Development of Dielectric

Resonator Antenna (DRA)

K. W. Leung
State Key Laboratory of Millimeter Waves &
Department of Electronic Engineering,
City University of Hong Kong

Presented to the IEEE LI Section Antenna and
Propagation Society on Monday October 8th 2012
I. Introduction

II. Circularly Polarized DRA Using a Parasitic Strip

III. Frequency Tuning Technique

IV. Omnidirectional Circularly Polarized DRAs

V. Dualband & Wideband DRAs

VI. Dualfunction DRAs









Outline
3

The DRA is an antenna that makes use of a radiating mode
of a dielectric resonator (DR).

It is a 3-dimensional device of any shape,
e.g., hemispherical, cylindrical, rectangular,
triangular, etc.

Resonance frequency determined by the its dimensions and
dielectric constant cr.
What is Dielectric Resonator Antenna (DRA) ?

4
Some DRAs :
5
Advantages of the DRA

Low cost
Low loss (no conductor loss)
Small size and light weight
Reasonable bandwidth (~10% for cr ~10)
Easy of excitation
High radiation efficiency ( generally > 95%)

6
Excitation schemes
(i) Microstrip line feed
7
Ground plane
Microstrip
feed line
Feed
Substrate
DRA
Slot
(ii) Aperture-couple feed
Excitation schemes
8
(iii) Coaxial feed
Excitation schemes
9
Bottom view Top view
Coaxial feed
10
Bottom view Top view
Aperture-coupled feed
11
Slot-fed DRA array using corporate
microstrip feed network
Corporate feedline for DRA array
12
Conformal-Strip Method
Rectangular Dielectric
Resonator Antennas
14
Proposed Antenna Geometry
W1
l1
a
d
b
x
y
z
Ground
Plane
Rectangular
DRA
Coaxial
Aperture
Conducting
Strip
c
r
a
(mm)
b
(mm)
d
(mm)
l
1
(mm)

W
1
(mm)

c
r
14.3 25.4 26.1 10 1 9.8
Resonant frequency of TE
mnl
(y) mode
Analytical Solution
Dielectric Waveguide Model (DWM)
2 2 2
0
2
z y x
r
k k k
c
f + + =
c t
d
l
k
b
n
k
a
m
k
z y x
2
, ,
t t t
= = =
2
0
2 2 2
k k k k
r z y x
c = + +
16
Numerical Solution
Advantages
- Very simple
- High modeling capability for general EM structures
- No spurious modes nor large matrix manipulation
- Provide a very wideband frequency response
Finite-Difference Time-Domain (FDTD) method
Disadvantages
- Time consuming, powerful computer required
17
Baseband Gaussian pulse
] ) 3 ( exp[
2 2
T T n t E
z
A =
T : pulse width
| |
| |
0
0
11
) (
) (
Z Z
Z Z
S
t I FFT
t V FFT
Z
in
in
in
+

=
=
Conformal Strip
Ground Plane
Ez
(V)
Hy Hy
Hx
Hx
(I)
Source occupies only one grid
Source model and extraction of S parameters
18
Uniform Cartesian grids
T = 0.083ns, t
0
= 3T
10-cell-thick PML with polynomial spatial scaling
(m = 4 and
max
= 1)
total grid size : 80x 110y 112z
total time steps : 10000
x = 0.715 mm, y = 0.508 mm, z = 0.5 mm
Parameters
19
3 3.5 4 4.5 5 5.5 6
-100
-50
0
50
100
150
200
Frequency (GHz)
2 2.5 3 3.5 4 4.5 5 5.5 6
-30
-25
-20
-15
-10
-5
0
Experiment
Theory
Frequency (GHz)
|
S
1
1
|
(
d
B
)
I
n
p
u
t

I
m
p
e
d
a
n
c
e

(
o
h
m
)
Experiment
Theory
Resistance
Reactance
Input Impedance/S
11
Reasonable agreement.
Wide Bandwidth of ~ 43%.
Dual resonant TE
111
y
and TE
113
y
modes are excited.
20
y
111
TE
y
113
TE
Resonant
Modes
Measured
resonant
frequencies
Calculated resonant
frequencies (FDTD)
Predicted resonant
frequencies (DWM)
f
mea
(GHz)
f
FDTD

(GHz)
error
(%)
f
DWM

(GHz)
error
(%)
3.81 3.90 2.3 3.95 3.6
N/A N/A N/A 4.26 N/A
4.57 4.60 0.7 4.7 1.7
Comparison between Theory and Measurement
y
112
TE
Reasonable agreement.
21
Field Distribution --- TE
111
y

Electric field
Magnetic field
z
2d
x
b
x-z
b
x
y
a
x-y
Imaged DRA (gound plane removed)
With gound plane
22
Field Distribution --- TE
112
y

x-z
b
x
y
a
x-y
Electric field
Magnetic field
z
2d
x
b
Imaged DRA (gound plane removed)
23
Field Distribution --- TE
113
y

x-z
Electric field
Magnetic field
b
x
y
a
x-y
z
2d
x
b
With gound plane
Imaged DRA (gound plane removed)
24
E (xz) - plane H (yz) - plane
(+x) (-x) (-y) (+y)
-40
0
o
-30-20-10 0
30
o
30
o
60
o
60
o
90
o
90
o
-40 -30-20-10 0
0
o
30
o
30
o
60
o
60
o
90
o
90
o
E (xz) - plane H (yz) - plane
(+x) (-y) (-x) (+y)
-40 -30-20-10 0
0
o
30
o
30
o
60
o
60
o
90
o
90
o
-40 -30-20-10 0
0
o
30
o
30
o
60
o
60
o
90
o
90
o
f = 3.5 GHz
f = 4.3 GHz
Radiation Patterns

Broadside radiation patterns are observed.
Measured E-plane crosspolarized fields mainly caused by finite
ground plane diffraction.
III. Circularly Polarized Design
using a Parasitic Strip

26
Proposed Antenna Geometry
a
(mm)
b
(mm)
d
(mm)
l
1
(mm)

W
1
(mm)

l
2
(mm)

