Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
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
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
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
Re Dmnpejxmnpt / mnp
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
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
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
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
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
Top and bottom walls are perfect magnetic conductors, and all four sidewalls are
PEC
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
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.
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
xv
xvi Contents
3.2.3 Model-3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.2.4 Model-4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.2.5 Basic Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
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
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
xix
Chapter 1
Rectangular DRA Fundamental
Background
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.
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).
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
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.
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:
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
E tan n E 0; 1:6
H nor n H 0; 1:7
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
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.
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.
xt C1 e jxt
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
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
npc
x kc ; when n 1; 2; 3. . . where k x=c; 1:17
L
References
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
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
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
@D
rH J 2:1
@t
D E
B lH r A
1
H r A
l
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
r re 0
E jxA re
E jxA re
r r A r r A r 2 A 2:5
16 2 Rectangular DRA Resonant Modes and Sources
Now,
r lH rr A r2 A
or
lr H rr A r2 A
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
r A jxele
or
1
e r A
jxle
Hence, r2 A k2 A lJ:
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
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
@Hz
Hz;y
dy
@Hz
Hz;x
dx
Similarly,
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
Za Zb mpx npy
2
Eix m; n Eix x; y cos sin dxdy
ab a b
00 2:12a
np
mp
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
20 2 Rectangular DRA Resonant Modes and Sources
w0; y; t wa; y; t 0
wx; 0; t wx; b; t 0
Y 00 X 00
ky2 ; kx2 ;
Y X
2.5 Two Dimensional Resonator 21
@2w
c2 r 2 w 0 2:16
@t2
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
r2 Ax k2 Ax 0; 2:19
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
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.
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
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
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
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
Curl E jxl H;
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
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
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
^ Jsy x; y; xY:
J Jsx x; y; xX ^ 2:39
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)
r2 W k2 W 0;
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.
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:
30 2 Rectangular DRA Resonant Modes and Sources
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
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
Reference
1. Okaya A, Barash LF (1962) The dielectric microwave resonator. Proc IRE 50:20812092
Chapter 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.
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.
36 3 Mathematical Analysis of Rectangular DRA
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;
n E 0;
n H 0;
x 0; a;
Ey Ez 0;
Hx 0;
At,
y 0; b;
Ex Ez 0;
Hy 0;
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
wz X xY yZz
where
TE mode Ez 0 and H z 6 0
wHz X xY yZz
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
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
3.2 Rectangular DRA Mathematical Modeling 39
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
wEz X xY yZ z;
40 3 Mathematical Analysis of Rectangular DRA
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
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;
n H 0;
n E 0;
At,
x 0; a;
Hy Hz 0;
Ex 0;
At,
y 0; b;
Hx Hz 0;
Ey 0;
1 1 @ 2 Ez @Hz
Ey ; 3:11
2
jx 1 kc2 jxl @z@y @x
1 @Ez 1 @ 2 Hz
Hx ; 3:12
2
jxl 1 kc2 @y jx @z@x
42 3 Mathematical Analysis of Rectangular DRA
1 1 @ 2 Hz @Ez
Hy ; 3:13
jxl 1 kc2 jx @z@y @x
2
wz X xY yZz;
where
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
3.2 Rectangular DRA Mathematical Modeling 43
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
TM mode H z 0 and Ez 6 0
wEz X xY yZ z;
Ey C0 X xY 0 yZ 0 z;
Now
Y 0 y A3 cos ky y A4 sin ky y;
44 3 Mathematical Analysis of Rectangular DRA
At,
y 0; b; Ey 0;
0 A3 cos ky 0 A4 sin ky 0;
np
A3 0 and ky ;
b
and
mp
kx ;
a
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
3.2 Rectangular DRA Mathematical Modeling 45
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
@Ez @Ey
jxlHx ; 3:19
@y @z
@Ex @Ez
jxlHy ; 3:20
@z @x
@Ey @Ex
jxlHz ; 3:21
@x @y
@Hz @Hy
jxEx ; 3:22
@y @z
46 3 Mathematical Analysis of Rectangular DRA
@Hx @Hz
jxEy ; 3:23
@z @x
@Hy @Hx
jxEz ; 3:24
@x @y
w X xY yZ z;
TM mode of propagation, H z 0;
Boundary conditions
Electrical walls ! Etan 0 n E;
! Hn 0 n H;
At, x = 0;
@
Z z 0;
@z
A5 coskz z A6 sinkz z 0;
Therefore, at
z 0; d;
A5 must be zero;
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
48 3 Mathematical Analysis of Rectangular DRA
Ez 0;
1
A1 A3 A5 ky
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
For, TE mode
For PEC walls, electric eld components are assumed to be varying with Hz in
direction of (x, y, z)
3.2 Rectangular DRA Mathematical Modeling 49
@
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
50 3 Mathematical Analysis of Rectangular DRA
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.
