='90deg'
-60 60
1.00
0.50
-90 90
-120 120
-150 150
-180
Fig. 12.47 Radiation pattern 1 of simulated RDRA for eld strength
m1 360.0000 -0.0000 1.8149 0 Curve Info

m2 60.0000 60.0000 0.6570 m1 GainTotal
1.60 Freq='10.05511022GHz' Phi='0deg'
GainTotal
Setup1 : Sweep
1.20 Freq='10.05511022GHz' Phi='90deg'
-60 60
0.80
0.40 m2
-90 90
-120 120
-150 150
-180
Fig. 12.48 Radiation pattern 2 of simulated RDRA for eld strength

274 12 Case Studies
m1 308.0000 -52.0000 8.6112 0 Curve Info

m2 54.0000 54.0000 7.1650 GainTotal
8.00 Freq='18.79759519GHz' Phi='0deg'
GainTotal
Setup1 : Sweep
m1 6.00 Freq='18.79759519GHz' Phi='90deg'
-60 m2
60
4.00
2.00
-90 90
-120 120
-150 150
-180
Fig. 12.49 Radiation pattern of simulated RDRA for eld strength
Fig. 12.50 Front and rear view of hardware implemented of dual-feed RDRA
Fig. 12.51 Short, open, and load termination for calibration of VNA
Fig. 12.52 Top and side view of single- and double-feed aperture couple feed of RDRA
Fig. 12.53 Back side view of

double-feed aperture couple
feed of RDRA
Fig. 12.54 Fabricated ferrite

RDRA
276 12 Case Studies
Fig. 12.55 Smith chart for impedance matching of RDRA
Fig. 12.56 Position of the

slot in ground plane of
single-feed RDRA
Fig. 12.57 RDRA investigation under testing setup
Fig. 12.58 RDRA H elds pattern showing magnetic eld strength

278 12 Case Studies
Fig. 12.59 VNA showing measured S11 of RDRA
Fig. 12.60 Fabricated model of RDRA under test with single feed and slot
Fig. 12.61 Measurements of return loss S11 of fabricated RDRA

12.4 Isolated and Grounded RDRA 279
Fig. 12.62 Slot in ground plane of fabricated RDRA
Fig. 12.63 Top view with single feed and SMA connector of fabricated RDRA
12.4 Isolated and Grounded RDRA
Design of Isolated DRA: Isolated and grounded RDRA has different lengths due to
image theory. Isolated RDRA is shown in Fig. 12.64. The RDRA is excited by a
coax feed. Ground plane is absent in the rst design. The rectangular DRA height
can be reduced to half if we use ground plane of nite dimensions (Table 12.2).
12.4.1 S11 Plot
Return loss of isolated DRA is shown in Fig. 12.65. It has resonant frequency of
3.99 GHz with 41.74 return loss.
280 12 Case Studies
Fig. 12.64 Isolated DRA
Table 12.2 Dimensions of Dimension of DRA in X-direction = 9.31 mm

isolated DRA
Dimension of DRA in Y-direction = 18.62 mm
Dimension of DRA in Z-direction = 4.6 mm
Permittivity of DRA = 37.84
Ansoft LLC XY Plot 2 HFSSDesign1

0.00 Curve Info
dB(St(coaxpin_T1,coaxpin_T1))
Setup1 : Sweep1
-5.00 Name X Y
m1 3.9900 -41.7414
m2 3.9133 -9.7575 m2 m3
-10.00
m3 4.0772 -9.8660
-15.00
-20.00
-25.00
-30.00
-35.00
-40.00 m1
-45.00
3.50 3.63 3.75 3.88 4.00 4.13 4.25
Freq [GHz]
Fig. 12.65 Return loss for isolated DRA

12.4.2 Gain Plot
Gain plot is shown in Fig. 12.66. It shows that isolated DRA has 4.7 dB gain at
resonant frequency.
12.4.3 Impedance (Z) Plot
Impedance plot of isolated DRA is shown in Fig. 12.67. This has the real imped-
ance nearly 50 at resonant frequency.
Ansoft
NameLLC Theta Ang Mag Radiation Pattern 1 HFSSDesign1
m3 90.0000 90.0000 4.7037 0 Curve Info
dB(GainTotal)
-30 30 Setup1 : Sweep1
0.00 Freq='3.99749499GHz' Phi='0deg'
dB(GainTotal)
Setup1 : Sweep1
-10.00
-60 60
-20.00
-30.00
m3
-90 90
-120 120
-150 150
-180
Fig. 12.66 Gain plot of isolated DRA

62.50 Curve Info
re(Zt(coaxpin_T1,coaxpin_T1))
Name X Y Setup1 : Sweep1
m8
50.00 m8 3.9930 50.7358 im(Zt(coaxpin_T1,coaxpin_T1))
m9 3.9915 -0.2234 Setup1 : Sweep1
37.50
25.00
Y1
12.50
m9
0.00
-12.50
-25.00
3.50 3.63 3.75 3.88 4.00 4.13 4.25
Freq [GHz]
Fig. 12.67 Impedance plot of isolated DRA

282 12 Case Studies
12.4.4 Design of RDRA with Ground Plane
Image theory has been applied to reduce the height of the antenna. HFSS model of
DRA with ground plane has been developed using excitation with coaxial feed.
Here, the height of DRA has been reduced to half as compared to isolated DRA
(Fig. 12.68; Table 12.3).
12.4.5 S11 Plot
Simulated return loss of DRA with ground is shown in Fig. 12.69. It has resonant
frequency 4.18 GHz with 28 dB return loss.
12.4.6 Gain Plot
Gain plot is shown in Fig. 12.70. It shows that antenna radiates in the end re
direction and holds the value of gain 4.62 dB at resonant frequency.
Fig. 12.68 HFSS model for

DRA with ground
Table 12.3 Dimensions of Dimension of DRA in X-direction = 9.31 mm

DRA with ground plane
Dimension of DRA in Y-direction = 9.31 mm
Dimension of DRA in Z-direction = 4.6 mm
Permittivity of DRA = 37.84

0.00
Name X Y
m1 4.1864 -28.6026 Curve Info

-5.00 m2 4.1343 -9.7735 dB(St(coaxpin_T1,coaxpin_T1))
Setup1 : Sweep1
m3 4.2345 -10.1310
m2 m3
-10.00
-15.00
-20.00
-25.00
m1
-30.00
3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00
Freq [GHz]
Fig. 12.69 Return losses for DRA with ground
Ansoft
NameLLC Theta Ang Mag Radiation Pattern 1 HFSSDesign1
m3 90.0000 90.0000 4.6291 0 Curve Info
dB(GainTotal)
-30 30 Setup1 : Sweep1
0.00 Freq='4.186372745GHz' Phi='0deg'
dB(GainTotal)
Setup1 : Sweep1
-10.00
Freq='4.186372745GHz' Phi='90deg'
-60 60
-20.00
-30.00
m3
-90 90
-120 120
-150 150
-180
Fig. 12.70 Gain pattern for DRA with ground
12.4.7 Impedance Plot
Impedance plot of DRA with ground is shown in Fig. 12.71. This has the real
impedance nearly 48.2 and reactive part is 0.17 at resonant frequency.
284 12 Case Studies

50.00 m6
Curve Info
Name X Y
re(Zt(coaxpin_T1,coaxpin_T1))
m6 4.1984 48.1805
Setup1 : Sweep1
37.50 m7 4.1824 0.1700
im(Zt(coaxpin_T1,coaxpin_T1))
Setup1 : Sweep1
25.00
12.50
Y1
m7
0.00
-12.50
-25.00
-37.50
3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00
Freq [GHz]
Fig. 12.71 Impedance plot for DRA with ground
12.4.8 Comparison of DRA With and Without Ground Plane
From the comparison in Table 12.4, it is seen that resonant frequency and gain for
isolated DRA and DRA with ground are approximately same but return loss and
impedance bandwidth are better in isolated DRA (Fig. 12.72).
Table 12.4 Comparison of DRA with and without ground plane

Parameter Isolated With ground
Resonant frequency (GHz) 3.99 4.16
Gain (dB) 4.7 4.629
B.W. (GHz) 0.17 0.1
Return loss (dB) 41.04 28

150.00 Curve Info
Name X Y im(Z(1,1))
Setup1 : Sweep1
125.00 m1 5.1303 49.9819
re(Z(1,1))
m2 5.1363 -2.4335 Setup1 : Sweep1
100.00
75.00
m1
50.00
Y1
25.00
0.00 m2
-25.00
-50.00
-75.00
4.00 4.50 5.00 5.50 6.00 6.50 7.00
Freq [GHz]
Fig. 12.72 Impedance plot of DRA

12.4.9 Detailed Design of Aperture-Coupled DRA
Figure 12.73 shows the view of a simple structure of rectangular DRA. The rect-
angular DRA of length L, width W, and height H is placed over a slot, cut at the
center of ground plane of size 50 50 mm2. The micro-strip line of length Lm and
width We is placed on the other side of the ground plane. The dielectric material
used for substrate is having permittivity, er 10:2 and thickness 0.64 mm. The
dielectric material used for DRA is Rogers RT/Duroid 6010/6010LM having per-
mittivity er 10:2. All other the dimensions have been shown in Table 12.5
(Fig. 12.74).
Fig. 12.73 HFSS model for reference antenna 1
Table 12.5 Dimensions of reference antenna

L W H Ls Ws Lm Wm Er er H (sub)
(mm) (mm) (mm) (mm) (mm (mm) (mm) (sub) (dra) (mm)
10.6 6 9.6 7.2 1.2 28 0.6 10.2 10.2 0.64
286 12 Case Studies
Fig. 12.74 Bottom view of aperture-coupled DRA

0.00 Name X Y Curve Info
m5 6.9000 -24.4423 dB(S(1,1))
m6 6.5800 -10.2532 Setup1 : Sweep1
k='28mm'
m7 7.3400 -9.8973
-5.00
dB(S(1,1))
m7
-10.00 m6
-15.00
-20.00
m5
-25.00
5.00 5.50 6.00 6.50 7.00 7.50 8.00
Freq [GHz]
Fig. 12.75 Return loss of aperture-coupled DRA
12.4.10 Return Loss
Return loss of antenna 1 is shown in Fig. 12.75. Resonant frequency is 6.89 GHz
with return loss 24 dB.
Fig. 12.76 Radiation pattern of gain for (antenna A) = 90 and = 90
300
250
im(Z(1,1))
200
re(Z(1,1))
150
Impedance
100
50
0
-50
-100
-150
5.0 5.4 5.8 6.2 6.6 7.0 7.4 7.8
Frequency [GHz]
Fig. 12.77 Impedance matching plots
12.4.11 Radiation Pattern
Antenna radiates in end re direction. Gain of antenna at = 90 and = 90 is

8.796 dBi (Figs. 12.76 and 12.77).
Annexure-1
Details of the Dielectric Materials and Their Suppliers
S. No. Material Permittivity Supplier or manufacturer

1. MgOSiO2 6.3 Countis Laboratories
(CD-6) 12295 Charles Dr, Grass Valley, CA 95945,
2. MgOSiO2TiO2 9.5 United States
(CD-9) +1 530-272-8334
tcountis@countis.com
3. MgOTiO2SiO2 13.0
(CD-13)
4. MgOTiO2 15.0
(CD-15)
5. MgOTiO2 16.0
(CD-16)
6. MgoCaOTiO2 18.0
(CD-18)
7. MgoCaOTiO2 20.0
(CD-20)
8. MgoCaOTiO2 30.0
(CD-30)
9. MgoCaOTiO2 50.0
(CD-50)
10. MgoCaOTiO2 100.0
(CD-100)
11. MgoCaOTiO2 140.0
(CD-140)
(continued)