W
2
(mm)

|
0
(degree)
c
r
24 23.5 12.34 10 1 12 1 225.6 9.5
W1
l1
a
d
b
x
y
z
Ground
Plane
Rectangular
DRA
Coaxial
Aperture
Conducting
Strip
c
r
C
W2
l2
Parasitic
Patch
(Center of
Parasitic Patch)
| 0
Top
View
Parasitic
Patch
Conducting
Strip
x
y
c
r
27
Input Impedance/S
11
Reasonable agreement.
Bandwidth ~ 14%.
Two nearly-degenerate TE
111
(y) modes are excited.
CP operation
2.5 3 3.5 4 4.5
-25
-20
-15
-10
-5
0
|
S
1
1
|

(
d
B
)
Frequency (GHz)
28
Axial Ratio in the boresight direction

3-dB AR bandwidth is ~ 2.7%, which is a typical value for
a singly-fed CP DRA.
3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8
0
5
10
15
20
A
x
i
a
l

R
a
t
i
o

(
d
B
)
Frequency (GHz)
Experiment
Theory
The H field of the DRA without and with parasitic
strip (Top view)
Without parasitic strip - LP field
With parasitic strip - CP field
3.4 GHz 3.4 GHz
Feeding strip Feeding strip
Parasitic
strip
29
30
Radiation Patterns (f = 3.4GHz, )

LHCP
RHCP
xz plane yz plane
(+x) (-x) (-y) (+y)
-40 -30-20-10 0
0
o
30
o
30
o
60
o
60
o
90
o
90
o
-40 -30-20-10 0
0
o
30
o
30
o
60
o
60
o
90
o
90
o
A broadside radiation mode is observed.
For each radiation plane, the LHCP field is more than 20dB
stronger than the RHCP field.
The maximum gain is 5.7 dBic (not shown here).
31
Effects of feeding strip length l
1
Input impedance changes substantially with l
1
.
AR is almost unchanged for different l
1
.
l
1
can be adjusted to match the impedance without changing AR.
2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4 4.2
-20
-15
-10
-5
0
|
S
1
1
|

(
d
B
)
Frequency (GHz)
l1 = 8 mm
l1 = 10 mm
l1 = 12 mm
3.1 3.2 3.3 3.4 3.5 3.6 3.7
0
5
10
15
20
A
x
i
a
l

R
a
t
i
o

(
d
B
)
Frequency (GHz)
l1 = 8 mm
l1 = 10 mm
l1 = 12 mm
II. Frequency Tuning Technique

The DRA for a paticular frequency may not be
available from the comericial market.

Fabrication tolerances cause errors between
measured and calculated resonant frequencies.

Frequency tuning methods:
(i) loading-disk; and
(ii) parasitic slot.
Backgruond
Frequency Tuning Technique
- using a loading disk

Side view Top view
The slot-coupled DRA with a conducting loading
cap
d
z
u
Slot
Hemispherical DRA
Dielectric
substrate crs
microstrip
line
Lt
Ground
plane
a
cra
x
Conducting
Loading Cap
o
Lt
x
y
|
Ls
Ws
Hemispherical
DRA
microstrip
line
Slot
Wf
Conducting
Loading Cap
Hemispherical DRA: radius a = 12.5 mm, dielectric constant
r
= 9.5.
Coupling slot : length L
s
, width W
s
Open-circuit stub: length L
t
Grounded dielectric slab:
rs
= 2.33, height d = 1.57 mm
Microstrip feedline: width W
f
= 4.7 mm
Calculated and measured return losses
(L
s
= 12 mm and W
s
= 1 mm)
3 3.2 3.4 3.6 3.8
-70
-60
-50
-40
-30
-20
-10
0
Frequency (GHz)
|
S
1
1
|

(
d
B
)
o = 26.38
Lt = 4.42 mm
o
o = 52.8
Lt = 13.6 mm
o
Theory
Experiments
Without cap (o = 0 )
Lt = 10.63mm
o
Resonance frequency:
3.52 GHz without any conducting cap ( = 0
0
), with L
t
= 4.42 mm
3.25 GHz ( = 26.38
o
and L
t
= 4.42 mm)
3.68 GHz ( = 52.8
o
and L
t
= 13.6 mm)
Calculated and measured radiation patterns
-40 -30 -20 -10 0
0
30
60
90
30
60
90
o o
o o
o o
o
(a) (b)
-40 -30 -20 -10 0
0
30
60
90
30
60
90
o o
o o
o o
o
Co-pol
X-pol
-40 -30 -20 -10 0
0
30
60
90
30
60
90
o o
o o
o o
o
(c)
-40 -30 -20 -10 0
0
30
60
90
30
60
90
o o
o o
o o
o
(d)
Co-pol
X-pol
3.58 GHz ( = 52.8
o
and L
t
= 13.6 mm)
Reasonable agreement
between theory and
experiment.

The effect of loading
cap on field pattern is not
significant.
3.25 GHz ( = 26.38
o
and L
t
= 4.42 mm)
Calculated and L
t
for having a good return loss
(minimum |S
11
| < -20dB)
3.2 3.3 3.4 3.5 3.6 3.7
0
10
20
30
40
50
60
2
4
6
8
10
12
14
16
Tuning frequency fr (GHz)
A
n
g
l
e

o

(
d
e
g
r
e
e
)
S
t
u
b

L
e
n
g
t
h

L
t
(
m
m
)
The resonant frequency can be tuned by varying and L
t
decreases from 26.38
o
to 0
o
(3.25 < f
r
< 3.5 GHz )
increases from 0
o
to 52.8
o
(3.5 < f
r
< 3.78 GHz)
Impedance bandwidth
The bandwidth decreases after a loading cap is added.
3.2 3.3 3.4 3.5 3.6 3.7
2
4
6
8
10
Frequency (GHz)
I
m
p
e
d
a
n
c
e

B
a
n
d
w
i
d
t
h

(
%
)
Frequency Tuning Technique
- using a parasitic slot

(a) Side view
(b) Top view
The annular-slot-excited cavity-backed DRA
z
x
Hemispherical DRA
Coaxial cable
Parasitic slot
a
c
r
metallic cavity
b
Ground plane
W
B
W
A
Feeding slot
r
B
r
A
Coaxial cable
Hemispherical DRA
Feeding slot
a
c
r
metallic cavity
b
Parasitic slot
A
|
r
A
r
B
W
B
W
A
y
x
IV. Omnidirectional Circularly
Polarized DRA

43
CP DRAs concentrated on broadside-mode designs only.
Provide larger coverage.
Advantages of omnidirectional CP antenna
44
Slotted omnidirectional CP DRA
Design I:
45
Antenna configurations
h
w
d
a
b
Dielectric
block

x
h
g
z
u
Slot
SMA connector
w
Probe
l
1
2r
Perspective view
Front view
Dielectric cube with oblique slots (polarizer) fabricated on
its four sidewalls.