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
3.2 Rectangular DRA Mathematical Modeling 51
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
Similarly,
@Ez
Hy C 00 ;
@x
C 00 X 0 xY y0 Z z;
X 0 x A1 cos kx x A2 sin kx x;
) Hy 0 at x 0; a;
) A1 0;
mp
kx ;
a
52 3 Mathematical Analysis of Rectangular DRA
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
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.
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
54 3 Mathematical Analysis of Rectangular DRA
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
3.2 Rectangular DRA Mathematical Modeling 55
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
Keywords Mathematical analysis Transcendental equation Rectangular DRA
Propagation constant Eigen vectors Effective electrical length Characteristic
equation
Fig. 4.1 a Rectangular DRA. b Eigen currents (current vectors) versus wavelength
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
kx mp=a 4:3a
ky np=b 4:3b
kz pp=d 4:3c
0 k0 2 kx 2 ky 2 kz 0
2
4:5
kz pp=d
k z 0 kz 2 0 k0 0 r k0 2
2 2
kz 02 kz 2 0 l0 x2 r l0 x2
kz 02 kz 2 x2 l1 r 4:6
60 4 Mathematical Analysis of Transcendental
) nH 0
And
nE 0
or,
Hy Hx 0
Ez 0
) nE 0
And
nH 0
At
x 0; a Ey Ez 0
Hx 0
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
jx 1 kc2 jxl @z@y @x
1 @Ez 1 @ 2 Hz
Hx 4:7c
2
jxl 1 kc2 @y jx @z@x
1 1 @ 2 Hz @Ez
Hy 4:7d
jxl 1 kc2 jx @z@y @x
2
wz X xY yZz
where
wHz X xY yZ z
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
4.1 Case-1: Top and Bottom Walls as PMC and Rest of the Four Walls are PEC 63
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
Hz A2 A4 A6 cos x cos y cos z 4:9
a b d
and
Ez 0
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
C1 e jkz z C2 ejkz z 0
C1 C2 0
4:13
or; C1 C2
@Hz @Hz0
Hz Hz0 and
@x @x
and
mp np
Hz0 B cos x cos y cos kz0 z 4:14b
a b
or,
0
C1 e jkz z C2 ejkz z C20 ejkz z 4:15
4.1 Case-1: Top and Bottom Walls as PMC and Rest of the Four Walls are PEC 65
or,
0
2kz C 1 coskz d kz0 C 02 ejkz z :
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
kz
tankz d q 4:18b
k02 r 1 kz2
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
mp
kx 4:20a
a
66 4 Mathematical Analysis of Transcendental
np
ky 4:20b
b
pp
kz 4:20c
d
) nH 0
And
nE 0
or,
Hy Hx 0
Ez 0
) 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
jx 1 kc2 @y jxl @z@x
1 1 @ 2 Ez @Hz
Ey 4:21b
2
jx 1 kc2 jxl @z@y @x
1 @Ez 1 @ 2 Hz
Hx 4:21c
2
jxl 1 kc2 @y jx @z@x
1 1 @ 2 Hz @Ez
Hy 4:21d
jxl 1 kc2 jx @z@y @x
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
wHz X xY yZz
Ez 0
we get
@Hz
0
Ey C
@x
68 4 Mathematical Analysis of Transcendental
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
Hz A2 A4 A6 cos x cos y cos z 4:23
a b d
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
kx2 ky2
Hz0 coskx x cosky y cosk 0z z
jxl0
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
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,
or,
or,
j tan kz d 1
0
kz kz
or,
kz
j tan kz d
kz0
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
kz0 6 pp=d
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,
74 4 Mathematical Analysis of Transcendental
Ey Ex 0
Hz 0
) 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
jx 1 kc2 @y jxl @z@x
1 1 @ 2 Ez @Hz
Ey 4:31b
2
jx 1 kc2 jxl @z@y @x
1 @Ez 1 @ 2 Hz
Hx 4:31c
2
jxl 1 kc2 @y jx @z@x
1 1 @ 2 Hz @Ez
Hy 4:31d
jxl 1 kc2 jx @z@y @x
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
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
76 4 Mathematical Analysis of Transcendental
Ex ; Ey 0
and
Hz 0
or,
0
C1 e jkz z C2 ejkz z C20 ejkz z 4:34
C1 e jkz z C2 ejkz z 0
C1 C2 0
4.2 Case-2 77
or,
C1 C2 4:35
@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
or,
0
C1 e jkz z C2 ejkz z C20 ejkz z
or,
0
2jkz C1 sinkz d kz0 C 02 ejkz z 4:38
Squaring both sides and substituting the value of kz02 from Eq. (4.3c), we get
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;
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---------->>>
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);
slice(xi,yi,zi,Ez,xslice,yslice,zslice)
colormap hsv
4.3 MATLAB Simulation Results 81
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
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
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
r 100
a 10 mm
b 10 mm
fr 7:97 GHz
4.3 MATLAB Simulation Results 83
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
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
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