Resonator Antennas, DOI 10.1007/978-81-322-2500-3
290 Annexure-1

12. Boron nitride 4.0 Emerson & Cuming Microwave
(ECCOSTOCK@) Products N.V.
13. Beryllium oxide 6.0 A unit of Laird Technologies
(ECCOSTOCK@) Hong Kong Holdings (4) Ltd.
Unit 2507-8, 25/F, Ofce Tower, Langham
14. Magnesium oxide 9.0
Place,
(ECCOSTOCK@)
8 Argyle Street, Mongkok
15. Magnesium titanate 10.0 Kowloon, Hong Kong
(ECCOSTOCK@) Tel: +852-2923 0600 Call: +852-2923 0605
16. Zirconia 20.0 Email: sales@hk.eccosorb.com
(ECCOSTOCK@)
17. Titanium dioxide (rutile) 50.0
(ECCOSTOCK@)
18. Strontium titanate >100.0
(ECCOSTOCK@)
19. Magnesium manganese 9.2 Hiltek Microwave Limited
Aluminum iron ferrite 15200 Shady Grove Road Suite 350
20. Magnesium titanate 16.0 Rockville, Maryland 20850
United States
21. Lithium ferrite 20.0
(301) 670-2833
22. Zirconium tin titanate 37.0 (301) 670-2831 Fax
23. Titania ceramic 80100 www.hiltek.com
24. MgSi (Steatile) 6.0 Morgan advanced Materials
(D6) 150 Kampong ampat
25. CaMgTi (Mg, Ca 20.0 05-06a
titanate) Ka Centre
(D20) Singapore
368324
26. ZrTiSn (Zr, Sn titanate) 37.0
t +65 6595 0000
(D36)
F +65 6595 0005
27. BaSmTi (Ba, Sm 76.5 asia.mc@morganplc.com
titanate)
(D37)
(continued)
Annexure-1 291

28. Titanate with other 6.5 Pacic Ceramics, Inc.
ingredients Advanced Microwave Ceramic Materials
(PD-6) 824 San Aleso Ave Sunnyvale, CA 94085
29. Titanate with other 9.5 USA (408) 747-4600
ingredients info@pceramics.com
(PD-9)
30. Titanate with other 12.0
ingredients
(PD-12)
ingredients
(PD-13)
ingredients
(PD-15)
ingredients
(PD-16)
ingredients
(PD-18)
ingredients
(PD-25)
ingredients
(PD-38)
ingredients
(PD-50)
(continued)
292 Annexure-1

ingredients
(PD-100)
ingredients
(PD-160)
ingredients
(PD-270)
41. Zr Sn Ti oxide 37.0 Temex Components & Temex Telecom,
(E2000) USA
42. E3000 34.0 Supplier 1
SM CREATIVE
43. Ba Zn Ta oxide 30.0
No 845, 2nd Cross, 7th Main HAL 2nd
(E4000)
Stage
44. Ba Sm Ti oxide 78.0 Indiranagar, Bangalore, 560 038
(E5000) India
45. Ti Zr Nb Zn oxide 45.0 sundar@smcel.com
(E6000) +91 (80) 25210268
+91 (80) 41255492
Mobile Phone Number: +91 (98) 45410417
http://www.smcel.com
S M Creative Electronics Ltd
#10, Electronic City, Sector-18
Gurgaon 122 015, Haryana
Tel: +91 124-4909850
Fax: +91 124-2455 212
smcel@smcel.com
Supplier 2
SIMAL
# 60 & 60/1, 18th Cross, 4th Main
Malleswaram, Bangalore, 560 055
India
agencies@simal.com.sg
+91 (80) 41532079
+91 (80) 23444410
Mobile Phone Number: +91 (99721) 24165
http://www.simal.com.sg
46. Cordierite (Mg, Al, 4.5 Trans-Tech
silicate) Skyworks Solutions, Inc.
(D-4) 5520 Adamstown Road
47. Forsterite (Mg, Si, 6.3 Adamstown, MD 21710
oxide) Supplier
(D-6) SM Electronic Technologies Pvt. Ltd.
#1790, 5th Main, 9th Cross, RPC Layout
48. MgTi 15.0
Vijayanagar 2nd Stage
(D-15)
Bangalore 560 040
49. MgTi 16.0 India
(D-16)
(continued)
Annexure-1 293

50. Ba, Zn, Ta-oxide 29.030.7 Mr. Manjunath
(D-29) +91-80-23301030
51. Ba, Zn, Ta-oxide 29.531.0 smgroup@vsnl.com
(perovskite)
(D-87)
52. BaZnCoNb 35.036.5
(D-83)
53. Zirconium titanate based 44.746.2
(D-43)
54. E-11 11.0 T-CERAM, RF & Microwave
55. E-20 20.0 Okrun 1144
500 03 Hradec Krlov
56. E-37 37.0
Czech Republic, EU
sales@t-ceram.com
www.t-ceram.com
+420 774 406 438
CZ 42196078
57. TE-21 21.0 Token Electronics Industry Co., Ltd.
58. TE-30 30.0 No. 137, Sec. 1, Chung Shin Rd., Wu Ku
Hsiang, Taipei Hsien, Taiwan, R.O.C
59. TE-36 36.0
TEL: 886-2-2981 0109; FAX: 886-2-2988
60. TE-45 45.0 7487
61. TE-80 80.0 http://www.token.com.tw rfq@token.com.
62. TE-90 90.0 tw
63. MgCaTi 20.0 MCV Microwave
(MDR20) 6640 Lusk Blvd, Suite A102 San Diego, CA
64. Ta with other ingredients 24.0 92121
(MDR24) Tel: 858-450-0468
Fax: 858-869-8404
65. Ta with other ingredients 30.0
www.mcv-microwave.com
(MDR30)
66. ZnSnTi 38.0
(MDR38)
67. LaBaTi 45.0
(MDR45)
68. DR-30 30.0 TCI Ceramics, Inc.
69. DR-36 36.0 18450 Showalter Rd., Hagerstown, MD
21742
70. DR-45 45.0
Ph: 301-766-0560 Fax: 301-766-0566
71. DR-80 80.0 E-mail: sales@tciceramics.com
www.tciceramics.com
72 RT6010, RT-6002 10.2
73 MCT-25 25 TRANS TECH
SMAT 27
BaTiO3 14
74 ECCOSTOCKSHIK 10,20,30,40
Annexure-2
Two-Dimensional Mathematical Model of Resonant Modes

in Cavity Resonator
See Figs. A2.1, A2.2 and A2.3.

Probe inserted dl
Characteristic equation of RDRA is given below:
kx2 ky2 kz2 er kmn

2
A2:1
The eld Ez can be expressed as follows:

r2? h2 Hz or Ez 0; Helmholtz equation
X
Ez Cmn sinmpx=a sinnpy=b ecmnz ejxt
In the above equation, Cmn are the amplitude coefcients and wave is propagating
in z-direction
q q
) cmn Propagation constant h2mn k2 h2mn x2 l
where hmn kc np=b; are possible eigenvalues.

Hence, computation of eld Ez when all the four sides of resonator are trans-
parent and magnetic walls (PMC walls) and top and bottom walls are PEC
(Electrical walls). We are well versed that Hz 0 at magnetic walls and Ez 0 at
electric walls.
The feed probe is inserted into rectangular DRA at point (a/2, b/2) in z-direction.
I(t) Current can be expressed in terms of magnetic vector potential Az.

296 Annexure-2
Fig. A2.1 RDRA without

ground plane
Fig. A2.2 Ground plane of

RDRA
Fig. A2.3 RF feed
^z lIdl jkr
Az e ; r is far field point:
4pr
jk^r ^z lIdl ejkr
div A
4pr
jk cos h lIdl ejkr
div A ;
4pr
jx6 9
2
c
Annexure-2 297
Let,
p kc lIdl
69 a 2 b2 ejkr
x 4pr
cos hlIdl jkr
e
4pr
Now,
E r jxA; Lorentzs gauge condition

1
Er 0 O 2
r
Eh jxAh jxAz; at z 0
E jxA 0 at z 0
Hence,
Eh Ez jxAz
jxlIdl jkr 2 mpx npy
e p sin sin
4pr ab a b
p
where r x2 y2
dm; m0 ; dn; n0 hunm ; um0 n0 i; where z = 0; (property of orthogonality as product
of basis function becomes zero)
X
Ez Cmn umn x; yecmnz
mn
hmn x
At z = 0
X jxlIdl p
p ejk x y ;
2 2
Ez ) Cmn umn x; y
mn 4p x y
2 2
Hence, amplitude coefcient
Za Zb r
jxlIdl umnx;y 2 2
Cmn pejk x a=2 x b=2 dxdy;
4p x2 y2
0 0
298 Annexure-2
Za Zb r
jxlIdl umnx;y jk
2 2
Cmn r
2 e x a=2 x b=2 dxdy
4p 2
0 0 x a=2 y b=2
A2:2

m 2 n 2
Hence p
2
x2
a b
if a [ b and m 1; 2; 3; . . . n 1; 2; 3; . . .

2 1 1 2 2 1
p x \p
2

a2 b2 a2 b2
r r
1 1 x 2 1
\
a2 b2 p a2 b2
jpp
cmn
d
k2 c2mn h2mn
p2 p2
hence, k2 h2mn
d2
Cmn Fourier coefcients of modes;

umn depends on input excitation;
hmn resonant mode (cut off frequency); and
k-propagation constant.
Generation of modes or characteristics frequencies xmnp e.m. of electro-
magnetic elds oscillations inside the cavity resonator has been discussed. The
basic Maxwells theory can be applied with boundary conditions to express res-
onator elds as superposition of these characteristics frequencies.
The elds
X Z
Ez x; y; z; t Re Cmnp ejxmnpt umnp x; y; z
mnp
or
X
Cmnp umnp x; y; z cosxmnp Umnp;
mnp
Annexure-2 299
npy
where umn x; y p2
ab
sin mpx
a sin b ; xmnp is the characteristic frequency and
xmnp is the phase of current applied. The rectangular cavity resonator is excited
at the centre with an antenna probe carrying current I(t) of some known frequency
xmnp. This generates the eld Ez inside the cavity of the form given below:
Z p2
jxlIdlx2 y2 jxtxc x2 y2 d
Ez x; y; d; t Gx; y 3=2 e I xejxt dx
4p x2 y2 d2
where G(x, y) are the constant terms associated with current.

Equating resonator eld with the antenna current elds at z d plane;
Antenna or resonator radiation current or elds
r
X 2 ppd
Cmnp sin cosxmnpt /mnp
p
d d
Antenna probe current

Z
jxlIdlx2 y2
Gx; y 3=2 I x
4p x2 y2 d2
p2
e dx e jxt c x y d wmnp umn x; ydxdy;
x 2 2
jxt
Multiply both sides by ejxmnpt and then taking time averaging (KAM) gives us
the following
r ZT
2 ppd jmnp 1 jxlIdlx2 y2 jxmnpt
Cmnp sin e lim Gx; y 3=2 e
d d T!/ 2p
4p x2 y2 d2
T
p2
I xejxt dx e jxt c x y d wmnp umn x; ydt
x 2 2
It is clear that for these two expressions to be equal, the probe current can be
dened as
1X h i
Ix jI mnpj dx xmnpejmnp ejmnp dfx xmnpg
2 mnp
The antenna probe current must contain only the resonator characteristics
frequencies xmnp, then
300 Annexure-2
r
X 2 ppd
Cmnp sin cosxmnpt /mnp
p
d d
Z
jxlIdlx2 y2 A2:3
Gx; y 3=2 I x
4p x2 y2 d2
p2
ejxt dxe jxt c x y d wmnp umn x; ydxdy
x 2 2
Antenna probe current = Resonator radiated current or magnetic elds, as per the
law of conservation of energy. The modes diagrams are given below (Figs. A2.4,
A2.5, A2.6, A2.7, A2.8, A2.9, A2.10, A2.11, A2.12, A2.13, A2.14 and A2.15):
Fig. A2.4 Mode diagram
Fig. A2.5 TE112
Fig. A2.6 TE113
Fig. A2.7 TE114

Annexure-2 301
Fig. A2.8 TE111
Fig. A2.9 TE112
Fig. A2.10 TE113
Fig. A2.11 TE115
Fig. A2.12 TE116