Centrally fed by a coaxial probe extended from a SMA
connector, whose flange used as the small ground plane.
Wave polarizer
h
w
d
a
b
Dielectric
block

LP omnidirectional DRA
Dielectric block
P
o
l
a
r
i
z
e
r
Dielectric block with the wave
polarizer
Proposed compact omnidirectional CP DRA
x
y
z
Dielectric
slabs
D
E
+
Antenna principle
47
Prototype for 2.4 GHz WLAN design
Top face and sidewalls Bottom face

Design parameters
c
r
= 15, a = b = 39.4 mm, h = 33.4 mm, w = 9.4 mm,
d =14.4 mm, r
1
= 0.63 mm, l = 12.4 mm, g = 12.7 mm
Photographs of the prototype
48
Simulated and measured results
Reflection coefficient


Impedance bandwidth: AR bandwidth:
Simulated: 20.3% (2.34-2.87 GHz) Simulated: 8.2% (2.34-2.54 GHz)
Measured: 24.4% (2.30-2.94 GHz) Measured: 7.3% (2.39-2.57 GHz)

Axial ratio
2 2.2 2.4 2.6 2.8 3
-30
-20
-10
0
Frequency (GHz)
|S11| (dB)
HFSS Simulation
Experiment
Axial Ratio (dB)
HFSS Simulation
Experiment
2 2.2 2.4 2.6 2.8 3
0
3
6
9
Frequency (GHz)
49
Very good omnidirectional characteristic

In the horizontal plane, LHCP fields > RHCP fields by
~20 dB .
Simulated and measured radiation patterns
LHCP
-40 -30 -20 -10 0
30
150
60
120
90 90
120
60
150
30
180
o
o o
o
o o
o
o
o
o
o
dB
xz-plane
0
o
u =
RHCP
(+x) (-x)
-40 -30 -20 -10 0
30
210
60
240
90 270
120
300
150
330
180
o
o
o
o
o
o
o
o
o
o
o
dB
xy-plane
0
o
| =
RHCP
50
Simulated and measured antenna gain
2 2.2 2.4 2.6 2.8 3
-3
-2
-1
0
1
2
3
Frequency (GHz)
Gain (dBic)
HFSS Simulation
Experiment
51
Wideband omnidirectional CP antenna
with parasitic metallic strips
Design II:
52
Perspective view Front view
x
h
g
z
u

ls
w
s
Slot
Strip
SMA connector
w
Probe
Hollow circular
cylinder
l
1
2r

Four parasitic metallic strips are embedded in the
lateral slots to enhance the AR bandwidth.

The hollow circular cylinder is introduced to enhance
the impedance bandwidth.
Antenna configurations
53
Photographs of the prototype

Top face and sidewalls
Bottom face
Design parameters
c
r
= 15, a = b = 30 mm, h = 25 mm, r = 3 mm, w =7 mm, d =10.5 mm
l
s
= 30.5 mm, w
s
= 1 mm, x
0
= 6.4 mm, r
1
= 0.63 mm, l = 19 mm.
Prototype for 3.4 GHz WiMAX design
54

Overlapping bandwidth: 22.0%; bandwidth widened by ~3 times.
Simulated and measured reflection
coefficient and axial ratio
Axial Ratio (dB)
2.8 3.2 3.6 4
-30
-20
-10
0
Frequency (GHz)
|S11| (dB)
HFSS Simulation
Experiment
2.8 3.2 3.6 4
0
6
12
18
Frequency (GHz)
Impedance bandwidth:
Simulated: 22.3% (3.11-3.89 GHz)
Measured: 24.5% (3.08-3.94 GHz)

AR bandwidth:
Simulated: 24.8% (3.11-3.99 GHz)
Measured: 25.4% (3.16-4.08 GHz)
55
Measured gain: wider bandwidth.

Measured antenna efficiency: 84-98% (3.1-3.9 GHz).
Antenna gain
Simulated and measured results
Radiation efficiency
Gain (dBic)
HFSS Simulation
Experiment
2.8 3.2 3.6 4
-4
-3
-2
-1
0
1
2
3
Frequency (GHz)
2.8 3.2 3.6 4
0
0.2
0.4
0.6
0.8
1
Frequency (GHz)
Efficiency
56
3.4 GHz
3.8GHz
LHCP fields > RHCP fields by more than 15 dB in horizontal plane.
Stable radiation patterns across the entire passband (3.1 3.9 GHz).

LHCP
-40 -30 -20 -10 0
30
210
60
240
90 270
120
300
150
330
180
o
o o
o
o o
o
o
o o
o
dB
xy-plane
0
o
| =
RHCP
-40 -30 -20 -10 0
30
150
60
120
90 90
120
60
150
30
180
o o
o o
o o
o o
o o
o
dB
xz-plane
0
o
u =
RHCP
(+x) (-x)
-40 -30 -20 -10 0
30
150
60
120
90 90
120
60
150
30
180
o
o o
o
o o
o
o
o o
o
dB
xz-plane
0
o
u =
LHCP
RHCP
(+x) (-x)
-40 -30 -20 -10 0
30
210
60
240
90 270
120
300
150
330
180
o
o o
o
o o
o
o
o o
o
dB
xy-plane
0
o
| =
RHCP
Simulated and measured radiation patterns
V. Dualband & Wideband DRAs

(i) Rectangular DRA

59
Dualband and wideband antennas are extensively used
(e.g., WLAN)
Multi-element DRA [1]
- requiring more DR elements and space
Hybrid slot-DRA [2]
- coupling slot used as the feed and antenna
- inflexible in matching the impedance
[1] Petosa, N. Simons, R. Siushansian, A. Ittipiboon and C. Michel, Design and analysis of
multisegmentdielectric resonator antennas, IEEE Trans. AP, vol.48, pp.738-742, 2000.
[2] Buerkle, K. Sarabandi, and H. Mosallaei, Compact slot and dielectric resonator antenna with
dual-resonance, broadband characteristics, IEEE Trans. AP , vol. 53, pp.1020-1027, 1983.