r 10
a 14 mm
b 8 mm
d 8 mm
86 4 Mathematical Analysis of Transcendental
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
4.3 MATLAB Simulation Results 87
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
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).
88 4 Mathematical Analysis of Transcendental
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@)
17. Strontium titanate 100.0 10 10 1 7.97 7.7587
(ECCOSTOCK@)
18. Strontium titanate 100.0 12.7 12.7 1 7.72 7.6628
(ECCOSTOCK@)
19. Strontium titanate 100.0 5 10 1 8.85 8.1828
(ECCOSTOCK@)
(continued)
4.4 Resonant Frequency of RDRA for Experimentations 89
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
4.4 Resonant Frequency of RDRA for Experimentations 91
%%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
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;
92 4 Mathematical Analysis of Transcendental
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
Results:
4.4 Resonant Frequency of RDRA for Experimentations 93
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:
94 4 Mathematical Analysis of Transcendental
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:
4.4 Resonant Frequency of RDRA for Experimentations 95
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:
96 4 Mathematical Analysis of Transcendental
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:
4.4 Resonant Frequency of RDRA for Experimentations 97
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;
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;
98 4 Mathematical Analysis of Transcendental
Program 7
Output:
4.4 Resonant Frequency of RDRA for Experimentations 99
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;
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:
100 4 Mathematical Analysis of Transcendental
Program 9
4.4 Resonant Frequency of RDRA for Experimentations 101
Program 10
102 4 Mathematical Analysis of Transcendental
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.
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
Re Dmnpejxmnpt / mnp
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
Re Dmnpejxmnpt / mnp
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
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
106 5 Mathematical Analysis of RDRA Amplitude Coefcients
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
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.
rr E jxlr H r M;
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
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,
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 ;
108 5 Mathematical Analysis of RDRA Amplitude Coefcients
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,
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
5.3 RDRA Maxwells Equation-Based Solution 109
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
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
ppz
d 1
mn z p fzmnp sin
pp d 2d d
r 5:14
1 d ppz
fzmnp sin
ppd 2 d
5.3 RDRA Maxwells Equation-Based Solution 111
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
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
Ex 0 where x 0; a or z 0; d;
Ey 0 where y 0; b or z 0; d;
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
112 5 Mathematical Analysis of RDRA Amplitude Coefcients
cy m; n k 2 h2 m; nja; d1=2
The equation
r H J jxE
gives
We assume that J on the walls is zero. Then, the boundary conditions yields
r E jxlH
r H jxeE
0 x aW
0 y bL
0 z dh
5.4 RDRA Inhomogeneous Permittivity and Permeability 113
E E x; yecz 5:19
H H x; yecz 5:20
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
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? H ? j w Ez^z 5:28
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
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
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
5.4 RDRA Inhomogeneous Permittivity and Permeability 115
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
v ve vm
vm vm
2
h2 h0
ck 2 ck2
jxlh 2 jxlh20
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
116 5 Mathematical Analysis of RDRA Amplitude Coefcients
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
5.4 RDRA Inhomogeneous Permittivity and Permeability 117
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
jpp
c ; p 1; 2; 3:
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
118 5 Mathematical Analysis of RDRA Amplitude Coefcients
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.