302 Annexure-2
Mode sketch
Fig. A2.13 TE114
Fig. A2.14 TE118
Fig. A2.15 TE114
Rectangular design
2 1
xmn pp
m n2 2
le a2 b2 , in two-dimensional case
@2X @2Y @2Z

kx2 x 0; ky2 y 0; kz2 z 0
@2x @2y @2z
er k02 kx2 ky2 kz2 ; where k is wave number
hkx x is the harmonic function and can be written as follows: sinkx x or coskx x:
These are solution of wave function and if boundary conditions are applied, then
eigenvalues can be dened as follows:
2pf0 q
k0 ; ky tanky d=2 er 1k02 ky2
c
c q
kx2 ky2 kz2 er k02 ; Resonant frequency f0 p kx2 ky2 kz2
2p er
kz d q
where kx m pa ; ky n pb ; and kz tag 2 er 1k02 kz2 :
Annexure-2 303
The resonance frequency of this antenna can be estimated using the approximate
analytical expressions for the resonance frequency of TE111 mode in the a rectan-
gular resonator (three dimensional) given by
r

c p2 p 2 p2
f111 p ;
2p er a 2b d
Three-dimensional case
Propagation constant, c2 k 2 kc2
A2.1 Fourier Series
1
a0 X 2np 2np
f x an cos x bn sin x
2 n1
a a
Za
2 2np
an f x cos x dx
a a
0
Za
2 2np
bn f x sin x dx
a a
0
Half-wave Fourier analysis will have odd or even terms, i.e., sinesine or
cosinecosine.
If f(x) = f( x), then even harmonics will take place and only cosine terms will
occur, i.e.,
X
1 pnx
f x Cn cos
n1
a
Z np
2 a
where Cn f x cos x dx
a 0 a
Similarly for odd terms, f x 6 f x;
X
1 pnx
f x Bn sin
n1
a
Z a np
2
where Bn f x sin x dx:
a 0 a
304 Annexure-2
A2.2 Spectral Resolution of EM Waves
Every wave can be subjected to the process of spectral resolution, i.e., can be
represented as a superposition of monochromatic waves of various frequencies. The
character of this expansion varies according to the character of the time dependence
of the elds.
One category consists of those cases where the expansion contains frequencies
forming a discrete sequence of values. The simplest case of this type arises in the
resolution of a purely periodic eld. This is the usual expansion in Fourier series. It
contains the frequencies which are integral multiples of the fundamental fre-
quency x0 2p=T; where T is the period of the eld. We therefore write it in the
form as follows:
X
1
f fnejx0 nt
n1
where f is any of the quantities describing the eld. The quantities fn are dened in
terms of the function f by the integrals
ZT=2
1
fn f tejnx0 tdt :
T
T=2
Because f(t) must be real
fn fn :
in more complicated cases, the expansion may contain integral multiples of several
different incommensurable fundamental frequencies. When the sum is squared and
averaged over the time, the product of terms with different frequencies is given zero
because they contain oscillating factors.
Only terms of the form fn fn jfn j2 remain. Thus, the average of the square of
the eld, i.e., the average intensity of the wave, is the sum of the intensities of its
monochromatic components.
P P
f 2 n 11 jfn j2 2 1 2
n1 jfn j , where it is assumed that the average of
the function f over a period is zero. Another category consists of elds which are
expandable in a Fourier integral containing a continuous distribution of different
frequencies. For this to be possible, the function f(t) must satisfy certain denite
conditions; usually we consider functions which vanish for t ! 1:
Similarly, fx fx ; let us express the total intensity of the wave, i.e., the
integrals of f 2 over all time, in terms of the intensity of the Fourier components.
Now, we have
Annexure-2 305
8 9 8 9
Z1 Z1 < Z1 = Z1 < Z1 = dx
dx
f 2 dt f fx ejxt dt fx fejxt dt
: 2p; : ; 2p
1 1 1 1 1
Z1
dx
fx fx ;
2p
1
or
Z1 Z1 Z1
2 dx dx
f dt
2
jfx j 2 jfx j2 :
2p 2p
1 1 0
R
1 1 jxt
f t 2p 1 fx e dx, where the
R 1Fourier components are given in terms of the
function f t by the integrals, fx 1 f tejxt dt:
A2.3 Coordinate System and Their Transformations
Rectangular x; y; z, cylindrical q; /; z, and spherical r; h; / coordinates can be

expressed as follows:
x q cos / r sin h cos /:

y q sin / r sin h sin / z r cos h:
p
q x2 y2 r sin h:
p
1 x2 y2 q
h tan tan1
z z
Transformations of the coordinate components of a vector among the three

coordinate systems are given by
Az Aq cos / A/ sin /
Ar sin h cos / Ah cos h cos / A/ sin /
Ay Aq sin / A/ cos /
Ar sin h sin / Ah cos h sin / A/ cos /
Az Ar cos h Ah sin h
Aq Az cos / Ay sin / Ar sin h Ah cos h
A/ Ax sin / Ay cos /
306 Annexure-2
Ar Ax sin h cos / Ay sin h sin / Az cos h

Aq sin h Az cos h
Ah Ax cos h cos / Ay cos h sin / Az sin h

Aq cos h Az sin h
unit vector in the three systems are denoted by ux ; uy ; uz ; uq ; u/ ; uz ;

and ur ; uh ; u/
dr dxdydz qdqd/dz r 2 sin hdrdhd/
Area Ds ux dydz uy dxdz uz dxdy

uq qd/dz u/ dqdz uz qdqd/
ur r 2 sin hdhd/ uh r sin hdrd/ u/ rdrdh
Length L ux dx uy dy uz dz
uq dq u/ qd/ uz dz
ur dr uh rdh u/ r sin hd/
Scalar multiplication is dened by
A B A1 B1 A2 B2 A3 B3
r rv r2 v
r r A r r A r 2 A
Rerej r cosxt h
Imrej r sinxt h
Kronecker Tensor
na2
f 2pr
1
2 e
2r2where a is the mean and is the variance and vector multipli-
cation can be dened as:

u1 u 2 u 3

A B A1 A2 A3
B1 B2 B3
The differential operators are the gradient rx;

Annexure-2 307
Divergence r A;
curl r A
Laplacian operator r2x
In rectangular coordinates, we can think of del r as the vector operator
@ @ @
r ux uy u z
@x @y @z
@x @x @x
r x ux uy uz
@x @y @z
@Ax @Ay @Az
rA
@x @x @x

ux uy u z
@ @ @
r A @x @y @z
A A A
z y z
@x2 @x2 @x2

r2x
@x2 @y2 @z2
In cylindrical coordinates, we have
@x 1 @x @x
rx uq u/ uz
@q q @/ @z
1 @ 1 @A/ @Az
rA qAq
q @q q @/ @z

1 @Az @A/ @Aq @Az 1@ 1 @Aq

r A ur u/ uz qA/
q @/ @z @z @q q @q q @/

1 @ @x 1@ 2
@ 2
r2x q 2 x2 x2
q @q @q q @/ @z
In spherical coordinates, we have

308 Annexure-2
@x 1 @x 1 @x
r x ur u/ u/
@r r @/ r sin h @/
1 @ 2 1 @ 1 @A/
rA 2 r Ar Ah sin h
r @r r sin h @h r sin h @/
1 @ @Ah
r A ur Ah sin h
r sin h @h @/

1 1 @Ar d 1 @ @Ar
uh rA/ u rAh
r sin h @/ dr r @r @h

1 @ @x 1 @ @x 1 @2
r2x 2 r2 2 sin h 2 2 x2
r @r @r r sin h @h @h r sin h @/
R ux x uy y uz z
And the source coordinates by
r 0 ux x 0 uy y 0 uz z 0
q
0
jr r j x x0 2 y y0 2 z z0 2
0
Ilejkjrr j
A
4pjr r 0 j
To emphasize that A is evaluated at the eld point (x, y, z) and Il is situated at the
source point x0 ; y0 ; z0 (Table A2.1),
0
Ilr 0 ejkjrr j
Ar A
4pjr r 0 j
Table A2.1 Frequency in Hz

Frequency Symbol Frequency in Hz
Tera T 1012
Giga G 109
Mega M 106
Kilo K 103
Hecto H 102
Deca Da 101
Deci d 101
Centi cm 102
Milli mm 103
Micro 106
Nano n 109
Pico p 1012
Femto f 1015
Annexure-3
Design Steps of RDRA Using ADS Software
Steps !
1. Export the model from HFSS and save in G drive or any le (without Path).
2. Now right click the ADS icon and click run as administration.
3. Then, click the yes button.
4. Then, click the cancel and go create new project.
5. Now on schematic will open, then go to layout button, then go to create update
layout.
6. Then go to le button and then go to import button, the layout model is
complete.
7. Now click on line which is connected to the patch on layout model then delete
it.
8. Then go to the view button then go to layer view then go to by name. Then go
to conductor 2 button, now then drag the feed or patch and date it.
9. Now go the each capacitor then click double and give it value according to the
formula.
Cv 26 f at v 0 Ci 1:298lf
Cf 0 f and add each capacitor by line by clicking on line icon.
10. Now go to S-parameter then click on termination which also given in Fig. A3.1.
11. Now go to the S.P (S-parameter) button and put on schematic window then
click the S-parameter which is on the schematic window and put frequency 1 to
3 by stepping 1 MHz frequency then ok.
12. Now go to simulate button and simulate it then after completing the button.
13. Now then go to EDS model, then go to substrate and create update then go to
open button put substrate (RT Duroid-5880) then put the thickness of the
substrate (1.524 mm) loss tangent (0.001) then go to apply and then go to ok.
14. Now again go to EDS model then go to component. Now go to create update
then put start frequency and stop frequency 3 GHz.
15. Now put the port on the patch by single clicking on the patch from port Ze on.
16. Now minimize it.
310 Annexure-3
Fig. A3.1 Then go to

button and connected to the
termination
17. Now go to schematic window and then go to library le and click on anywhere
on schematic window.
18. Now go to lumped element and select on capacitor and put three by pressing
control button.
PCB manufacturing from HFSS model
1. Save HFSS model bottom as view .dxf le after going to modeler and
exporting it
2. open .dfx in AutoCAD to generate .pdf or image as .jpg format.
3. use butter paper to place this design on to PCB
4. now connect SMA connectors and it is ready for testing antenna parameters.
II. HFSS design steps:
APPLY MAGNETIC AND ELECTRIC BIAS TO MHD ANTENNA
MAGNETIC BIASING STEPS WITH HFSS:
1. MAKE THREE SLOTS
2. SLOTS SHOULD BE ENCLOSING MICRO-STRIP FEED LINE
3. THE UPPER EDGE OF ALL THE SLOTS SHOULD TOUCH EACH
OTHER
4. THE SUBS AND SLOTS SHOULD NOT INTERSECT
5. UNITE ALL THE SLOTS
6. SELECT MATERIAL
7. (A) FERRITE
8. (B) MAGNETIC SATURATION EG 500 TESLA
9. GO TO BOX ASSIGN EXCITATION MAGNETIC BIAS
10. NEXT
11. PERMEABILITY
12. X, Y, Z VALUE-DESIRED
13. FINISH
14. CHECK FOR VALIDATION
Annexure-3 311
15. RELOCATE SLOT IF REQUIRED

16. SIMULATE
ELECTRIC BIASING STEPS WITH HFSS:

1. INSERT TWO BOXES OF COPPER INSIDE THE DRA OVER THE
SLOT
2. NOW APPLY VOLTAGE BIAS BY RIGHT CLICK AND APPLY +15 V

312 Annexure-3
3. WHEN CLICK ON VOLTAGE, THIS WINDOW COME WHERE WE

ENTER VOLTAGE AND E FIELD DIRECTION
4. IN THE SAME WAY, WE APPLY ELECTRIC BIAS TO SECOND

ELECTRODE
Annexure-4
Resonating Modes in Rectangular Resonators
See Fig. A4.1.
Rectangular waveguide solution:

Helmholtz equation
r2 w k2 w 0 source less medium

r2 w k2 w lj medium with source
Maxwells equations
@H
r E l
@t
@E
rH j
@t
Solving LHS of both sides rst

i j k
@ @Ez @Ey @Ez @Ex @Ey @Ex
r E @x @
@y
@
@z i j k
Ex Ey Ez @y @z @x @z @x @y

i j k
@ @Hz @Hy @Hz @Hx @Hy @Hx
r H @x @
@y
@
@z i j k
Hx Hy Hz @y @z @x @z @x @y
Comparing with RHS in both equations and getting value of Hx, Hy, Hz and Ex,
Ey, Ez, we get