Background

60


Wideband DRA [1]

Dualband DRA [2]

Trial-and-error approach is normally used

Systematic design approach is desirable
[1] B. Li and K. W. Leung, Strip-fed rectangular dielectric resonator antennas with/without a
parasitic patch, IEEE Trans. Antennas Propagat., vol.53, pp.2200-2207, Jul.2005.
[2] T. H. Chang and J. F. Kiang, Dual-band split dielectric resonator antenna, IEEE Trans.
Antennas Propagat., vol.55, no.11, pp.3155-3162, Nov.2007.

Use of higher-order DRA
mode
61
Design Formulas for Dual-Mode rectangular DRA
Electric field
Magnetic field
z
2d
x
b
x-z
b
x
y
a
x-y
x-z
b
x
y
a
x-y
Electric field
Magnetic field
z
2d
x
b
x-z
Electric field
Magnetic field
b
x
y
a
x-y
z
2d
x
b
y
TE
111
y
TE
112
y
TE
113
The E-field should vanish on the PEC and the TE
112
mode
cannot be excited properly.
The TE
111
mode and TE
113
mode are used in the dual-
mode design.
62
x
y
z
b
a
d
Ground plane
Rectangular DRA (c )
r
Formula Derivation
The wavenumbers k
x1, x2
and
k
z1, z2
can be written as follows:
a
k k
x x
t
= =
2 1
d
k
z
2
1
t
=
d
k
z
2
3
2
t
=
From the DWM model, the frequencies f
1
, f
2
are given by:
2
2 , 1
2
2 , 1
2
2 , 1 2 , 1
2
z z y y x x
r
k k k
c
f + + =
c t
2
2 , 1
2
2 , 1
2
2 , 1 2 , 1 z z x x y y
k k k k =
in which are wavenumbers in the
dielectric, with c being the speed of light in vacuum.
where
c f k
r
/ 2
2 , 1 2 , 1
c t =
(*)
64
d
k k
d A +

=
2
1
2
2
2
t
) 96 . 3 (
2
2
2
1
1
2
32 . 10
9
32 . 10
f
f
k k
a

+

=
2 1
35 . 0 65 . 0 b b b + =
Engineering Formulas for the DRA dimensions

|
|
.
|

\
|
+
|
|
.
|

\
|

|
|
.
|

\
|
= A
2
1
2
3
1
2
4
1
2
4422 . 11 3209 . 2 1393 . 0
f
f
f
f
f
f
d
3
1
2
10 4437 . 18 4984 . 23

(
(

(
+
|
|
.
|

\
|

f
f
(m)
1
1
1 tan
2
2
2 , 1
2 , 1 1
2 , 1
2 , 1

|
|
.
|

\
|
|
|
.
|

\
|
=

y y r y y
k
k
k
b
c
where
65
) 96 . 3 (
2
2
2
1
1
2
32 . 10
9
32 . 10
f
f
k k
a

+

=
Limit of frequency ratio f
2
/f
1
From
3k
1
> k
2
or 3f
1
> f
2


We have
giving
f
2
/f
1
< 3
which is the theoretical limit that is not known before.
> 0 9
2
2
2
1
d k k
66
1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8
0
1
2
3
4
5 =
r
c
10 =
r
c
30 =
r
c
70 =
r
c
Error of (%)
1
f '
f f
2 1
/
Compared with DWM results, errors of f
1
, f
2
are both
less than 2.5% for 1 < f
2
/f
1
2.8 , 5
r
70.
f
1
kept constant at 2.4 GHz.
Error analysis
67
A. Example for Dual-band Rectangular DRA Design
a = 20.8 mm, b = 10.5 mm, and d = 18.5 mm.
Given: f
1
= 3.47 GHz (WiMAX)
f
2
= 5.2 GHz (WLAN),
r
=10
Using dual-mode
formulas
68
Configuration of the dualband DRA
b
a
d
Ground plane
Microstripline
Coupling
slot
LS
W
L
Wf
h
Substrate
(c ) rs
Rectangular
DRA (c )
r
z
y
x
W = 2.6 mm, L =10.6 mm, Ls=7.2 mm, W
f
=1.94 mm,
h=0.762mm,
rs
= 2.93
69
Measured and simulated reflection coefficients
Measured bandwidths:
Lower band: 15% (3.25-3.78 GHz) covering WiMAX (3.4-3.7 GHz).
Upper band: 8.3% (5.03-5.47 GHz) covering WLAN (5.15-5.35 GHz).

3.2 3.6 4 4.4 4.8 5.2 5.6
-40
-30
-20
-10
0
HFSS Simulation
Measurement
Frequency (GHz)
Reflection coefficient |S | (dB)
11
70
COMPARISON OF DESIGN, SIMULATED, AND MEASURED
RESONANCE FREQUENCIES OF TE
111
y
AND TE
113
y
MODES
Resonant
Mode
Measured
frequency
(GHz)
Design
frequency
Simulated HFSS
frequency
f
1,2