d
a
b
5.5 RDRA with Probe Current Excitation 119
l2 a a l2 b b
\jx j\ ; \jy j\
2 2 2 2 2 2
^ z l^Idl
@A @ ejkr l^idl cos h jk cos h jkr jx ^
e 2/
@z 4p @z r 4p r2 r c
^ w.r.t. z
Differentiating /
^
@/ l^idlc2 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
^ jxl^Idl l^Idlc2 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:
120 5 Mathematical Analysis of RDRA Amplitude Coefcients
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
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
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
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
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
npy ppz
where umnp x; y; z p2 sin mpx
3=2
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
124 5 Mathematical Analysis of RDRA Amplitude Coefcients
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
@/ 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,
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
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
1
XL jxL; XC
jxC
128 5 Mathematical Analysis of RDRA Amplitude Coefcients
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
r2 Hz c2 xHz
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
r2 c2 x Ez 0
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
5:55
X
r2 ~c2 x Hz Jz mnp; xumnp x; y; z
mnp
X
Hz f Hz mnp; xumnp x; y; z;
mnp
130 5 Mathematical Analysis of RDRA Amplitude Coefcients
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
Z a Z b Zd
Jy n; m; p; x JSy x; y; xd0 z d0 vmnp x; y; zdxdydz
0 0 0
^z r? Ez ^zc ^z E? ^z jxl H ?
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
On differentiating
di2 di 1
L R 0
d2 t dt c
di2 R di 1
2
0
d t L dt LC
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
Let
R
k
s2L
2 2
R 1
x2
2L Lc
S1 k x 2
S2 k x 1
I A1 es1 t Az es2 t
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
1 1
It q es1 t q es2 t
L L2 LC
R 4
L L2 LC
R 4
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
L s2 R s C1 0
134 5 Mathematical Analysis of RDRA Amplitude Coefcients
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
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
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
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
!
lI dl ejkr !
A; where A is magnetic vector potential 6:5
4pr
From Helmholtz equation ~
A
E jx ~
A
138 6 Mathematical Analysis of Radiation Pattern of RDRA
~ 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
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 :
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
!
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
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
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
142 6 Mathematical Analysis of Radiation Pattern of RDRA
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
CI0 ejkr
Az Ph0 ; 0 ; 6:19
r
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
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
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
2xW x
Qx ;
jI xj2 Rr x
(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
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
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
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
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
Lumped port
W (slot)
Fig. 7.10 a H eld distribution inside RDRA and b E eld distribution inside RDRA
152 7 Rectangular DRA Higher-Order Modes and Experimentations
-4.00 0.00
dB(GainTotal)
-6.00
dB(S(1,1))
-10.00
Name X Y
-10.00 dB(GainTotal)
Setup1 : LastAdaptive
Freq='12GHz' Phi='0deg'
-30.00 dB(GainTotal)
-12.00
Setup1 : LastAdaptive
Freq='12GHz' Phi='90deg'
-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
-15.00 dB(GainTotal)
-14.00 Setup1 : LastAdaptive
Freq='10GHz' Phi='0deg'
-16.00 -20.00 dB(GainTotal)
Setup1 : LastAdaptive
Freq='10GHz' Phi='90deg'
-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
Freq [GHz] Theta [deg]
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,
7.1 Introduction to Higher Modes 153
(a) (b)
-5.00
5.00
dB(GainTotal)
dB(S(1,1))
-7.50
0.00
-10.00
-5.00
Name X Y
dB(GainTotal)
Setup1 : LastAdaptive
-15.00 -15.00 Freq='12GHz' Phi='0deg'
dB(GainTotal)
Setup1 : LastAdaptive
Freq='12GHz' Phi='90deg'
-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
Freq [GHz] Theta [deg]
Fig. 7.13 TE113 mode at frequency 12 GHz. a H eld distribution, b E eld distribution, c return
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
-20.00 dB(GainTotal)
Setup1 : LastAdaptive
Freq='15GHz' Phi='0deg'
-25.00 dB(GainTotal)
Setup1 : LastAdaptive
Freq='15GHz' Phi='90deg'
-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
154 7 Rectangular DRA Higher-Order Modes and Experimentations
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
7.1 Introduction to Higher Modes 155
(a) (b)
0.00
-10.00
dB(GainTotal)
dB(S(1,1))
-10.00
-20.00
Name X Y
dB(GainTotal)
-30.00 Setup1 : LastAdaptive
-30.00 Freq='10GHz' Phi='0deg'
dB(GainTotal)
Setup1 : LastAdaptive
Freq='10GHz' Phi='90deg'
-40.00 -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. 7.17 TE113 mode at frequency 10 GHz. a H eld distribution, b E eld distribution, c return
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.