314 Annexure-4
Fig. A4.1 Resonating modes

in rectangular resonators

1 @Ez @Ey
Hx A4:1
jxl @y @z

1 @Ez @Ex
Hy A4:2
jxl @x @z

1 @Ey @Ex
Hz A4:3
jxl @x @y

1 @Hz @Hy
Ex A4:4
jxe @y @z

1 @Hz @Hx
Ey A4:6
jxe @x @z

1 @Hy @Hx
Ez A4:7
jxe @x @y
@
Substituting: @z c;
jxe @E @Hz
@y c @x
z
jxl @H @Ez
@y c @x
z
Hx Ex
c2 x2 le c2 x2 le
jx @E @Hz
@x c @x
z jxl @H @Ez
@x c @y
z
Hy Ey
c2 x2 le c2 x2 le
d2 H z d2 H z
c2 x2 le Hz 2
dx2 dy
d2 Ez d2 Ez
c2 x2 le Ez 2
dx2 dy
On looking above equations, we get that Hz, Ez in 2-D Helmholtz equation

Now, rewriting Helmholtz equation for source-free medium
Annexure-4 315
r2 w k 2 w 0
Here, k is the wave number
W X xY yZ z
2
1 dX 1 d2 Y 1 d2 Z
k2 0
X dx2 Y dy2 Z dz2
Separating the independent terms, we get

1 d2 X
kx2
X dx2

1 d2 Y
ky2
Y dy2

1 d2 Z
kz2
Z dz2
k2 kx2 ky2 kz2

W A sin kx x B cos kx x C sin ky y D cos ky y ejkz z
Solving above function and keeping propagation in +z-direction only, we get TE

mode
Xn mpx npyo
Hz Cmn cos cos ejkz z ; Cmn Fourier Coefficients A4:8
mn
a b
TM mode
Xn mpx npyo
Ez Dmn sin sin ejkz z ; Dmn Fourier Coefficients A4:9
mn
a b
These Fourier coefcients are resultant of mode amplitude and propagation

constant at any instant.
Let c jkz and m, n are integers and a, b are dimensions;
mp2 np2
kc mn ; cut off frequency
a b

mp 2 np 2
kz2 x2 le
a b
316 Annexure-4
Hence, EM wave will propagate in z-direction if:

mp 2 np 2
x le
2
[0
a b
This gives cutoff frequency as follows:

s
1 mp2 np2
xc p
le a b
It means, waveguide will support all waves having greater than xc to

propagate.
Now, rewriting Hz and Ez
Xn mpx npyo
Hz Cmn cos cos ejkz z
mn
a b
Xn mpx npyo
Ez Dmn sin sin ejkz z
mn
a b
Here Cmn and Dmn are coefcients of Hz and Ez elds

Eixx;y Incident EM wave in x-direction;
Eiyx;y Incident EM wave in y-direction;
" #
X jxlDmn np cmn Cmn mp mpx npy
Eixx;y b a
cos sin expcmn z;
h2m;n a b
Similarly,
" #
X jxlDmn mp cmn Cmn np mpx npy
Eiyx;y a b
cos sin expcmn z;
h2m;n a b
On simplication
np mp
jxlDmn cmn Cmn
Eixm;n b a
h2m;n
Annexure-4 317
Similarly
mp np
jxlDmn cmn Cmn
Eiym;n a b
h2m;n

" mp cm;n np jxl
#
Eixm;n a h2m;n b h2m;n Cmn

c np jxl mp ;
Eiym;n m;n
b h2m;n a h2m;n
Dmn
we can now get the value of Cmn , Dmn after substitution of Eixm;n ; Eiym;n values.
2 2
Where h2m;n ma nb and
q
cmn h2m;n x2 le
Hence, Cmn and Dmn gives us relative amplitudes of Ez and Hz elds in TM or TE

modes.
Hence, we get solution of possible amplitudes and phase of wave propagating
through rectangular waveguide called as modes of propagation.
Half-wave Fourier expansion in waveguide is given as follows:
Za mpx mpx Zb
npy npy
fmn cos cos dx sin sin dy;
a a b b
0 0
even or odd terms, i.e., f(x) = f(x) for even term (all cosine terms) or even modes.
Where m, m and n, n 1
Za Zb mpx npy
2
Eixm;n Eixx;y cos cos dxdy
ab a b
0 0
Za Zb mpx npy
2
Eiym;n Eiyx;y sin sin dxdy
ab a b
0 0
cosinecosine.
If f(x) = f(x), even harmonics will take place and only cosine terms will occur,
i.e.,
X
1 pnx
f x Cn cos
n1
a
318 Annexure-4
where
Za np
2
Cn f x cos x dx
a a
0
Similarly for odd terms, f(x) f(x);
X
1 pnx
f x Bn sin
n1
a
where
Za np
2
Bn f x sin x dx
a a
0
Solving wave equation with boundary conditions Etan 0; we nd E elds and

then H elds. Now shape and size of resonator is given, wave equation shall give
solution of characteristic frequencies xmnp called eigenvalues or eigenfrequen-
cies of e-m oscillations of cavity resonator.
Lowest eigenfrequency x1 is cl ; where l is the dimension of resonator.
2
Higher frequency x cl ; then x is 2pvx2 c3 :
Hence, it depends on volume and net on shape of resonator.
X pmx pny prz
For resonator: fmnr sin sin sin f x; y; z
mnr
a b d

@ 2
@ @ 2 2
wx; y; z k 2 wx; y; z f x; y; z;
@x2 @y2 @z2
Helmholtz equation
X pmx pny prz
wx; y; z Cmnr sin sin sin
mnr
a b d

m2 h2 r 2
k 2 p2 C mnr fmnr
a2 b2 d 2
fmnr
Amplitude coefficient; C mnr 2 2 2
.
k2
m
a2 hb2 dr 2
xmnr
Annexure-4 319
p
k x l
x xmnp d; where d small deviation and r is different from p.

Hence,
fmnr
C mnr
xmnp d2 xmnr2
fmnr

dxmnp xmnr
A4-3 Solution of Single-String Resonator
x00 t x20 xt Bejxt
xt Aejxt
x20 x2 A B
Hence,
B
A
x20 x2
jxt
xt xBe
2 x2 , if x0 x; then x(t) will be 1
0
Now, x x0 d when d is small deviation
Bejxt

x0 xx0 x
Hence, the solution of spring resonator is in one dimension
Bejxt

d2x0
320 Annexure-4

@2 1 @2
ux; t 0; at boundaries
@x2 c2 dt2
u0; t 0 and uL; t 0
Taking Fourier transform of above equation

@ 2 x2
u^x; x 0
@x2 c2
Writing above terms in sine and cosine form, we have

xx xx
A sin B cos 0
c c
û0; x 0
ûL; x 0

c 0; Hence k L np; sine values to be zero.
sin xL
x kc npc
L , when n = 1, 2, 3 where k = /c;
when 2L, it is fundamental frequency x1
when L, the frequency is 2x1
when 2L/3, the frequency 3x1 ;
which can be generalized as:
X npx
C nsin
n
L
A-4 Solution of Two-Dimensional Resonator
General Helmholtz equation is given below (Fig. A4.2):
Fig. A4.2 Rectangular

resonator
y=b
x=a
Annexure-4 321
@ 2 wx; y; t @ 2 wx; y; t 1 @ 2 wx; y; t

2 0
@x2 @y2 c @t2
Applying boundary conditions
w0; y; t wa; y; t 0
wx; 0; t wx; b; t 0
Let input excitation be some tension T

@2w @ @w @ @w
rdxdy 2 Tdy dx Tdx dy
@t @x @x @y @y
Y 00 X 00
kY2 ; kX2 ;
Y X
Now from Helmholtz equation:
@2w
c2 r 2 w 0
@t2
Using separation of variables:
wx; y; t X xY yT t
00
T 00 t X x Y 00 y
x2 c2 A4:10
T t X x Y y
Let
X x sinkx x

Y y sin ky y
322 Annexure-4
x2
kx2 ky2
c2
where kx and ky can be

np mp
kx ; ky
b a
q
2 2
Equation (A4.1) can be written as xmn cp ma nb :
From Fourier series analysis
X
1 mpx npyh i
xm; n sin sin Cm; nejxm;nt Dm; nejxm;nt
mn1
a b
A4:12
At t 0; wx; y; 0 w0 x; y
On differentiating equation w0 x; y, we get w1 x; y; 0 w1 x; y:
When t 6 0;
X
1 mpx npy
w0 x; y Cm; n Dm; n sin sin A4:13
mn1
a b
X
1 mpx npy
w1 x; y jxm; nC m; n Dm; n sin sin A4:14
mn1
a b
Za Zb mpx npy
2
p w0 x; y sin sin dxdy C m; n Dm; n
A4:15
ab a b
0 0
Similarly,
Z mpx npy
2
jxm; n p w1 x; y sin sin dxdy C m; n Dm; n
ab a b
A4:16
Hence, obtain the value of Cm; n; Dm; n from Eqs. (A4.3) and (A4.4)
Annexure-4 323
ZZ mpx npy
1
Cm; n; Dm; n
p w0 x; y sin sin dxdy
ab a b
Z mpx npy
A4:17
1
w; x; y sin sin dxdy
jxm; n a b
Hence, from Eq. (A4.17),

m px n py
0 0
w0 x; y A sin sin
a b
m px n py
0 0
w1 x; y B sin sin
a b
due to force, perturbation occurs (Fig. A4.3)

Solving equation (A4.17)
Cm; n; Dm; n dm m0
dn n0

A a b 1 B a b
p p
ab 2 2 ab jxm0 ; n0 2 2
p
p A j B ab
C m; n; Dm; n ab A jBdm m0
dn n0
4 4 4
p m px n py
ab
ejxm0 ;n0 t
0 0
wx; y; t ReA jB sin sin
2 a b
Hence, we complete solution of two-dimensional resonator.

p m px n py
ab 0 0
wx; y; t A cosxm0 n0 t B sinxm0 n0 t sin sin
2 a b
A4:18
Fig. A4.3 Deformation due T(x+ x,y+ y)

to excitation T(x, y)
T(x,y + y) T(x+ x,y)

324 Annexure-4
Alternate method
m 2b sin h;
n 2a cos h;
Dividing both sides of above equations by 2a and 2b and adding them gives us
1 n2 m2
2 2; where k2 kx2 ky2
k2 4a 4b
Thus, resonant frequency of resonator can be determined.

Half-wave Fourier analysis:
1
a0 X 2np 2np
f x an cos x bn sin x
2 n1
a a
Z a
2 2np
an f x cos x dx
a a
0
Za
2 2np
bn f x sin x dx
a a
0
cosinecosine.
If f(x) = f(x), even harmonics will take place and only cosine terms will occur,
i.e.,
X
1 pnx
f x Cn cos
n1
a
where
Za np
2
Cn f x cos x dx
a a
0
Similarly for odd terms, f x 6 f x;
X
1 pnx
f x Bn sin
n1
a
Annexure-4 325
where
Za np
2
Bn f x sin x dx
a a
0
Spectral resolution of EM waves

Every wave can be subjected to the process of spectral resolution, i.e., can be
represented as a superposition of monochromatic waves of various frequencies. The
character of this expansion varies according to the character of the time dependence
of the elds.
One category consists of those cases, where the expansion contains frequencies
forming a discrete sequence of values. The simplest case of this type arises in the
resolution of a purely periodic eld. This is the usual expansion in Fourier series. It
contains the frequencies which are integral multiples of the fundamental fre-
quency x0 2p=T , where T is the period of the eld. We therefore write it in the
form as follows:
X
1
f fnejx0 nt
n1
(where f is any of the quantities describing the eld). The quantities fn are dened in
terms of the function f by the integrals
ZT=2
1
fn f tejnx0 tdt:
T
T=2
Because f(t) must be real
fn fn :
in more complicated cases, the expansion may contain integral multiples of several
different incommensurable fundamental frequencies. When the sum is squared and
averaged over the time, the product of terms with different frequencies is given zero
because they contain oscillating factors.
Only terms of the form fn fn jfn j2 remain. Thus, the average of the square of
the eld, i.e., the average intensity of the wave, is the sum of the intensities of its
P P1
monochromatic components. f 2 1 2
n1 jfn j 2
2
n1 jfn j ; where it is
assumed that the average of the function f over a period is zero. Another category
consists of elds which are expandable in a Fourier integral containing a continuous
distribution of different frequencies. For this to be possible, the function f(t) must
satisfy certain denite conditions; usually we consider functions which vanish for
t ! 1.
326 Annexure-4
Similarly, fx fx ; Let us express the total intensity of the wave, i.e., the
integrals of f 2 over all time, in terms of the intensity of the Fourier components.
Now, we have
8 9 8 9
Z1 Z1 < Z1 = Z1 < Z1 = dx
dx
f 2 dt f fx ejxt dt fx fe jxt dt
: 2p; : ; 2p
1 1 1 1 1
Z1
dx
fx fx ;
2p
1
or
Z1 Z1 Z1
2 dx dx
f dt
2
jfx j 2 jfx j2 :
2p 2p
1 1 0
R
1 1 jxt
f t 2p 1 fx e dx; where the Fourier components are given in terms of the
function R1
f(t) by the integrals, fx 1 f tejxt dt:
Power and Energy Signals:

Let x(t) is the input signal, i.e., voltage signal. As per Parsevals power theorem,
energy associated with this signal be
Z1
E jxtj2 dt; in time domain
1
Z1
1
jXxj2 dx; in frequency domain
2p
1
The amount of energy radiated by this signal, when applied across

Antenna having radiation resistance Rr shall be
Z1 Z1
1 2 1
E jxtj dt jXxj2 dx
Rr 2pRr
1 1
Now if input signal is x(t) having current signal
Z1 Z1
2 R
E Rr jxtj dt jX xj2 dx
2p
1 1
Annexure-4 327
ESD energy spectral density; energy spread per unit volume across 1 resister
ESD jX xj2
Discrete Fourier transform (DFT) in time domain into frequency domain spectral
analysis
X
N 1
j2pnk
x k xne ; k 0; 1; . . .; N 1:
n0
N
1XN 1
j2pkn
X n X ke ; n 0; 1; 2; . . .; N 1:
N k0 N
X(n) nite sequence.