(GHz)
Error
(%)
f
HFSS

(GHz)
Error
(%)
TE
111
y
3.40 3.47 2.05 3.47 2.05
TE
113
y

5.18 5.30 2.32 5.24 1.15
71
TE
111
y
mode: measured (3.40 GHz), simulated (3.47 GHz).
Broadside radiation patterns are observed for both planes.
Co-polarized fields > cross-polarized fields by more than 20 dB in
the boresight direction.
Measured and simulated radiation patterns
-40 -30 -20 -10 0
30
150
60
120
90 90
120
60
150
30
180
0
o
o
o o
o
o o
o
o
o
o
o
dB
90 90
150
180
o
o
o
o
o
dB
-40 -30 -20 -10 0
30
150
60
90
120
60
150
30
180
0
o
o o
o o
o o
o
o
o o
o
dB
(a)
Simulation
H-plane (y-z plane)
Measurement
E-plane (x-z plane)
-x
120
Co-pol
Cross-pol
Co-pol
Cross-pol
+x -y
+y
72
Measured and simulated radiation patterns
TE
113
y
mode: measured (5.18 GHz), simulated (5.24 GHz).
Broadside radiation patterns are observed for both planes.
Co-polarized fields > cross-polarized fields by more than 20 dB in
the boresight direction.
-40 -30 -20 -10 0
30
150
60
120
90 90
120
60
150
30
180
0
o
o
o o
o
o o
o
o
o
o
o
dB
90 90
150
180
o
o
o
o
o
dB
-40 -30 -20 -10 0
30
150
60
90
120
60
150
30
0
o
o o
o o
o o
o o
o o
o
dB
(b)
120
180
o
180
o
Simulation
H-plane (y-z plane)
Measurement
E-plane (x-z plane)
-x +x -y
+y
Co-pol
Co-pol
Cross-pol Cross-pol
73
TE
111
y
mode: Maximum gain of 4.02 dBi at 3.48 GHz.
TE
113
y
mode: Maximum gain of 7.52 dBi at 5.13 GHz.
Electrically larger antenna has a higher antenna gain.
Measured antenna gain
3 3.5 4 4.5 5 5.5 6
-5
0
5
10
Gain (dBi)
Frequency (GHz)
74
B. Example for Wideband DRA Design
a = 30.7 mm, b = 24.7 mm, and d = 47.7 mm.
Given: f
1
= 1.98 GHz (PCS)
f
2
= 2.48 GHz (WLAN),
r
=10
Using formulas for
dual-mode
rectangular DRA
75
Configuration of the wideband DRA
b
Conducting
feeding strip
Coaxial
aperture
a
d
W
l
Ground plane
Rectangular
DRA
(c )
r
x
y
z
l = 17 mm, W = 1 mm
76
Measured and simulated reflection coefficients
Measured bandwidths : 30.9% (1.83-2.50 GHz)
PCS (1.85-1.99 GHz), UMTS (1.99-2.20 GHz)
& WLAN (2.4-2.48 GHz)
1.5 2 2.5 3
-40
-30
-20
-10
0
HFSS Simulation
Measurement
Reflection coefficient |S | (dB)
11
Frequency (GHz)
77
Measured and simulated radiation patterns

Measured (2.16 GHz), simulated (2.11 GHz).
Broadside radiation patterns are observed.
Co-polarized fields > cross-polarized fields
by more than 20 dB in the boresight direction.
-40 -30 -20 -10 0
30
150
60
120
90 90
120
60
150
30
180
0
o
o
o o
o
o o
o
o
o
o
o
dB
90 90
150
180
o
o
o
o
o
dB
-40 -30 -20 -10 0
30
150
60
90
120
60
150
30
180
0
o
o o
o o
o o
o o
o o
o
dB
120
Simulation
H-plane (y-z plane)
Measurement
E-plane (x-z plane)
-x +x -y
+y
Co-pol
Co-pol
Cross-pol
Cross-pol
(a)
78
Measured and simulated radiation patterns

Measured (2.41 GHz), simulated (2.46 GHz).
Broadside radiation patterns are observed.
Co-polarized fields > cross-polarized fields by more
than 20 dB in the boresight direction.
-40 -30 -20 -10 0
30
-150
60
-120
90 -90
120
-60
150
-30
180
0
o
o
o o
o
o o
o
o
o
o
o
dB
90 -90
150
180
o
o
o
o
o
dB
-40 -30 -20 -10 0
30
-150
60
90
120
-60
150
-30
180
0
o
o o
o o
o o
o
o
o o
o
dB
H-plane (y-z plane)
Measurement
E-plane (x-z plane)
Simulation
Co-pol Co-pol
Cross-pol Cross-pol
(b)
-x +x
-y
+y
79
Measured antenna gain
The maximum gain of 6.98 dBi at 2.47GHz.
TE
113
y
-mode gain > TE
111
y
-mode gain.

1.6 1.8 2 2.2 2.4 2.6 2.8
0
2
4
6
8
Frequency (GHz)
Gain (dBi)
(ii) Cylindrical DRA

81
2
0
2 2
i r zi i
k k k c

= +
Ground plane
z
a
Cylindrical
DRA
h

c
r
k
i
& k
zi
:dielectric wavenumbers along the & z directions

k
0i
= 2tf
i
/c : wavenumber in air
(1)
Resonance frequency of the HEM
mnr
mode of the cylindrical DRA
i = 1, 2 for f
1
, f
2
f
1
: HEM
111
mode frequency
f
2
: HEM
113
mode frequency
82

u

z
x
y
|
c
r
a

Infinite
dielectric rod
Resonance frequency of the HEM
mnr
mode of the cylindrical DRA
For k

:
2 2
0
) 1 ( '
i i r i
k k k

c =
where
is the radial wavenumber outside the
dielectric rod

J
m
(x) : Bessel function of the first kind
K
m
(x): modified Bessel function of the second kind.
(2)
(3)
D. Kajfez and P. Guillon, Dielectric resonators, Norwood, MA, Artech House, Inc., 1986.
( ) ( )
( )
( ) ( )
( )
2 4
2 2 2 2 2
) ' (
) ' )( ' (
'
' '
'
1
) (
'
'
' '
'
1
) (
'
1
a k k
k k k k m
a k K
a k K
k a k J
a k J
k a k K
a k K
k a k J
a k J
k
i i
i r i i i
i m
i m
i i m
i m
i
r
i m
i m
i i m
i m
i

c
c
+ +
=
|
|
.
|

\
|
+
|
|
.
|

\
|
+
83

h
z

c
r
Infinite dielectric
slab
|
|
|
.
|

\
|

=

zi
zi i r r
i
zi
k
k k
p
hk
2 2
0 1
) 1 (
tan
c c
For k
z
: approximated by the
TM
01
-mode wavenumber
Resonance frequency of cylindrical DRA
(i = 1, 2 for f
1
, f
2
)
where p
1
= 1 and p
2
= 3
correspond to the HEM
111
and
HEM
113
modes, respectively.
(4)
R. K. Mongia and P. Bhartia, Dielectric resonator antennas- a review and general design relations for
resonant frequency bandwidth, International Journal of Microwave and Millimeter-Wave Computer-
Aided Engineering, vol. 4, no. 3, pp 230-247, 1994.
84