156 7 Rectangular DRA Higher-Order Modes and Experimentations
(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)
Setup1 : LastAdaptive
-15.00 Freq='15GHz' Phi='0deg'
dB(GainTotal)
Setup1 : LastAdaptive
Freq='15GHz' Phi='90deg'
-20.00
-200.00 -100.00 0.00 100.00 200.00
Theta [deg]
Fig. 7.18 TE115 mode at frequency 12 GHz. a H eld distribution, b E eld distribution, c gain
(a) (b)
Freq='12GHz' Phi='90deg'
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. 7.19 TE117 mode at frequency 15 GHz. a H eld distribution, b E eld distribution, c gain
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]
158 7 Rectangular DRA Higher-Order Modes and Experimentations
(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
dB(GainTotal)
-30.00 -20.00 Setup1 : LastAdaptive
Freq='9.5GHz' Phi='0deg'
-25.00 dB(GainTotal)
Setup1 : LastAdaptive
Freq='9.5GHz' Phi='90deg'
-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
Freq [GHz] Theta [deg]
Fig. 7.21 TE115 mode at frequency 9.5 GHz. a H eld distribution, b E eld distribution, c return
loss, d gain
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
(a) (b)
m1
(c) 10.00
5.00
0.00
dB(GainTotal)
-5.00
-10.00
Name X Y
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:
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
(a) (b)
m1 m2 m1 -30.0000 5.7498
5.00 m2 30.0000 6.1212
-0.00
dB(GainTotal)
-5.00
-10.00
Curve Info
-15.00 dB(GainTotal)
Setup1 : LastAdaptive
Freq='12.55GHz' Phi='0deg'
-20.00 dB(GainTotal)
Setup1 : LastAdaptive
Freq='12.55GHz' Phi='90deg'
-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
-10.00
-20.00
-30.00
-40.00
-200.00 -100.00 0.00 100.00 200.00
Theta [deg]
7.2 Resonant Frequency and RDRA Structure 161
r
m2 n2 p2
c
fmnp p ; 7:9
2 er a b d
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
7:10
where
166 7 Rectangular DRA Higher-Order Modes and Experimentations
" #
amnp cos/mnp bmnp sin/mnp
1
wmnp tan ; Phase: 7:13
amnp sin/mnp bmnp cos/mnp
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.
168 7 Rectangular DRA Higher-Order Modes and Experimentations
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.
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.
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.
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.
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.
170 7 Rectangular DRA Higher-Order Modes and Experimentations
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]
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.
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
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.
172 7 Rectangular DRA Higher-Order Modes and Experimentations
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
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.
7.13 Effect of a/b and d/b Aspect Ratio 173
m2
-10.00
m3
-15.00
S11 [dB]
-20.00 m1
-25.00
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
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]
7.13 Effect of a/b and d/b Aspect Ratio 177
Fig. 7.41 Structure of the antenna after insertion of moat inside RDRA and eld distribution at
frequency 4.56 GHz
178 7 Rectangular DRA Higher-Order Modes and Experimentations
-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
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]
7.13 Effect of a/b and d/b Aspect Ratio 179
Reference
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
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.