DFT has nite length N, period N

wh; / k !rn r0 n 1kd sin h
X
N
E sin h ejn1kd sin h
n1
Annexure-5
Resonant Mode Generation and Control in RDRA
In this annexure, resonant modes TE and TM have been generated inside RDRA
whose dimensions are a, b, and d. Two parallel plates are attached along with
dielectric slab in between these plates to RDRA. This slab forms non-resonant part
and RDRA is main resonant. This is shown in Fig. A5.1a, b. The resonant modes
dominant and higher-order modes are being generated by maintaining appropriate
aspect ratio of RDRA. Then, the non-resonant slab inductance and capacitance is
introduced into main RDRA. This lumped value of inductance and capacitance is
seen in the resonant frequency.
(a) The increase in the length of internal strip introduce shift in the higher
resonant modes frequency, as they shift toward lower side and vice versa.
Hence, resonant frequency is reduced.
(b) On the other side, increase in the length of external strip introduces shift in
the lower resonant modes frequency shifts toward higher side and vice
versa. Hence, frequency is increased with strip length.
(c) Increase in spacing between parallel plates introduces the combined effect
of internal as well as external strip length variation, i.e., higher- and
lower-order resonant modes shift toward the centre frequency which can be
seen as mode-merging effect.
(d) Finally, the effect of placing a lumped varactor diode between parallel plates is
seen. The increase in the capacitance value of lumped varactor diode causes
shift in the higher resonant frequency toward lower resonant frequency
side.
These results have been investigated using HFSS and they shown with S11
results along with each RDRA model. By varying length, a, width b, and
height d of RDRA modes are generated. The internal strip, external strip, and
dielectric slab and dielectric constant provided several degrees of freedom in the
RDRA design. This has extended the control on the amount of coupling, hence
resonant frequency. This shall have large impact on resonant modes, compactness
of antenna, radiation pattern, and polarization.

330 Annexure-5
Fig. A5.1 a RDRA with two

parallel standing strips having
rectangular non-resonant slab
in between. b RDRA with
lumped varactor diode
between strips
A5.1 Effect of Change of Aspect Ratio (a/b) and (a/d) of RDRA

on Resonant Modes
See Fig. A5.2.
A5.2 Effect of Strip Length, Separation, r on the Modes Developed

Inside the RDRA
See Fig. A5.3 and Table A5.1.

Annexure-5 331
(a)
Name XY XY Plot 4 HFSSDesign1 ANSOFT

-2.50 14.9500 -17.6485
m1 Curve Info
m2 17.1000 -17.6193 dB(S(waveport,waveport))
m3 18.0500 -19.9365 Setup6 : Sweep
dB(S(waveport,waveport))
-5.00
-7.50
-10.00
-12.50
-15.00
m1 m2
-17.50
m3
-20.00
10.00 12.00 14.00 16.00 18.00 20.00
Freq [GHz]
(b)
Name XY XY Plot 4 HFSSDesign1 ANSOFT

0.00 16.0500 -12.9889
m1 Curve Info
Setup6 : Sweep
-2.50
-5.00
-7.50
-10.00
m1
-12.50
-15.00
m2
-17.50
10.00 12.00 14.00 16.00 18.00 20.00
Freq [GHz]
Fig. A5.2 a Higher-order modes generated in RDRA with square base. b Higher-order modes
generated in the rectangular base RDRA
332 Annexure-5
Fig. A5.3 RDRA with parallel standing strips
Table A5.1 Effect of strip length

Structure x (mm) y (mm) z (mm) er
DRA 4.6 9 10.8 9.8
substrate 20 30 0.8 2.2
Micro-strip 2.4 (width)
Rect. SLAB 1 9 10.8 1
External strip 2.4 10.5
Internal strip 2.4 3
A5.2(a) Effect of Internal Strip Length Variation on Resonant

Modes Inside RDRA
The effect of the internal strip length is seen on resonance frequency and resonant
modes of RDRA.
The reflection coefcient plot can be seen for the possible changes as given in
Fig. A5.4a.
The effective electrically length of RDRA is changed by introducing change in
length of internal strip as given below.
Changing the effective dimension of the dielectric resonator changes the reso-
nant frequency.
Annexure-5 333
Fig. A5.4 a External strip (xed) = 10.5 mm and variation in internal strip from (2 mm).
b External strip (xed) = 10.5 mm and variation in internal strip from (2.5 mm). c External strip
(xed) = 10.5 mm and variation in internal strip from (3 mm)
334 Annexure-5
Fig. A5.4 (continued)
A5.2(b) Effect of External Strip Length Variation on Resonant

Modes Inside RDRA
The effect of the external strip length on resonance frequency and resonant modes is
shown in Fig. A5.5b. Internal strip (xed) = 3 mm and variation in external strip
from 10.5, 7, 0 mm is investigated. Contrary to the previous case, the third reso-
nance stays mainly xed at the same frequency, while the rst and second resonant
frequencies are considerably decreased with increasing external strip length.
A5.2(c) Effect of Separation Width Between the Two Parallel

Standing Strips and er
The effect of the spacing between parallel plates and permittivity of the rectangular
slab between parallel plates is seen on resonance frequency and modes (Figs. A5.6
and A5.7).
The separation width variation ranges as 0.5, 1.5, and 2.5 mm [er 1, external
strip = 10.5 mm and inner strip = 3.5 mm (xed)].
we will change the variable separation Width (0.5, 1.5, 2.5) for er 2 keeping
external strip = 10.5 mm, inner strip = 3.5 mm constant (Fig. A5.8).
Annexure-5 335
Fig. A5.5 a Internal strip (xed) = 3 mm and variation in external strip from (0 mm). b Internal
strip (xed) = 3 mm and variation in external strip from (7 mm). c Internal strip (xed) = 3 mm and
variation in external strip from (10.5 mm)
336 Annexure-5
Fig. A5.5 (continued)
Fig. A5.6 RDRA with sepa-

rated plates
Annexure-5 337
(a) Name X Y XY Plot 4 HFSSDesign1 ANSOFT

m1 10.2000 -17.1071 Curve Info
m2 14.6000 -15.6812 dB(S(waveport, waveport))
m3 16.8500 -16.0264 Setup6 : Sweep
m4 17.6500 -16.4270
m5 19.5000 -16.2692
-7.50
-10.00
-12.50
-15.00 m2
m3 m5
m4
m1
-17.50
10.00 12.00 14.00 16.00 18.00 20.00
Freq [GHz]
(b) Name X Y XY Plot 4 HFSSDesign1 ANSOFT

m1 10.4500 -17.9665 Curve Info
m2 14.3500 -16.3577 dB(S(w aveport,w aveport))
m3 16.8000 -16.7685 Setup6 : Sw eep
m4 17.5000 -16.4319
m5 19.5000 -21.9771
-7.50
-10.00
-12.50
-15.00
m2 m4
m3
m1
-17.50
-20.00
m5
-22.50
10.00 12.00 14.00 16.00 18.00 20.00
Freq [GHz]
(c) Name X Y XY Plot 4 HFSSDesign1 ANSOFT

m1 14.6000 -16.6991 Curve Info
m2 16.9500 -16.3861 dB(S(waveport, waveport))
m3 17.9000 -20.2745 Setup6 : Sweep
-5.00
-7.50
-10.00
-12.50
-15.00 m2
m1
-17.50
m3
-20.00
-22.50
10.00 12.00 14.00 16.00 18.00 20.00
Freq [GHz]
Fig. A5.7 a Separation width (0.5 mm) for er 1, external strip = 10.5 mm, inner strip = 3.5 mm.
b Separation width (1.5 mm) for er 1, external strip = 10.5 mm, inner strip = 3.5 mm.
c Separation width (2.5 mm) for er 1, external strip = 10.5 mm, inner strip = 3.5 mm
338 Annexure-5
(a) Name X Y
5.00 14.0000 -13.9564
m1
ANSOFT
Curve Info
m2 16.9000 -22.9742 dB (S(waveport,waveport))
m3 19.3500 -20.7517 Setup6 : Sweep
-0.00
dB (S(waveport,waveport))
-5.00
-10.00
m1
-15.00
-20.00 m3
m2
-25.00
10.00 12.00 14.00 16.00 18.00 20.00
Freq [GHz]
(b) Name X Y
-4.00 14.3500 -15.3507
m1
Curve Info
m3 19.5000 -16.1834 Setup6 : Sweep
-6.00
-8.00
-10.00
-12.00
-14.00
m1
m2 m3
-16.00
-18.00
10.00 12.00 14.00 16.00 18.00 20.00
Freq [GHz]
(c) Name X Y
-2.00 14.4500 -13.4533
m1 Curve Info
m3 Setup6 : Sweep
-4.00 17.1500 -13.6281
m4 17.6500 -14.9946
m5 19.5500 -16.0278
-6.00
-8.00
-10.00
m2
-12.00
m1 m3
-14.00
m4
m5
-16.00
-18.00
10.00 12.00 14.00 16.00 18.00 20.00
Freq [GHz]
Fig. A5.8 a Separation width (0.5) for er 2, external strip = 10.5 mm, inner strip = 3.5 mm.
b Separation width (1.5) for er 2, external strip = 10.5 mm, inner strip = 3.5 mm. c Separation
width (2.5) for er 2, external strip = 10.5 mm, inner strip = 3.5 mm
Annexure-5 339
A.5.2(d) Effect of Variable Capacitance (Varactor Diode)

in Between the Plate
The effect of the varactor diode capacitance placed in between the parallel standing
strips is seen. The resonant modes get shifted lower side (Fig. A5.9).
The separation width = 1.0, er 1, external strip = 10.6 mm, inner
strip = 3.0 mm, varactor diode (variation from 1 to 5 F with step of 1 F) at
position (z = 2.3) in vertical direction. The resulting effect is shown in Fig. A5.10.
A5.3 Designing Steps
HFSS steps_Project1
1. Open HFSS.
2. Create le name project1.
3. Dene in the Cartesian co-ordinate system origin as (x = 0, y = 0, z = 0).
4. Choose 3-D rectangular box for substrate by dening the desired substrate
material and its dimensions such as (RT Duroid and x = 20 mm, y = 30 mm,
z = 0.8 mm).
5. Create DRA structure with desired material and dimensions on the substrate top
surface (e.g., If substrate dimension from origin was 0.8 mm in z-direction.
Then choose DRA #d dimension keeping substrate dimension as reference).
Fig. A5.9 RDRA with lumped capacitance