=

|
|
|
.
|

\
|
+
+
+ =
4
1
4 4
1
2
1
i
i
i
f
f B
i
i
r r
S
D
C e
A E
a
h
i
c c
(
(
(
(

0
0
0
4 4 4 4
3 3 3 3
2 2 2 2
1 1 1 1
D C B A
D C B A
D C B A
E D C B A
s
=
(
(
(
(

0 996 . 1 982 . 3 162 . 1 23 . 19
0 2 . 160 713 . 4 511 . 1 15 . 36
0 7 . 3682 402 . 4 2 . 625 3 . 680
116500 34800 937 . 0 234 . 0 7 . 489
(1)
f
1
: HEM
111
mode frequency (lower band)
f
2
: HEM
113
mode frequency (upper band)
Ground plane
z
a
Cylindrical
DRA
h

c
r
Design formula of ratio h/a for given f
1
, f
2
, and c
r

Using the covariance matrix adaptation
evolutionary strategy again,
85
(
(
(

|
|
|
.
|

\
|
+
+
+ =

=
i
i
a
h B
i
i
i
r r
S
D
C e
A E
f
a
i
4
1
4 4
1 r
1
2
c
c c c t
(
(
(
(

0
0
0
4 4 4 4
3 3 3 3
2 2 2 2
1 1 1 1
D C B A
D C B A
D C B A
E D C B A
s
(
(
(
(

0 057 . 0 114 . 6 659 . 5 429 . 4
0 814 . 17 9764 . 0 0368 . 0 152 . 0
0 1 . 304 9973 . 0 005 . 0 0571 . 0
10932 8 . 3107 00152 . 0 751 . 1 109 . 1
=
(2)
Design formula of radius a
Radius a can be found by inserting h/a into (2) below:
After a is found, h can be determined from h/a.
Maximum error of a: 2.1% for 1 s h/a s 3.5, 9 s c
r
s 27
Maximum error of h: 3.0% for 1.28 s h/a s 1.85, 9 s c
r
s 27
86
A. Example for dualband cylindrical DRA design
a = 17.9 mm & h = 42.5 mm
Given: f
1
= 1.71 GHz (DCS:1.71- 1.88 GHz )
f
2
= 2.4 GHz (WLAN:2.4 - 2.48 GHz ),

r
=9.4
Using formulas
(1) & (2)
87

Matching slot Excitation
strip
Via
x
y
|
Cylindrical
DRA
(c )
r
a
Aperture
Feedline
Ground plane
W
f
W
s
D
s
L
s

a
h
u
x
z
Cylindrical
DRA
(c )
r
Matching slot
Via
Feedline
d
Ground plane
Aperture
for via
Excitation strip
w
l
Configuration of the dualband LP DRA
Top view
Side view
a = 18.7 mm, h = 42.5 mm, c
r
= 9.4, l = 12.5 mm, w = 1 mm,
Ls = 20 mm, Ws = 1.5 mm, and Ds = 12.75 mm.
Radius a has been slightly increased to reduce the merging effect
88

1.6 1.8 2 2.2 2.4 2.6
-30
-20
-10
0
Frequency (GHz)
HFSS Simulation
Measurement
Reflection Coefficient |S11| (dB)
Measured and Simulated Reflection coefficients
Reasonable agreement
Lower band impedance bandwidth: 15.5% (1.70-2.00 GHz)
Upper band impedance bandwidth: 3.7% (2.39-2.48 GHz)
89

oo
-40 -30 -20 -10 0
30
150
60
120
90 90
120
60
150
30
180
o
o o
o
o o
o
o
o
o
o
dB
90 90
150
180
o
o
o
o
dB
-40 -30 -20 -10 0
30
150
60
90
120
60
150
30
180
o o
o o
o
o
o
o o
o
dB
120
Col-pol
Cross-pol
Cross-pol
Simulation
H-plane (y-z plane)
Measurement
E-plane (x-z plane)
-x +x
-y
+y
u
= 0
o
u
= 0
o
-40 -30 -20 -10 0
30
150
60
120
90 90
120
60
150
30
180
o
o o
o
o o
o
o
o
o
o
dB
90 90
150
180
o
o
o
o
o
dB
-40 -30 -20 -10 0
30
150
60
90
120
60
150
30
180
o o
o o
o o
o o
o o
o
dB
120
Col-pol
Cross-pol
Cross-pol
Simulation
H-plane (y-z plane)
Measurement
E-plane (x-z plane)
-x +x -y
+y
u
= 0
o
u
= 0
o
Measured and simulated radiation patterns
HEM
111
mode: measured (1.8 GHz), simulated (1.8 GHz)
HEM
113
mode: measured (2.42 GHz), simulated (2.45 GHz)
(a) (b)
Broadside radiation patterns are observed.
Co-polarized fields > cross-polarized fields by more than 20 dB in the
boresight direction.
90

1.6 1.65 1.7 1.75 1.8 1.85 1.9
0
2
4
6
Frequency (GHz)
Lower band gain (dBi)
HFSS Simulation
Measurement
Frequency (GHz)
Upper band gain (dBi)
2.3 2.35 2.4 2.45 2.5 2.55
0
2
4
6
8
10
10
8
6
4
2
0
2.3 2.35 2.5 2.4 2.45 2.55
Measured and simulated gain
HEM
111
mode: Maximum measured gain of ~6 dBi (1.75 GHz)
HEM
113
mode: Maximum measured gain of ~ 8 dBi (2.43 GHz)
91

L2
L3
L1
W2
W0
WS
W3
W1
LS
DS
Input port
Via
Cylindrical
DRA
(c )
r
x
|
y
a
quadrature
coupler
Dualband
grounding
via
To
plane
Ground
port
Isolation
slot
Matching
strip
Excitation
via
To grounding
Dualband CP DRA
a = 18.7 mm, h = 42.5 mm, c
r
= 9.4, l = 12.5 mm, w = 1 mm, Ls = 21 mm, Ws = 1.5 mm, Ds =
12.75 mm, L
1
= 26.9 mm, L
2
=26.5 mm, L
3
= 56.65 mm, W
0
= 4.66 mm, W
1
= 7.3 mm, W
2
=
4.44 mm, and W
3
= 0.46 mm.

a
h
u
x
z
Cylindrical
DRA
(c )
r
Matching slot
Via
Feedline
d
Ground plane
Aperture
for via
Excitation strip
w
l
Top view
Side view

1.6 1.8 2 2.2 2.4 2.6
-40
-30
-20
-10
0
Reflection Coefficient |S11| (dB)
Frequency (GHz)
HFSS Simulation
Measurement
Reasonable agreement
Lower band bandwidth:18.9% (1.58-1.91 GHz).
Upper band bandwidth:7.8% (2.33-2.52 GHz).
Measured and simulated reflection coefficients
93

1.6 1.65 1.7 1.75 1.8 1.85 1.9
0
2
4
6
8
10
Lower band AR (dB)
Frequency (GHz)
HFSS Simulation
Measurement
Upper band AR (dB)
Frequency (GHz)
2.3 2.4 2.5
0
2
4
6
8
8
6
4
2
0
2.3 2.4 2.5
Measured and simulated axial ratios (ARs)
Reasonable agreement
Lower band AR bandwidth: 12.4% (1.67-1.89 GHz)
Upper band AR bandwidth: 7.4% (2.34-2.52GHz)
94

-40 -30 -20 -10 0
30
150
60
120
90 90
120
60
150
30
180
o
o o
o
o o
o
o
o o
o
dB
90 90
150
180
o
o
o
o
o
dB
-40 -30 -20 -10 0
30
150
60
90
120
60
150
30
180
o o
o o
o o
o o
o o
o
dB
120
LHCP
RHCP
RHCP
Simulation
H-plane (y-z plane)
Measurement
E-plane (x-z plane)
-x +x -y
+y
u = 0
o
u
= 0
o
-40 -30 -20 -10 0
30
150
60
120
90 90
120
60
150
30
180
o
o o
o
o o
o
o
o
o
o
dB
90 90
150
180
o
o
o
o
o
dB
-40 -30 -20 -10 0
30
150
60
90
120
60
150
30
180
o o
o o
o o
o o
o o
o
dB
120
RHCP
LHCP
RHCP
Simulation
H-plane (y-z plane)
Measurement
E-plane (x-z plane)
-x +x -y
+y
u
= 0
o
u = 0
o
Measured and simulated radiation patterns
(a) (b)
HEM
111
mode: measured (1.8 GHz), simulated (1.8 GHz)
HEM
113
mode: measured (2.42 GHz), simulated (2.45 GHz)
Broadside radiation patterns are observed.
LHCP fields > RHCP fields by ~20 dB in the boresight direction.
95
B. Example for wideband cylidnrical DRA design
a = 10.3 mm & h = 34.3 mm
Given: f
1
= 2.90 GHz, f
2
= 3.72 GHz,
r
= 9.4
Using formula
(5) & (6)
96

Cylindrical
DRA
(c ) r
Conducting
feeding strip
Coaxial
aperture
l
w
h
a
Ground plane
z
x
y

2.8 3 3.2 3.4 3.6 3.8 4 4.2
-40
-30
-20
-10
0
Reflection Coefficient |S11| (dB)
Frequency (GHz)
HFSS Simulation
Measurement
Configuration Reflection coefficient
a = 10.3 mm, h = 34.3 mm, c
r
= 9.4,
l = 12 mm, and w = 1 mm.
Good agreement
Measured impedance bandwidth:
23.5% (3-3.8 GHz)
Wideband LP cylindrical DRA
97

3 3.2 3.4 3.6 3.8 4
0
2
4
6
8
10
12
Frequency (GHz)
Antenna Gain (dBi)
HFSS Simulation
Measurement
Measured and simulated gain
HEM
111
mode: Maximum measured gain of ~7 dBi (3.29 GHz)
HEM
113
mode: Maximum measured gain of ~10 dBi (3.83 GHz)

98

L1
L1
W0
W0
W1
Input port
Excitation
strip
Isolation
port
Ground
plane
via
Excitation strip
Cylindrical
DRA
(c )
r
Wideband
quadrature
coupler
via
x
|
y
a

a
h
u
x
z
Cylindrical
DRA
(c )
r
Via
d
Ground plane
Aperture
for via
Excitation strip
w
l
Wideband quadrature coupler
a = 10.3 mm, h = 34.3 mm, c
r
= 9.4, l = 11.5 mm, w = 1 mm,
L
1
= 14.67 mm, W
0
= 1.94 mm, and W
1
= 3.21 mm.
Wideband CP cylindrical DRA
Top view
Side view
99

3 3.2 3.4 3.6 3.8 4
-30
-20
-10
0
Frequency (GHz)
Reflection Coefficient |S11| (dB)
HFSS Simulation
Measurement

3 3.2 3.4 3.6 3.8 4
0
2
4
6
8
Frequency (GHz)
Axial ratio (dB)
HFSS Simulation
Measurement
Measured 3-dB AR bandwidth :
24.7% (3.05-3.91 GHz).
Measured impedance bandwidth:
25.5% (3.04-3.93 GHz).
Wideband CP DRA
Reflection coefficient Axial ratio
VI. Dualfunction DRAs

101
Advantage
System size and cost can be reduced by
using dualfunction DRAs.

Additional functions
- Packaging cover
- Oscillator
102
Packaging Cover
103
Conventional
Aperture
Metallic supports
for grounding
Dielectric resonator
antenna/oscillator
load and packaging
cover
Metal ground
Power supply
z
y
Front view
h
H
Microstrip
feedline
d
Transistor (other
RF components
not shown)
Proposal
104
Antenna Configuration
Aperture
Feedline and the
RF/MIC circuits
Metal ground
Coaxial line
Dielectric resonator antenna
and packaging cover
Metallic supports
for grounding
H
h
y
z
Side view

Coaxial
x
y
Microstrip
feedline
Aperture
L
W
a
b
Top view
Resonant frequency
f
0
= 2.4GHz

Parameters:
Hollow DRA:
L=30mm, W=29mm,
H=15mm, & c
r
=12

Metallic Cavity:
a = 15mm, b = 21.6mm, h = 5mm
Top face : Duroid c
r
=2.94
thickness 0.762mm
Aperture: 0.2063
e

105
Design Procedure (Simulation):
Step 2
Remove the lower center
portion concentrically to
form a notched DRA.
As a result, the resonant
frequency >2.4GHz
Step 3
Cover the two sides with the
same material. Move the
frequency back to 2.4GHz
by increasing the thickness.
(thickness f
0
)
Step 1
Use the DWM to design
a solid rectangular DRA
at 2.4-GHz fundamental
TE111 Mode.
x
z
106
Experimental Verification:


- Hard-clad foam (c
r
1) is used to form the
container.

- ECCOSTOCK HiK Powder of c
r
=12 is used as
the dielectric material.
107
2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3
Frequency (GHz)
-60
-50
-40
-30
-20
-10
0
R
e
t
u
r
n

L
o
s
s

(
d
B
)
Theory
Experiment
2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3
Frequency (GHz)
-60
-40
-20
0
20
40
60
80
I
n
p
u
t

I
m
p
e
d
a
n
c
e

(
o
h
m
)
Return Loss and Input Impedance
(Passive hollow RDRA with a metallic cavity)
Good agreement.
Bandwidth ~ 5.6%.
Measured resonance frequency: 2.42GHz (error < 0.83%)

108
-40 -30 -20 -10
0
30
-150
60
-120
90
-90
120
-60
150
-30
180
0
co-pol
cross-pol
-40 -30 -20 -10 0
30
-150
60
-120
90 -90
120
-60
150
-30
180
0
cross-pol
co-pol
Theory
Experiment
(a) (b)
x-z plane y-z plane
Radiation Patterns
(Passive hollow DRA with a metallic cavity)
Broadside TE
111
y

mode is observed.

Co-polarized fields generally stronger than the cross-
polarized fields by 20dB in the boresight direction.
109
2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3
Frequency (GHz)
-40
-30
-20
-10
0
R
e
t
u
r
n

L
o
s
s

(
d
B
)
A A
DRA DRA
Return
Loss
Return
Loss
Receiver Transmitter
DRA (passive)
DRA (active, receiver)
DRA (active, transmitter)
Return Loss of the Active Integrated Antenna
Integrated with Agilent AG302-86 low noise amplifier (LNA)
(gain of 13.6dB at 2.4GHz)
LNA prematched to 50O at the input.
A small hole is drilled on the ground plane to supply the DC bias to the LNA.
110
-180 -120 -60 0 60 120 180
Observation angle (degree)
-80
-70
-60
-50
-40
-30
-20
-10
0
A
m
p
l
i
t
u
d
e

(
d
B
)
co-pol (passive)
co-pol (active)
cross-pol (passive)
cross-pol (active)
Amplified Radiation Pattern
Compared to the passive DRA, the active DRA has a gain of
7 - 12dB across the observation angle from -90
o
to 90
o
.
The gain is less than the specification due to unavoidable
impedance variations and imperfections in the measurement.
Dielectric Resonator Antenna
Oscillator (DRAO)

112
The DRA is used as the oscillator load,
named as DRAO.
Methodology
The reflection amplifier method is used to
design the antenna oscillator.
113
DRA
jX
Zin=Rin+Xin
Lm
Transistor
C
ZL=RL+XL
DRAO Schematic Diagram
- Oscillate condition: X
L
+X
in
=0 & R
L
<|R
in
|
- DRA first replaced by a 50O load at 1.85GHz.
114
L
w
Top view
Microstrip
feedline
Aperture
y
x
Lm Ls
La
Wa
Transistor (other
RF components
not shown)
Resonance frequency
f
o
= 1.85GHz at TE
111
y


Parameters:
DRA
L=52.2mm,
W=42.4mm,
H=26.1mm,
cr = 6.

Aperture
L
a
= 0.3561
e
, W
a
= 2mm
L
s
= 9.5 mm, L
m
= 40 mm.

Duroid substrate
crs=2.94, d=0.762mm

Antenna Configuration:
Dielectric resonator
antenna and
oscillating load
Microstrip
feedline Aperture
Ground
L
H
z
x
Substrate
Side view
d
Transistor (other
RF components
not shown)
cr
115
Return Loss and Input Impedance
Good agreement.
Bandwidth ~ 22.14%.
Resonance frequency: Measured 1.86GHz
Simulated 1.83GHz (1.5% error).
1.6 1.7 1.8 1.9 2 2.1 2.2
-60
-50
-40
-30
-20
-10
0
Frequency (GHz)
Return Loss (dB)
HFSS Simulation
Experiment
1.6 1.7 1.8 1.9 2 2.1
-80
-60
-40
-20
0
20
40
60
80
Frequency (GHz)
Input Impedance (ohms)
116
Spectrum of the Free-running DRAO
Transmitting power P
t
= 16.4dBm
DC-RF efficiency: ~ 13% (2-25% in the literature).
Phase noise: 103dBc/Hz at 5MHz offset
Second harmonic < fundamental by 22dB
117
Radiation Pattern
Broadside TE
111
y

is observed.

Co-polarized fields are generally 20dB stronger
than the cross-polarized fields in the boresight direction.
Solid DRAO (measur.)
Passive Solid DRA (measur.)
HFSS Simulation
-40 -30 -20 -10 0
30
60
90 -90
120
150
-30
180
0
-60
-120
-150
co-pol
cross-pol
x-z plane
(a)
-40 -30 -20 -10 0
30
60
90 -90
120
150
-30
180
0
-60
-120
-150
y-z plane
(b)
co-pol
cross-pol
118
DRA can be of any shape. Can it be made like a swan?
Yes!
DRA is simple made of dielectric. Can glass be used for
the dielectric?
It leads to probably the most beautiful antenna in
the world .
Yes!
Glass-Swan DRA
Distinguished Lecture
Transparent antennas: From 2D to 3D
119
120

The DRA can be easily excited with various excitation schemes.

Frequency tuning of the DRA can be achieved by using
a loading-disk or parasitic slot.

The dualband and wideband DRAs can be easily designed using
higher-order modes.

Compact omnidirectional CP DRAs have been presented

Dualfuncton DRAs for packaging and oscillator designs have
been demonstrated.



Conclusion
121
122
Q & A