8.1 Introduction 183
3.20 2.40
-90 90 -90 90
3.00 3.00
-90 90 -90 90
-180 -180
HFSSDesign1 HFSSDesign1
Radiation Pattern 8 Curve Info Radiation Pattern 8 Curve Info
0 rETotal 0 rETotal
Setup1 : Sweep Setup1 : Sweep
-30 30 Freq='16.64328657GHz' Phi='0deg' -30 30 Freq='16.64328657GHz' Phi='0deg'
14.00 rETotal 14.00 rETotal
Setup1 : Sweep Setup1 : Sweep
Freq='16.64328657GHz' Phi='90deg' Freq='16.64328657GHz' Phi='90deg'
10.50 10.50
-60 60 -60 60
7.00 7.00
3.50 3.50
-90 90 -90 90
2.80 2.80
-90 90 -90 90
3.00 3.00
-90 90 -90 90
3.00 3.50
-90 90 -90 90
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
8.1 Introduction 185
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.
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 ;
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
186 8 RDRA Angular Excitation Mathematical Model
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
Eh jxAh and Ah ^h A
Eh ; E/ ; H/ Eh =g; Hh E/ g:
8.2 Angular Shift in Excitation 187
1 2
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
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
188 8 RDRA Angular Excitation Mathematical Model
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
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
^
~ 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
^n H Js
jpp
c ; for all wave guide and
d
d d
let c ; jx ; for all cavity resonator
@z dt
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
190 8 RDRA Angular Excitation Mathematical Model
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
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
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
192 8 RDRA Angular Excitation Mathematical Model
Jsx x; y; t H y jz0
X n o
2p2 mpx
npy
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
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. 8.10 S11 RDRA with shifted slot resonant frequency 3.67
8.6 Experimentations 195
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
d d dxmnpja; b; d
; ;
dd db da
b
a
Fig. 9.7 RDRA having three different heights for increasing resonant modes
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
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
XY Plot 1 HFSSDesign1
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]
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
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
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
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.
206 9 Sensitivity Analysis of Rectangular DRA
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
See below Figs. 9.10, 9.11, 9.12, 9.13 and Table 9.2.
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
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.
10.1 Introduction
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
10.1 Introduction 213
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
214 10 Hybrid Modes in RDRA
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
with duality:
^
n H Js surface current density on walls;
Hsx x; y; 0 Jsy x; y
Hsy x; y; 0 Jsx x; y:
Hence,
X
jxmnpt
Jsx x; y; dt CmnpwH
mnpy x; y; 0e
mnp
X 10:7
jxmnpt
DmnpuH
mnpy x; y; 0e :
mnp
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
10.2 Mathematical Model 217
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 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
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
218 10 Hybrid Modes in RDRA
@ 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:
@ 2 Ez
0; z 0; d;
@x@z
@Ez
0; z 0; d:
@z
where r x; y; z
X X
H hom x; y; z; t C mnpejxmnpt wH
mnp
r Dmnpejxmnpt wH
mnp
r;
10:13
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
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
c2 Jsy;x x; y Jsx;y x; y
Similarly, we can compute Cxs Cys terms which are the desired solutions of
hybrid resonant modes.
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.
n H Js on each wall
n E Ms on each wall
jxl
r? Ez cE ? r? Hz ^z jxE ? 10:23
c
2
jxl x l
r? E z r? Hz ^z c E?
c c
c2 x2 l h2
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:
X
Ez Re Cmnpejxmnpt umnp r 10:29
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,
Hence,
pp
bmn
d
mpx npy ppz
Hz sin sin sin 10:32
a b d
224 10 Hybrid Modes in RDRA
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
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
226 10 Hybrid Modes in RDRA
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
10.7 General Solution of Hybrid Modes (HEM) 227
and,
0 jlxmnp
1
@
@y umnp
2
B hmn C
mnp r
WH @ jlxmnp A
@
@x umnp
h2mn
0
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:
^ 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
10.7 General Solution of Hybrid Modes (HEM) 229
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,
230 10 Hybrid Modes in RDRA
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.
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
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
-120 120
-150 150
-180
HFSSDesign1
Name X Y XY Plot 4 Curve Info
80.00
m1 70.0000 4.3036 dB(AxialRatioValue)
Setup1 : Sweep
Freq='13.34170854GHz' Phi='0deg'
70.00
dB(AxialRatioValue)
Setup1 : Sweep
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]
Keywords Inhomogeneous Permittivity Permeability Linear combinations
Sequential RLC circuits Discrete modes Complex solution
11.1 Introduction
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
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
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,
236 11 Inhomogeneous Permittivity, Permeability, and Conductivity
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
ve vm v1 x; y; ve vm v2 x; y 11:7c
@ @ @ @
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
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
and likewise,
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,
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 ;
11.2 Mathematical Model 239
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
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
D kw0 0; k h20
240 11 Inhomogeneous Permittivity, Permeability, and Conductivity
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 ;
11.2 Mathematical Model 241
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
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
n 2 D2
0\x\a; 0\y\b;
To get
0
D kmn w1 k1 wmn
0 0
V1 kmn 0
wmn 0
A 1 A
x k
B B
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.
11.2 Mathematical Model 243
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
where
r; h0 ; /0 r; h0 ; /0 ^r
Ad t; r Ad t; r; h; / r
constant
1
T I1 x21 I2 x22 I3 x23
2
where
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:
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
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
ZT
_ /;
S n /; /; h; h;
_ h; w; w;
_ w dt
0
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 :
248 11 Inhomogeneous Permittivity, Permeability, and Conductivity
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
11.2 Mathematical Model 249
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
250 11 Inhomogeneous Permittivity, Permeability, and Conductivity
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
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).
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. 12.4 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
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. 12.13 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
12.2 RDRA with ManganeseManganese Material as Dielectric 259
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. 12.20 Prototype RDRA under test setup for measurements with VNA
12.2 RDRA with ManganeseManganese Material as Dielectric 261
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).
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
0 Curve Info
rETotal
-30 30 Setup1 : Sweep
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
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]
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
Name X Y
XY Plot 4 HFSSDesign1 ANSOFT
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]
-10.00
-90 90
-120 120
-150 150
-180
1.00
-90 90
-120 120
-150 150
-180
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
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
30mm
20mm
-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) -------->>
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
-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)------->>>
-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) ----->>
12.3 Dual-Feed RDRA Hardware and Measurements 273
0.50
-90 90
-120 120
-150 150
-180
0.40 m2
-90 90
-120 120
-150 150
-180
2.00
-90 90
-120 120
-150 150
-180
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
12.3 Dual-Feed RDRA Hardware and Measurements 275
Fig. 12.52 Top and side view of single- and double-feed aperture couple feed of RDRA
Fig. 12.60 Fabricated model of RDRA under test with single feed and slot
Fig. 12.63 Top view with single feed and SMA connector of fabricated 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).
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
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]
Gain plot is shown in Fig. 12.66. It shows that isolated DRA has 4.7 dB gain at
resonant frequency.
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
Freq='3.99749499GHz' Phi='90deg'
-60 60
-20.00
-30.00
m3
-90 90
-120 120
-150 150
-180
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]
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).
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.
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.
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]
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
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
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]
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).
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]
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).
-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]
Return loss of antenna 1 is shown in Fig. 12.75. Resonant frequency is 6.89 GHz
with return loss 24 dB.
12.4 Isolated and Grounded RDRA 287
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]
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
^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,
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
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
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
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):
Mode sketch
Rectangular design
2 1
xmn pp
m n2 2
le a2 b2 , in two-dimensional case
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
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
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
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:
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
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
Divergence r A;
curl r A
@ @ @
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
@x 1 @x @x
rx uq u/ uz
@q q @/ @z
1 @ 1 @A/ @Az
rA qAq
q @q q @/ @z
@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
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
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.
Springer India 2016 309
R.S. Yaduvanshi and H. Parthasarathy, Rectangular Dielectric
Resonator Antennas, DOI 10.1007/978-81-322-2500-3
310 Annexure-3
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
Maxwells equations
@H
r E l
@t
@E
rH j
@t
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
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
r2 w k 2 w 0
W X xY yZ z
2
1 dX 1 d2 Y 1 d2 Z
k2 0
X dx2 Y dy2 Z dz2
TM mode
Xn mpx npyo
Ez Dmn sin sin ejkz z ; Dmn Fourier Coefficients A4:9
mn
a b
" #
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
#
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
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
Half-wave Fourier analysis will have odd or even terms, i.e., sinesine or
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
X
1 pnx
f x Bn sin
n1
a
where
Za np
2
Bn f x sin x dx
a a
0
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
fmnr
C mnr
xmnp d2 xmnr2
fmnr
dxmnp xmnr
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
Bejxt
d2x0
320 Annexure-4
@2 1 @2
ux; t 0; at boundaries
@x2 c2 dt2
^uL; 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
x=a
Annexure-4 321
w0; y; t wa; y; t 0
wx; 0; t wx; b; t 0
Y 00 X 00
kY2 ; kX2 ;
Y X
@2w
c2 r 2 w 0
@t2
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
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
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
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
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
Half-wave Fourier analysis will have odd or even terms, i.e., sinesine or
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
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
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
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:
Z1
E jxtj2 dt; in time domain
1
Z1
1
jXxj2 dx; in frequency domain
2p
1
Z1 Z1
1 2 1
E jxtj dt jXxj2 dx
Rr 2pRr
1 1
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
E sin h ejn1kd sin h
n1
Annexure-5
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.
(a)
-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)
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
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
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.
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
m4 17.6500 -16.4270
m5 19.5000 -16.2692
dB(S(waveport,waveport))
-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]
m4 17.5000 -16.4319
m5 19.5000 -21.9771
dB(S(waveport,waveport))
-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]
-5.00
dB(S(waveport,waveport))
-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
XY Plot 4 HFSSDesign1
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
XY Plot 4 HFSSDesign1 ANSOFT
Curve Info
m2 17.7000 -15.9854 dB (S(waveport,waveport))
m3 19.5000 -16.1834 Setup6 : Sweep
-6.00
dB (S(waveport,waveport))
-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
XY Plot 4 HFSSDesign1 ANSOFT
m1 Curve Info
m2 16.7500 -12.0819 dB (S(waveport,waveport))
m3 Setup6 : Sweep
-4.00 17.1500 -13.6281
m4 17.6500 -14.9946
m5 19.5500 -16.0278
-6.00
dB (S(waveport,waveport))
-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
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.
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).
c='1pF'
dB(S(waveport,waveport))
-10.00 Setup1 : Sweep
c='2pF'
dB(S(waveport,waveport))
Setup1 : Sweep
-15.00 c='3pF'
dB(S(waveport,waveport))
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]
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
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).
~ ~ @ @ @
r x ~ y ~z
@x @y @z
~ q @ ^ 1 @ ^z @
r ^ /
@q q @/ @z
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
@/ @ 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 @/
@ @
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
^p ^x cos / ^y sin / ^p
^ ^x sin / ^y cos / /
/ ^
^z ^z ^z
~ q @ 1^ @ @
r ^ / ^z
@q q @/ @z
~ @ ^x @ ^y @ ^z
r A6:5
@x @y @z
@ @ @r @ @h @ @/
^x ^x A6:6
@x @r @x @h @x @/ @x
@ @ @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
@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 /
@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 /
@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
z x y2 z 2
r sin h sin /
r 2 tan h
cos h sin /
r
348 Annexure-6
@/ @ 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
@/ @ y
tan1
@y @y x
1 1
y2 x
1 x2 A6:16
x
2
x y2
cos /
r sin h
Annexure-6 349
@/ @ 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).
@ @ @ sin h
^z ^z cos h A6:20
@z @r @h r
We get
~ ^r @ ^h 1 @ /
r ^ 1 @
@r r @h r sin h @/
%%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);
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
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 :
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
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;
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;
Annexure-6 355
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;
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;
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;
Annexure-6 357
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
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
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;
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;
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
dielectric resonator antennas. IEEE Trans Antennas Propag 45(9):13481356
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
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
dielectric resonator antennas. IEEE Trans Antennas Propag 50(10):114151419
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
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
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