340 Annexure-5
Name X Y XY Plot 11 HFSSDesign1 ANSOFT

0.00 10.9619 -26.4758
m1 Curve Info
m3 17.4549 -26.3304 Setup1 : Sweep
c='0pF'
-5.00 dB(S(waveport,waveport))
Setup1 : Sweep
c='1pF'
-10.00 Setup1 : Sweep
c='2pF'
Setup1 : Sweep
-15.00 c='3pF'
Setup1 : Sweep
c='4pF'
-20.00 dB(S(waveport,waveport))
m2 Setup1 : Sweep
c='5pF'
-25.00 m1 m3
-30.00
10.00 12.00 14.00 16.00 18.00 20.00
Freq [GHz]
Fig. A5.10 Variation of resonant frequency with lumped capacitance
6. Create two parallel strips adjacent to DRA above the substrate surface with
rectangular slab in between them keeping substrate dimension as reference.
7. Apply micro-strip feeding to the DRA structure by dening the micro-strip port
with appropriate length and width for impedance matching (e.g., wave port)
assigning in the desired direction of input excitation.
8. Variation in height of external strip keeping the internal strip height xed and
vice versa.
9. Effect of the permittivity of rectangular slab can be seen by varying the material
property and thickness of the slab in between two xed parallel plates.
10. Placing a lumped capacitor between two parallel standing strips with desired
value (e.g., 2 F) and perform parametric analysis for variable capacitance
value of lumped element.
11. Performing the simulation for the steps 8, 9, 10 mentioned above separately and
for mode analysis of DRA which give modal frequency response and effect of
the variation of radiation parameters associated with DRA and non-resonant
slab with parallel standing strip geometry.
12. Analysis of the simulated structure can be performed by taking various
response quantities such as S11, radiation pattern, gain, and eld distribution.
13. The above mechanism can also be validated in RDRA by VNA with anechoic
chamber on prototype model after structure is simulated.
Annexure-6
Cartesian, Cylindrical, and Spherical Coordinate System
There are three different coordinate systems, i.e., Cartesian, cylindrical, and
spherical systems. Cartesian are (x, y, z), cylindrical are q; /; z, and spherical are
r; h; / representation (Figs. A6.1 and A6.2).
(a) Cylindrical to cartesian (b) Cartesian to cylindrical

p
X q cos / q x2 y2
Y q sin y
tan1
Zz x
Zz
(c) Spherical to cartesian (d) Cartesian to cylindrical
p
X r sin h cos / r x2 y2 z2
Y r sin h sin / p!
x2 y2
Z r cos h h tan1
z
y
/ tan1
x
(e) Cylindrical to spherical (f) Spherical to cylindrical
p
r q2 z2 q r sin h

q //
h tan1
z z r cos h
//
1. DEL r derivation in cylindrical system:

The Cartesian r (Del) is given as follows:
~ ~ @ @ @
r x ~ y ~z
@x @y @z

342 Annexure-6
Fig. A6.1 Cartesian system
Fig. A6.2 Cylindrical

components
Cylindrical r (Del) is given below:
~ q @ ^ 1 @ ^z @
r ^ /
@q q @/ @z
Converting differential operators in terms of the cylindrical system by chain rule:

@ @ @q @ @/ @ @z
^x 6 6 ^x
@x @q @x @/ @x @z @x
@q
@ p 1 2x x
x2 y2 p p cos /
@x @x z x2 y2 x y2
2
p
) q x2 y2
p
q x2 y2
Hence,
x
cos /
q
@/ @ h 1 y 1 x0 y1 x2 y
tan 2
@x @x x 2
1 2
y x 2 x2 y2 x
x
@/ 1 y 1
p p sin /
@x x2 y2 x2 y2 q
@/
0 ) z is the same z as in Cartesian system it doesn't depend on x
@x
Annexure-6 343
As per chain rule

Thus, we have

@ @ @P @ @/ @ @z
^x ^x
@x @p @x @/ @x @z @x
A6:1
@ @ 1
^x cos / sin / 0
@q @/ p
Using the same technique to convert the differential for y:

@ @ @q @ @/ @ @z
^y ^y
@y @q @y @/ @y @z @y
@q @ p 2y y
x2 y2 p p sin /
@y @y 2 x y2 2 x y2
2
@/ @ 1 1 1 x 1
tan yx 2 cos /
@y @y y2 x
1 2 x y 2 q
x
@z
0
@y
Thus,

y @ @q @ @/ @ @z
^y ^y
@y @q @y @/ @y @z @y
A6:2
@ 1 @
^y sin / cos / 0
@q q @/
Finally, since z is not transformed between coordinate systems
@ @
A6:3
@z @z
~ ^x @ ^y @ ^z @
r
@x @y @z

~ ^x cos / @ 1 sin / @ ^y sin / @ 1 cos / @ ^z @
r
@q q @/ @q q @/ @z
Cylindrical
~ ^x cos / ^y sin / @ 1 ^y cos / ^x sin / @ ^z @

r A6:4
@q q @/ @z
344 Annexure-6
Hence, denition to cylindrical unit vector is given as follows:
^p ^x cos / ^y sin / ^p
^ ^x sin / ^y cos / /
/ ^
^z ^z ^z
Thus, Del cylindrical can be written as follows:
~ q @ 1^ @ @
r ^ / ^z
@q q @/ @z
which is the desired solution of r in cylindrical coordinates.

2. DEL r expression as spherical system (Figs. A6.3, A6.4 and A6.5):
Spherical to Cartesian Cartesian to Spherical

p
X r sin h cos / r x2 y2 z2
Y r sin h sin / p!
1 x2 y2
Z r cos h h tan
z
y
/ tan1
x
~ @ ^x @ ^y @ ^z
r A6:5
@x @y @z

@ @ @r @ @h @ @/
^x ^x A6:6
@x @r @x @h @x @/ @x
Fig. A6.3 Spherical system

Annexure-6 345
Fig A6.4 Spherical

components
Fig. A6.5 Spherical

subcomponents
@ @ @r @ @h @ @/
^y y A6:7
@y @r @y @h @y @/ @y

@ @ @r @ @h @ @/
^z z A6:8
@z @r @z @h @z @/ @z
346 Annexure-6
Now, partially differentiate r with respect to x
@r @ p
x2 y2 z 2
@x @x
2x
p
2 x y2 z 2
2
x A6:9
p
x2 y2 z 2
r sin h cos /

r
sin h cos /
Similarly partially differentiate r with respect to y
@r @ p
x2 y2 z 2
@y @y
2y
p
2 x2 y2 z 2
y
p A6:10
x y2 z 2
2
r sin h sin /

r
sin h sin /
Partially differentiate r with respect to z
@r @ p
x2 y2 z 2
@z @z
2z
p
2 x y2 z 2
2
z A6:11
p
x y2 z 2
2
r cos h

r
cos h
Annexure-6 347
Partially differentiate h with respect to x

p!
@h @ x2 y2
tan1
@x @x z
1 1 2x
x2 y2
p
1 z 2 x2 y2
z2
2
z x
p
x y z x y2
z2 2 2 2 A6:12
x
p
x2 y2 2
z x y2 z 2
r sin h cos /

r 2 tan h
cos h cos /

r
Partially differentiate h with respect to y

p!
@h @ x2 y2
tan1
@y @y z
1 1 2y
x2 y2
p
1 z 2 x2 y2
z2
2
z y
p
z 2 x2 y2 z x2 y2 A6:13
y
p
x y 2
2 2
z x y2 z 2
r sin h sin /

r 2 tan h
cos h sin /

r
348 Annexure-6
Partially differentiate h with respect to z

p!
@h @ x2 y2
tan1
@z @z z
p
1 x2 y2

1 x zy
2 2
z2
2
A6:14
p!
z x2 y2
2
x y z
2 2 z
sin h

r
Partially differentiate / with respect to x
@/ @ y
tan1
@x @x x
1 x0 y1
2
1 yx2 x2
x2 h y i A6:15
2 2
x y2 x
y
2
x y2
sin /

r sin h
Partially differentiate / with respect to y
@/ @ y
tan1
@y @y x

1 1
y2 x
1 x2 A6:16
x
2
x y2
cos /

r sin h
Annexure-6 349
Partially differentiate / with respect to z
@/ @ y
tan1 0 A6:17
@z @z x
Put Eqs. (A6.5), (A6.8), (A6.11) in Eq. (A6.2), Put Eqs. (A6.6), (A6.9), (A6.12)
in Eq. (A6.3) and Put Eqs. (A6.7), (A6.10), (A6.13) in Eq. (A6.4).

@ @ @ cos h cos / sin / @

^x ^x sin h cos / A6:18
@x @r @h r r sin h @/

@ @ @ cos h sin / cos / @

^y ^y sin h sin / A6:19
@y @r @h r r sin h @/

@ @ @ sin h
^z ^z cos h A6:20
@z @r @h r
Put Eqs. (A6.13), (A6.14) and (A6.15) in Eq. (A6.1).

And by using original denition to Spherical unit vector,
^r ^x sin h cos / ^y sin h sin / ^z cos h

^h ^x cos h cos / ^y cos h sin / ^z sin h
^ ^x sin / ^y cos /
/
We get
~ ^r @ ^h 1 @ /
r ^ 1 @
@r r @h r sin h @/
Fig. A6.6 E and H elds pattern in RDRA

350 Annexure-6
Fig. A6.7 Rectangular DRA
Table A6.1 Transcendental equation solution

S. Permittivity Dimension (a(length) b Resonant Effective Multiple %
No. (width) d(depth)) (mm) frequency width (b) factor change
in width
1. 10.0 14.3 25.4 26.1 3.5 34.22 1.3474 34.7381
2. 10.0 14 8 8 5.5 14.13 1.7665 76.6535
3. 10.0 15.24 3.1 7.62 6.21 8.33 2.8872 168.7230
4. 20.0 10.2 10.2 7.89 4.635 15.31 1.5014 50.1419
5. 20.0 10.16 10.2 7.11 4.71 15.15 1.4858 48.5797
6. 35.0 18 18 6 2.532 24.12 1.34 33.9973
7. 35.0 18 18 9 2.45 25.64 1.4244 42.4423
8. 100.0 10 10 1 7.97 11.24 1.1242 12.4237
~ from Cartesian to spherical converted.

Hence r
3. E and H elds in RDRA
Fields converting into TE and TM modes inside rectangular DRA (Fig. A6.6).
4. Transcendental equation solution using MATLAB programs (simulated
rectangular DRA) (Fig. A6.7; Table A6.1).
Annexure-6 351
%%Dimensions of DRA
%%length
d=[14.3,14.0,15.24,10.2,10.16,18,18,10];
%%width
w=[25.4,8,3.1,10.2,10.2,18,18,10];
%%height
h=[26.1,8,7.62,7.89,7.11,6,9,1];
%%Mode
m=1;
n=1;
p=1;
c=3e8;
cons=[10.0,10.0,10,20,20,35,35,100];
syms y real
for i=drange(1:8)
kx(i)=pi/d(i);
kz(i)=pi/2/h(i);
end
352 Annexure-6
m=1;
n=1;
p=1;
E_r=10;
a=15.24e-03;
b=3.1e-03;
d=7.62e-03;
c=3e+08;
kx=m*pi/a;
ky=n*pi/b;
kz=p*(pi/d)/2;
ko=sqrt(kx^2+ky^2+kz^2)/sqrt(E_r);
fo=(c*ko/pi)/2;
foghz=fo/(1e+09);
Annexure-6 353
MATLAB programs taking parameters a,b,d same and comparing frequency using :
Program 1 : Characteristic Equation
m=1
n=1
p=1
E_r=10
a=14.3e-03
b=25.4e-03
d=26.1e-03
c=3e+08
k_x=m*pi/a
k_y=n*pi/b
k_z=p*(pi/d)/2
f_o=(c*k_o/pi)/2
f_oGHz=f_o/1e+09
354 Annexure-6
m=1;
n=1;
p=1;
E_r=10;
a=14.3e-03;
b=25.4e-03;
d=26.1e-03;
c=3e+08;
syms y real
kx=pi/a;
kz=pi/d/2;
effwidth=pi/ky;
factor=effwidth/b;
Annexure-6 355
: Characteristic Equation
Where a=17mm
b=25mm
c=10mm
m=1;
n=1;
p=1;
E_r=10;
a=17e-03;
b=25e-03;
d=10e-03;
c=3e+08;
k_x=m*pi/a;
k_y=n*pi/b;
k_z=p*(pi/d)/2;
k_o=sqrt(k_x^2+k_y^2+k_z^2)/sqrt(E_r);
f_o=(c*k_o/pi)/2;
f_oGHz=f_o/1e+09;
356 Annexure-6
m=1;
n=1;
p=1;
E_r=10;
a=17e-03;
b=25e-03;
d=10e-03;
c=3e+08;
syms y real
kx=pi/a;
kz=pi/d/2;
effwidth=pi/ky;
factor=effwidth/b;
Annexure-6 357
Characteristic Equation
m=1
n=1
p=1
E_r=10
a=14.3e-03
b=25.4e-03
d=26.1e-03
c=3e+08
k_x=m*pi/a
k_y=n*pi/b
k_z=p*(pi/d)/2
f_o=(c*k_o/pi)/2
f_oGHz=f_o/1e+09
358 Annexure-6
m=1;
n=1;
p=1;
E_r=10;
a=14.3e-03;
b=25.4e-03;
d=26.1e-03;
c=3e+08;
syms y real
kx=pi/a;
kz=pi/d/2;
effwidth=pi/ky;
factor=effwidth/b;
Annexure-6 359
360 Annexure-6
Bibliography
1. Long SA, Mcallister MW, Shen LC (1983) The resonant cylindrical dielectric cavity antenna.
IEEE Trans Antennas Propag 31(3):406412
2. Kajfez D, Guillon P (eds) (1986) Dielectric resonators. Artech House, Dedham
3. Kranenburg RA, Long SA, Williams JT (1991) Coplanar wave guide excitation of resonator
antennas. IEEE Trans Antennas Propag 29(1):119122
4. Lee RQ, Simons RN (1994) Bandwidth of dielectric resonator antennas. In: IEEE antennas
propagate society of international symposium. Seattle, pp 15001503
5. Wu JY, Hung CY, Wong KL (1994) Low prole very high permittivity DRA excited by
coplanar waveguide. Microw Opt Technol Lett 22(2):9697
6. Yung EKN, Lee WWS, Luk KM (1994) A Dielectric resonator on a micro strip antennas. In:
IEEE antennas propagate society of international symposium serial. Washington, pp 1504
1507
7. Mongia RK, Luk KM, Yung EKN (1994) Low prole dielectric resonator antennas using a
very high permittivity material. Electron Lett 30(17):13621363
8. Oliver MB, Antar YMM, Mongia RK, Iittipiloon A (1995) Circularly polarized rectangular
dielectric resonator antenna. Electron Lett 31(6):418419
9. Pozar DM, Schubert DH (1995) Micro strip antennas: the analysis of micros trip antennas and
arrays. IEEE Press, New York
10. Keller MG, Roscone DJ, Oliver MB, Mongia RK, Antar YMM, Iittipiloon A (1995) Active
aperture coupled rectangular dielectric resonator antenna. IEEE Microw Guided Wave Lett 5
(11):376378
11. Esselle KP (1996) A low prole rectangular dielectric resonator antennas. Electron Lett 44
(9):12961297
12. Hwang Y, Zhang YP, Mulk K, Yung ENK (1997) Gain enhanced miniaturized rectangular
dielectric resonator antennas. Electron Lett 33(5):350352
13. Leung KW, To MW (1997) Slot coupled dielectric resonator antennas using a proximity feed
on a perpendicular substrate. Electron Lett 33(20):16651666
14. Hwang Y, Zhang YP, Luk KM, Yung EKN (1997) Gain-enhanced miniaturised rectangular
dielectric resonator antenna. IEEE Electron Lett 33(5):350352
15. Mongia RK, Ittipiboon A (1997) Theoretical and experimental investigation on rectangular
dielectric resonator antennas. IEEE Trans Antenna Propag 45:13481356
16. Mongia RK, Iittipiboon A (1997) Theoretical and experimental investigation on rectangular
dielectric resonator antennas. IEEE Trans Antennas Propag 45(9):13481356
17. Mongia RK, Ittipiboon A (1997) Theoretical and experimental investigations on rectangular
18. Junkar GP, Kishk AA, Glisson AW (1998) Multiport network description and radiation
characteristics of coupled dielectric resonator antennas. IEEE Trans Antennas Propag 46
(8):12421243

362 Bibliography
19. Kranenburg RA, Long SA (1998) Micro strip transmission line excitation of dielectric
resonator antennas. Electron Lett 24(18):11561157
20. Wu JY, Hung CY, Wong KL (1999) Low prole very high permittivity DRA excited by a
coplanar wave guide. Microw Opt Technol Lett 22(2):9697
21. Hong CS (1999) Adjustable frequency dielectric resonator antennas. Proc Natl Sci Counc.
ROC (A) 23(6):736738
22. Henry B, Petosa A, Antar YMM, Morin GA (1999) Mutal coupling. Opt Technol Lett 21
(1):4448
23. Luk KM, Lee MT, Leung KW, Yung EVN (1999) Technique for improving coupling between
micro strip line and dielectric resonator antennas. Electron Lett 35(5):357358
24. Sangiovanni A, Garel PY, Dauvigenc JY, Pichor Ch (2000) Numerical analysis of dielectric
resonator antennas. Int J Numer Model Electr Netw Devices Fields 13:199215
25. Leung KW (2000) Conformal strip excitation of dielectric resonator antennas. IEEE Trans
Antennas Propag 48(6):961967
26. Ge Y, Esselle KP (2000) Computation of radiation pattern of a rectangular dielectric resonator
antenna using the method of moments. Microw Opt Technol Lett 27(6):382384
27. Hang CY, Kuo JS (2000) Frequency adjustable circularly polarized dielectric resonator
antennas. Microw Opt Technol Lett 34(3):211213
28. Sangiovanni A, Dauvignac JY, Pichod C (2001) Embedded dielectric resonator antenna for
bandwidth enhancement. Electron Lett 33(7):406408
29. Laisne A, Gillard R, Piton G (2001) Robust slot fed dielectric resonator antennas using an
intermediate. Electron Lett 37(25):14971498
30. Jerry Fiedziuszko S, Home S (2001) Dielectric resonatorrise your high SQ. IEEE Microw
Mag 5160
31. Deng SM, Chen TW, Kan HH (2001) A CPW fed rectangular dielectric resonator antennas. In:
Proceedings of IEEE APMC, Taipei, pp 954956
32. Al Salmeh MS, Antar YMM, Seguin G (2002) Coplanar wave guide fed coupled rectangular
33. Fiedziuszko SJ, Hunter IC, Itoh T, Kobyashi Y, Nishokawa T, Suitzer SN, Wakino K (2002)
Dielectric materials, devices and circuits. IEEE Trans Microw Theory Tech 50(3):706720
34. Petosa A, Littipiboon A, Thirkoune S (2002) Perforated dielectric resonator antennas. Electron
Lett 38(24):14931495
35. Kishk AA, Yan Y, Glisson AW (2002) Conical dielectric resonator antennas for wide band
applications. IEEE Trans Antenna Propag 50(4):469474
36. So KK, Leung KW (2002) Annular-slot excited dielectric resonator antenna with a backing
cavity. IEEE Antenna Propagation Society international symposium, San Antonio, Texas,
June 2002, pp 470473
37. Ferrando-Bataller M, Valero-Nogueira A, Cabedo-Fabrs M, Antonino Daviu E (2003)
Design of ultra-wideband antennas using characteristic modes. INICA/COST 284 workshop,
Berlin, Germany
38. Li B, So KK, Leung KW (2003) A circular polarized dielectric resonator antenna excited by
asymmetrical U-slot with a backing cavity. IEEE Antenna Wirel Propag Lett. 2:133135
39. Guo YX, Luk KM (2003) Dual polarized dielectric resonator antennas. IEEE Trans Antennas
Propag 51(5):11201123
40. Kishk AA (2003) Wideband truncated tetrahedron dielectric resonator antenna excited by a
coaxal probe. IEEE Trans Antenna Propag 51(10):29132917
41. Gentilli GB, Morni M, Selleri S (2003) Relevance of coupling effects on DRA array design.
IEEE Antenna Propaga 51(3):399404
42. Kishk AA, Zhang X, Glission AW, Kajfez D (2003) Numerical analysis of stacked dielectric
resonator excited by a coaxial probe for wideband application. IEEE Trans Antenna Propaga
51(8):19962006
43. Deng SM, Tsai CL, Chiu CW, Chnag SF (2003) CPW fed rectangular ceramic dielectric
resonator antennas 51(8):21422144
Bibliography 363
44. Petosa A, Ittipiboon A, Antaar Y (2003) Broadband dielectric resonator antennas. In:
Dielectric resonator antennas. Research Studies Press Ltd., UK
45. Ghosh B, Antar YMM, Petosa A, Littipiboon A (2004) Feed conguration of CPW fed DRA.
IEEE Antennas Propagation Society international symposium, Monterrey, CA, June 2004
46. Bit-Babik G, DiNallo C, Faraone A (2004) Multimode dielectric resonator antenna of very
high permittivity. In: Proceedings of IEEE antennas and propagation symposium, Monterrey,
CA, June 2004, pp 13831386
47. Bit Babik B, Di Nallo C, Fraone A (2004) Multimode dielectric resonator antennas of very
high permittivity. IEEE Antennas Propagation Society international symposium, Monterrey,
CA, June 2004
48. Sung Y, Ahn CS, Kim YS (2004) Micro strip fed dual frequency dielectric resonator antennas.
Microw Opt Technol 42(5):388390
49. Littipiboon A, Mongia RK, Antar YMM, Bhartia P, Cuhaci M (2004) An integrated
rectangular dielectric resonator antenna. IEEE Antennas Propagation Society international
symposium, Michigan, June 2004, pp 604607
50. Petosa A, Thirkoune S, Lttipiboon A (2004) Array of performed dielectric resonator antennas.
IEEE Antennas Propagation Society international symposium, Monterrey, CA, June 2004
51. Li B, Leung KW (2004) A wideband strip fed rectangular dielectric resonator antenna. In:
Proceedings of IEEE antennas and propagation symposium, DC, Washington, July 2004,
pp 13831386
52. Sheng XQ, Leung KW, Yung EKN (2004) Analysis of waveguide fed dielectric resonator
antenna using a hybrid innite element method moment method. IEEE Proc Microw Antenna
Propag 151:9195
53. Eshrah IA, Kishk AA, Yakovlev AB, Gilsson AW (2004) Modeling of a waveguide probe
excitation for dielectric resonator antenna. IEEE Antennas Propagation Society international
symposium, Monterrey, CA, June 2004
54. Nannini C, Ribero JM, Dauvignac JY, Pichot C (2004) Bifrequency behavior and bandwidth
enhancement of a dielectric resonator antenna. Microw Opt Technol Lett 42(5):432434
55. Nannini C, Ribero JM, Dauvignac JY, Pichot C (2004) Bifrequency and bandwidth
enhancement of a dielectric resonator antennas. Microw Opt Technol Lett 42(5):432434
56. Joliveau E, Smith J, Wolfe J (2004) Tuning of vocal tract resonances by sopranos. Nature
427:116
57. Li B, Leung KW (2005) Strip-fed rectangular dielectric resonator antennas with/without a
parasitic patch. IEEE Trans Antenna Propag 53(7):22002207
58. de Young CS, Long SA (2006) Wideband cylindrical and rectangular dielectric resonator
antenna. IEEE Antennas Wirel Propag Lett 5:426429
59. Petosa A, Thirakoune S, Littipiboon A (2009) Higher order modes in rectangular DRAs for
gain enhancement. Presented at the ANTEM 2009, Banff, Canada, Feb 2009
60. Chang T-H, Kiang J-F (2009) Bandwidth broadening of dielectric resonator antenna by
merging adjacent bands. IEEE Trans Antennas Propag 57(10):33163320
61. Sebastian ML (2009) Materials for wireless communication. Elsevier, Amsterdam
62. Kong LB (2010) Electrically tunable materials. Elsevier, Amsterdam
63. Huff GH, Rolando DL, Walters P, McDonald J (2010) A frequency recongurable dielectric
resonator antenna using colloidal dispersions. IEEE Antenna Wirel Propag Lett 9
64. Huff GH, Rolando DL, Walters P, McDonald J (2010) A frequency recongurable dielectric
resonator antenna using colloidal dispersions. IEEE Antennas Wirel Propag Lett 9
65. Petosa A, Thirakoune S (2011) Rectangular dielectric resonator antennas with enhanced gain.
IEEE Trans Antennas Propag 59(4):13851389
66. Yaduvanshi RS (2012) Hybrid MHD antenna. Intersci J IJUWBCS 2(4):201206
67. Wang YF, Denidni TA, Zeng QS, Wei G (2014) IEEE Lett
68. Khan QU Higher mode excitation for high gain micro strip antenna. AIU-Int J Electron
Commun
364 Bibliography
69. Guha D, Antaar YMM (2012) Higher order mode excitation for high-gain broadside radiation
from cylindrical dielectric resonator antennas. IEEE Antenna Propag
70. Popov A (1999) Multi-mode dielectric resonator antenna with controllable radiation pattern.
IEEE Trans Antenna Propag
71. Kreyszig E (2002) Advanced engineering mathematics. Wiley for Half wave Fourier analysis
72. Jackson JD (1962) Classical electrodynamics. Wiley, New York
73. Cheng X (2011) Hybrid mode radiation in patch antenna loaded with corrugated electric
arrays. Department of Electrical and Computer Engineering, University of Florida,
Gainesville, FL, USA
74. Chang K, Hsieh LH (2004) Microwave ring circuits and related structures. Wiley Interscience,
New York
75. Adam JJ, Bernhard JT (2013) Broadband equivalent circuit models for antenna impedances
and elds using characteristic modes. IEEE Trans Antenna Propag
76. Jaggi VP, Mathur AB (1985) Advanced engineering mathematics. Khanna publisher (Chapter
13 Series solutions and special functions I.e. Bessel function)
77. Neshati MH, Wu Z (2001) The determination of the resonant frequency of the TEy111 mode in a
rectangular dielectric resonator for antenna application. 11th international
78. Colline BS, Nahar V, Kingsley SP, Zhang SQ, Krupa S (2004) A dual band hybrid dielectric
antenna for laptop computer. IEEE Antenna Propagation Society international symposium
79. Jordan EC, Balmain KG (1968) Electromagnetic and radiating systems. Wiley, New York
80. Landau LD, Liftshiz EM (1984) Electrodynamics of continuous media. Elsevier, Amsterdam
81. Jacksion JD (1962) Classical electrodynamics. Wiley, New York
82. Balanis CA (2005) Antenna theory analysis and design. Wiley, New York
83. Harrington RF (1961) Time harmonic electromagnetic elds. IEEE Press series on
electromagnetic wave theory
84. Sadiku MNO (2009) Principles of electromagnetics. Oxford University Press, Oxford
85. Feynman RP (2013) Lecture series on physics volume II. Pearson Publishing
86. Guha D, Antar YMM (2011) Microstrip and printed antennas. Wiley, New York
Index
A Even order modes, 154, 166

Antenna gain, 148, 149, 167, 229 Excitation angle, 126
Antenna recongurability, 230
Aperture coupled feed, 34, 193, 197 F
Field and pattern, 4, 6, 14, 21, 36, 119, 148,
B 165
Boundary conditions, 5, 6, 11, 14, 20, 22, Flux density, 144
2528, 3335, 46, 47, 50, 52, 57, 73, 107, Fourier transform, 9, 105, 118, 124, 125
111, 112, 116, 118, 121, 123, 128, 144, Frequency recongurability, 230
148, 181, 185, 212, 217, 218, 222, 225,
233, 239, 247 G
Gain of RDRA, 166
C
Cavity resonator, 107, 121, 123, 145 H
Characteristic equation, 12, 57, 59, 161 Helmholtz equation, 3, 1215, 20, 23, 25, 27,
Characteristic modes, 7, 12 52, 53, 107, 136, 137, 212, 213, 217, 222,
Charge conservation equations, 211 223, 225, 233
Cherenkov principle, 11 Higher modes, 14, 126, 148, 149, 154,
Circular polarization, 12, 126, 181, 213, 262, 166169, 213, 214, 229, 259
265, 269 Higher-order even modes, 126, 155, 166
Current density, 7, 12, 25, 27, 130, 135, 136, Higher-order odd modes, 126, 155
185, 191, 212, 213, 219, 243 Hybrid modes (HEM), 6, 14, 148, 155, 185,
214, 225, 230
D
Dielectric polarization, 11, 103 K
Dielectric resonator, 1, 2, 1113, 33, 165, 166, KAM (KalmogorovArnoldMoser) theory, 3,
213, 252 13, 24, 212, 214, 217, 225
Dipole moment, 8, 11, 14, 103, 123, 136, 149,
168, 185, 230 L
Directivity, 148, 168, 213 Longitudinal elds, 3, 33, 35, 123, 148
Discrete modes, 23, 53, 233 Lorentzs gauge condition, 118, 140
Dominant mode, 5, 59, 136, 148 Loss tangent, 4
E M
Eigen currents, 58, 148 Magnetic charge density, 106
Eigen values, 7, 166, 241, 242 Magnetic dipoles, 6, 126, 166, 167
Eigen vectors, 7, 12, 57, 148 Magnetic scalar potentials, 136
Electric charge density, 106 Magnetic vector potential, 12, 21, 118, 135,
Electric scalar potentials, 119, 136 138, 140143, 181, 186

366 Index
Magneto hydrodynamics (MHD), 310 Reactive power, 125

Maxwells equation, 11, 12, 14, 103, 106, 136, Rectangular cavity resonator, 107, 121
148, 212214, 221 Rectangular dielectric resonator antenna
MHD antenna, 310 (RDRA), 18, 12, 13, 24, 28, 29, 3336,
Miniaturization, 168 45, 54, 57, 79, 82, 87, 102, 103, 106, 112,
118, 121, 123, 126, 135, 137, 141, 144,
O 146, 148, 149, 154156, 159, 165171,
Odd order modes, 126, 154, 168, 229 193, 198, 199, 205, 207, 211214, 217,
Orthogonal polarization, 231 219, 225, 227, 230, 232, 251, 262, 269,
Orthogonal radiation, 5, 12, 154 279, 282
Orthogonality property, 24, 54, 126, 202, 212 Relative permittivity, 11, 103
Orthonormality, 6, 34, 35, 104, 105, 122, 185, Resonant cavity, 11, 183
213, 216 Resonant frequency, 1, 68, 11, 13, 22, 2931,
Orthonormality principle, 6, 13, 216, 225 33, 35, 57, 59, 72, 78, 82, 85, 87, 103, 126,
136, 141, 144, 146, 148, 156, 159, 167,
P 168, 193, 198, 203, 225, 229, 282, 284, 286
Parsevals power theorem, 12, 136 Resonant modes, 47, 1113, 21, 2729, 31,
Patch antenna, 1, 165, 166 3436, 81, 103, 106, 120, 125, 130,
Pattern recongurability, 230 148150, 155, 165168, 172, 174, 185,
Perfect electric conductors (PEC), 107, 118, 186, 198, 199, 207, 218, 220, 229
212 Resonating modes of RDRA, 221
Perfect magnetic conductors (PMC), 107, 211 Resonator current, 105
Permittivity, 2, 4, 13, 30, 87, 112, 149, 156, Rigorous theoretical analysis, 11
169, 285
Perturbation, 4, 13, 112, 148, 166, 212 S
Perturbed modes, 13 Smith chart, 195, 260, 276
Polarization, 168, 213 Surface current density, 12, 185, 191, 213
Polarization recongurability, 229
Poynting theorem, 13 T
Poynting vector, 12, 137, 140, 143, 181 Taylors expansion, 199, 204
Probe current, 3, 12, 104, 122, 125, 127, 146 Transcendental equation, 33, 35, 57, 59, 65, 72,
Propagation constant, 6, 7, 12, 14, 33, 35, 52, 78, 102, 158, 161, 225
57, 59, 121, 129, 136, 159, 161 Transverse electric (TE), 14, 35, 148
Propagation parameter, 109 Transverse magnetic (TM), 14, 148
Q U
Quality factor, 3, 12, 34, 144146 Unperturbed modes, 242
R V
Radiating lobes, 79, 80 Variance function, 199
Radiation pattern, 7, 12, 14, 34, 35, 80, 125, Vector potential, 12, 21, 118, 119, 135, 136,
135, 136, 138, 139, 141143, 152, 166, 141, 142, 243
168, 181, 185, 186, 188, 192, 193, 198,
213, 229, 243, 246, 251, 256, 269 W
Random variables, 199 Wave number, 6, 21, 22, 28, 57
RDRA mathematical modeling, 22, 36, 135 Wave vector, 5, 12, 21, 135
RDRA modes, 4, 5, 11, 103, 121, 123, 168
RDRA resonant modes, 4, 5, 7, 11, 12, 28, 34,
79, 103

Book - 2016 Rectangular Dielectric Resonator Antennas Theory and Design

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Book - 2016 Rectangular Dielectric Resonator Antennas Theory and Design

Caricato da

Copyright:

Formati disponibili

RajveerS.

ISBN 978-81-322-2499-0 ISBN 978-81-322-2500-3 (eBook)

Library of Congress Control Number: 2015942632

Springer New Delhi Heidelberg New York Dordrecht London

Printed on acid-free paper

Springer (India) Pvt. Ltd. is part of Springer Science+Business Media (www.springer.com)

Microwave dielectric resonator antenna (DRA) materials or ceramics were dem-

The discrete modes (mnp) enable us to visualize the resonator as collection of

where R is autocorrelation and contain components fcos; sing fcos; sing

is orthogonal to each function of the set:

{Cmnp, Dmnp} magnitude components can be determined based on principle of

E x; y; 0}; for different (m, n),

supported with practically implemented case studies. Devising control on resonant

(c) Surface current density is equated with generated magnetic elds

at z = 0; amplitude coefcients (Dmnp and Cmnp) are obtained on expansion of

vmnp cos cos cos Hz

Sidewalls and top walls all are perfect electric conductors

umnp sin sin cos Ez

vmnp sin sin sin Hz

Sidewalls and top walls all are perfect magnetic conductors

umnp cos cos sin Ez

vmnp sin sin cos Hz

umnp cos cos cos Ez

vmnp sin sin sin Hz

er k02 kx2 ky2 kz2

the nal result of transcendental equation is thus achieved.

New Delhi, India Rajveer S. Yaduvanshi

1 Rectangular DRA Fundamental Background . . . . . . . . . . . . . . . . 1

2 Rectangular DRA Resonant Modes and Sources. . . . . . . . . . . . . . 11

3 Mathematical Analysis of Rectangular DRA. . . . . . . . . . . . . . . . . 33

4 Mathematical Analysis of Transcendental Equation

5 Mathematical Analysis of RDRA Amplitude Coefficients . . . . . . . 103

6 Mathematical Analysis of Radiation Pattern of RDRA . . . ...... 135

7 Rectangular DRA Higher-Order Modes and Experimentations . . . 147

8 RDRA Angular Excitation Mathematical Model

9 Sensitivity Analysis of Rectangular DRA . . . . . . . . . . . . . . . . . . . 199

10 Hybrid Modes in RDRA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211

11 Inhomogeneous Permittivity, Permeability, and Conductivity

12 Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251

Dr. Rajveer S. Yaduvanshi is working as an associate professor in the

Abstract This chapter introduces rectangular dielectric resonator antenna.

Keywords RDRA (rectangular DRA)

Springer India 2016 1

1.2 History of DRA

Dielectric resonator antenna is a microwave antenna consists of block of ceramics

1.3 Working Mechanism of RDRA

The electromagnetic waves were generated by rapid oscillations of electrons in

Fig. 1.1 Ceramics rectangular DRA with a, b, and d dimensions

Fig. 1.2 RDRA equivalent

Fig. 1.3 RDRA with ground

1.4 Antenna Radiation Parameters

Antenna radiation parameters are as follows:

1.5 Advantage of RDRA

Lower conduction losses due to use of dielectric material;

1.6 Resonant Modes

time-average electrical energy is equated to magnetic energy at any instant of time

Fig. 1.4 Resonant modes in

1.7 Characterization of Resonant Modes

er k02 kx2 ky2 kz2 ; 1:4

Fig. 1.5 Electric elds and magnetic elds are associated

The solution resonant frequency of RDRA can be determined depending upon

Magnetic walls of RDRA:

This book contents are lucid, simple, and pedagogical.

1.8 Magnetic Dipole Moment

r2 A k2 A lJ rjxle jxele 2:6

Curl E r E jxlH B; t 2:8a

Curl H r H jxE J D; t 2:8b

Ez;y cEy jxlHx

cEx Hz;x jxEy

" mp cm;n np jxl

C1 expfcm; n; xzg C2 expfcm; n; xzg

r2 H r J e jxlr jxH; 2:33

umnz x; y; x sin hfcm; n; xz d g w.r.t. x; y; z: