Development of Dielectric Resonator Antenna 2

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
Outline
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
What is Dielectric Resonator Antenna (DRA) ?
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 r.
3
Some DRAs :
Advantages of the DRA

Low cost
Low loss (no conductor loss)
Small size and light weight
Reasonable bandwidth (~10% for r ~10)
Easy of excitation
High radiation efficiency ( generally > 95%)
Excitation schemes
(i) Microstrip line feed
Excitation schemes
(ii) Aperture-couple feed
DRA
Ground plane
Slot
Microstrip
feed line
Feed
Substrate
Excitation schemes
(iii) Coaxial feed
Coaxial feed
Top view
Bottom view
Aperture-coupled feed
Bottom view
Top view
10
Corporate feedline for DRA array
Slot-fed DRA array using corporate

microstrip feed network
11
Conformal-Strip Method
12
Rectangular Dielectric
Resonator Antennas
Proposed Antenna Geometry

z
Ground
Plane
Rectangular
DRA
W1
Coaxial
Aperture
l1
Conducting
Strip
a
(mm)
b
(mm)
d
(mm)
l1
(mm)
W1
(mm)
14.3
25.4
26.1
10
9.8
14
Analytical Solution
Dielectric
Waveguide Model (DWM)
Resonant frequency of TEmnl(y) mode

f0
c
2 r
k x2 k y2 k z2
m
n
l
kx
,ky
, kz
a
b
2d
k x2 k y2 k z2 r k02
Numerical Solution
Finite-Difference
Time-Domain (FDTD) method
Advantages
- Very simple
- High modeling capability for general EM structures
- No spurious modes nor large matrix manipulation
- Provide a very wideband frequency response
Disadvantages
- Time consuming, powerful computer required
16
Source model and extraction of S parameters

Baseband Gaussian pulse
E z exp[ (t n 3T ) 2 T 2 ] T : pulse width
Conformal Strip
FFT V (t )
FFT I (t )
Z in Z 0
S11
Z in Z 0
Z in
Ground Plane
(I)
Hy Ez (V) Hy
Hx
Hx
Source occupies only one grid

17
Parameters
Uniform Cartesian grids
x = 0.715 mm, y = 0.508 mm, z = 0.5 mm
T = 0.083ns, t0 = 3T
10-cell-thick PML with polynomial spatial scaling
(m = 4 and max = 1)
total grid size : 80x
110y
112z
total time steps : 10000
18
Input Impedance/S11
200
Experiment
Theory
-5
|S11|(dB)
Input Impedance (ohm)
150
-10
Resistance
-15
-20
-25
100
-30
2.5
3.5
4.5
5.5
Frequency (GHz)
50
0
-50
Experiment
Theory
-100
3
3.5
Reactance
4.5
5.5
Frequency (GHz)
Reasonable agreement.
Wide Bandwidth of ~ 43%.
Dual resonant TE111y and TE113y modes are excited.
19
Comparison between Theory and Measurement

Resonant
Modes
Measured
resonant
frequencies
Calculated resonant
frequencies (FDTD)
Predicted resonant
frequencies (DWM)
fmea (GHz)
fFDTD
(GHz)
error
(%)
fDWM
(GHz)
error
(%)
y
TE111
3.81
3.90
2.3
3.95
3.6
y
TE112
N/A
N/A
N/A
4.26
N/A
4.57
4.60
0.7
4.7
1.7
y
TE113
Reasonable agreement.
20
Field Distribution --- TE111y
z
y
2d
b
x-z
Electric field
Magnetic field
x-y
Imaged DRA (gound plane removed)
With gound plane

21

z
2d
b
x-z
Electric field
Magnetic field
x-y

22
z
y
2d
a
b
x-z
Electric field
Magnetic field
x-y
With gound plane

23
Radiation Patterns
o
0o
30
30
60o
30
60o
0o
30o
60o
60o
f = 3.5 GHz
90o
90o90o
-40 -30 -20-10 0
(-x)
30
0o
90
(-x)
30
60o
0o
90 90
-40 -30 -20-10 0
(+x) (-y)
(+y)
30o
60o
E (xz) - plane
H (yz) - plane
o
30
60o
-40 -30 -20-10 0
(+x) (-y)
E (xz) - plane
o
90o
60o
90
-40 -30 -20-10 0
H (yz) - plane
f = 4.3 GHz
(+y)
Broadside radiation patterns are observed.

Measured E-plane crosspolarized fields mainly caused by finite
ground plane diffraction.
24
III. Circularly Polarized Design

using a Parasitic Strip
Proposed Antenna Geometry

Conducting
Strip
0
Top
Parasitic View
Patch
Ground
Plane
W1
Rectangular
DRA
l2
W2
l1
Parasitic
Patch
C (Center of
Coaxial
Aperture
Conducting Parasitic Patch)

Strip
a
b
d
l1
W1
l2
(mm) (mm) (mm) (mm) (mm) (mm)
24
23.5
12.34
10
12
W2
0
(mm) (degree)
1
225.6
r
9.5
26
Input Impedance/S11
0
|S11| (dB)
-5
-10
-15
-20
-25
2.5
3.5
4.5
Frequency (GHz)
Reasonable
agreement.
Bandwidth ~ 14%.
Two nearly-degenerate TE111(y) modes are excited.
CP operation
27
Axial Ratio in the boresight direction
Axial Ratio (dB)
20
15
10
Experiment
Theory
3
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
Frequency (GHz)
3-dB AR bandwidth is ~ 2.7%, which is a typical value for

a singly-fed CP DRA.
28
The H field of the DRA without and with parasitic

strip (Top view)
Feeding strip
Without parasitic strip - LP field
3.4 GHz
Feeding strip
Parasitic
strip
With parasitic strip - CP field

3.4 GHz
29
Radiation Patterns (f = 3.4GHz, )

30o
LHCP
0o
60o
90o
(-x)
30o
30o
60o
0o
60o
90o90o
-40 -30 -20-10 0
(+x) (-y)
xz plane
30o
60o
90o
-40 -30 -20-10 0
(+y)
yz plane
RHCP
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).
30
Effects of feeding strip length l1

20
Axial Ratio (dB)
|S11| (dB)
-5
-10
l1 = 8 mm
l1 = 10 mm
l1 = 12 mm
-15
-20
2.4
2.6
2.8
3.2
3.4
3.6
Frequency (GHz)
3.8
4.2
l1 = 8 mm
l1 = 10 mm
l1 = 12 mm
15
10
0
3.1
3.2
3.3
3.4
3.5
3.6
3.7
Frequency (GHz)
Input impedance changes substantially with l1.

AR is almost unchanged for different l1.
l1 can be adjusted to match the impedance without changing AR.
31
II. Frequency Tuning Technique
Backgruond
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.
Frequency Tuning Technique

- using a loading disk
The slot-coupled DRA with a conducting loading

cap
z
Conducting
Loading Cap
ra
Conducting
Loading Cap
Hemispherical DRA
Ws
x
Ground
plane
Ls
Wf
Lt
Slot
Dielectric
substrate rs
Slot
Lt
Side view
microstrip
line
Hemispherical
DRA
microstrip
line
y
Top view
Hemispherical DRA: radius a = 12.5 mm, dielectric constant r = 9.5.

Coupling slot : length Ls , width Ws
Open-circuit stub: length Lt
Grounded dielectric slab: rs = 2.33, height d = 1.57 mm
Microstrip feedline: width Wf = 4.7 mm
Calculated and measured return losses

(Ls = 12 mm and Ws = 1 mm)
0
-10
-20
|S11| (dB)
-30
= 26.38o
Lt = 4.42 mm
-40
Without cap ( = 0 )
Lt = 10.63mm
-50
Theory
-60
-70
= 52.8 o
Lt = 13.6 mm
Experiments
3.2
3.4
3.6
3.8
Frequency (GHz)
Resonance frequency:
3.52 GHz without any conducting cap ( = 00), with Lt = 4.42 mm
3.25 GHz ( = 26.38o and Lt = 4.42 mm)
3.68 GHz ( = 52.8o and Lt = 13.6 mm)
Calculated and measured radiation patterns
Co-pol
0o
o
0o
30
30
30o
30
60o
60o
60o
60o
X-pol
90o
-40
-30
-20
-10
90o
90o
-40
(a)
-30
-20
-10
90o
(b)
Reasonable agreement
between
theory
and
experiment.
3.25 GHz ( = 26.38o and Lt = 4.42 mm)

0o
30o
30o
30o
60o
60o
90o
-40
(c)
The effect of loading

cap on field pattern is not
significant.
Co-pol
0o
-30
-20
-10
60o
X-pol
90o
30o
60o
90o
-40
-30
(d)
3.58 GHz ( = 52.8o and Lt = 13.6 mm)
-20
-10
90o
60
16
50
14
12
40
10
30
8
20
10
0
3.2
Stub Length Lt (mm)
Angle (degree)
Calculated and Lt for having a good return loss

(minimum |S11| < -20dB)
4
3.3
3.4
3.5
3.6
3.7
Tuning frequency fr (GHz)
The resonant frequency can be tuned by varying and Lt

decreases from 26.38o to 0o (3.25 < fr < 3.5 GHz )
increases from 0o to 52.8o (3.5 < fr < 3.78 GHz)
Impedance bandwidth
Impedance Bandwidth (%)
10
2
3.2
3.3
3.4
3.5
3.6
3.7
Frequency (GHz)
The bandwidth decreases after a loading cap is added.
Frequency Tuning Technique

- using a parasitic slot
The annular-slot-excited cavity-backed DRA

Hemispherical DRA
z
Coaxial cable
Parasitic slot
rA r
rB
a
Ground plane
Feeding slot
x
WB WA
b
metallic cavity
(a) Side view

Hemispherical DRA
Feeding slot
Parasitic slot
r
rA
rB
WB
WA
Coaxial cable
x
y
(b) Top view
metallic cavity
IV. Omnidirectional Circularly

Polarized DRA
Advantages of omnidirectional CP antenna

Provide larger coverage.
CP DRAs concentrated on broadside-mode designs only.
43
Design I:
Slotted omnidirectional CP DRA
44
Antenna configurations
Dielectric
block
d
w
Slot
h
h
b
a
Probe
2r1
g
SMA connector
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.
45
Antenna principle
Po
la
riz
er
Dielectric block
LP omnidirectional DRA
y
Dielectric block with the wave

polarizer
Dielectric
block
Dielectric
slabs
d
w
h
E
x
z
Wave polarizer
b
a
Proposed compact omnidirectional CP DRA
Photographs of the prototype

Prototype for 2.4 GHz WLAN design
Top face and sidewalls
Bottom face
Design parameters
r = 15, a = b = 39.4 mm, h = 33.4 mm, w = 9.4 mm,
d =14.4 mm, r1 = 0.63 mm, l = 12.4 mm, g = 12.7 mm
47
Simulated and measured results

Reflection coefficient
Axial ratio
|S11| (dB)
-10
Axial Ratio (dB)
-20
3
HFSS Simulation
Experiment
-30
HFSS Simulation
Experiment
0
2
2.2
2.4
2.6
Frequency (GHz)
2.8
Impedance bandwidth:
Simulated: 20.3% (2.34-2.87 GHz)
Measured: 24.4% (2.30-2.94 GHz)
2.2
2.4
2.6
Frequency (GHz)
2.8
AR bandwidth:
Simulated: 8.2% (2.34-2.54 GHz)
Measured: 7.3% (2.39-2.57 GHz)
48
Simulated and measured radiation patterns

30 o
0 o
30
RHCP
LHCP
60 o
60
90 o
-40
-30
-20
-10
90 o
330 o
0 o
30 o
RHCP
60 o
300 o
270 o
-40
-30
-20
-10
dB
120 o
120 o
150 o
150 o
180 o
(-x)
xz-plane
90 o
dB
120 o
240 o
150 o
210 o
180 o
(+x)
xy-plane
Very good omnidirectional characteristic

In the horizontal plane, LHCP fields > RHCP fields by
~20 dB .
49
Simulated and measured antenna gain

Gain (dBic)
3
2
1
0
-1
HFSS Simulation
Experiment
-2
-3
2.2
2.4
2.6
Frequency (GHz)
2.8
50
Design II:
Wideband omnidirectional CP antenna

with parasitic metallic strips
51
Antenna configurations
z
Hollow circular
cylinder
Slot
Strip
ls
ws
Probe
2r1
SMA connector
Perspective view
Front view
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.
52
Photographs of the prototype

Prototype for 3.4 GHz WiMAX design
Top face and sidewalls
Bottom face
Design parameters
r = 15, a = b = 30 mm, h = 25 mm, r = 3 mm, w = 7 mm, d =10.5 mm

ls = 30.5 mm, ws = 1 mm, x0 = 6.4 mm, r1 = 0.63 mm, l = 19 mm.
53
Simulated and measured reflection

coefficient and axial ratio
Axial Ratio (dB)
|S11| (dB)
18
HFSS Simulation
Experiment
Impedance bandwidth:
Simulated: 22.3% (3.11-3.89 GHz)
Measured: 24.5% (3.08-3.94 GHz)
-10
12
-20
-30
2.8
0
2.8
3.2
3.2
3.6
Frequency (GHz)
3.6
Frequency (GHz)
AR bandwidth:
Simulated: 24.8% (3.11-3.99 GHz)
Measured: 25.4% (3.16-4.08 GHz)
Overlapping bandwidth: 22.0%; bandwidth widened by ~3 times.

54
Simulated and measured results

Antenna gain
Radiation efficiency
Gain (dBic)
Efficiency
0.8
1
0
0.6
-1
0.4
-2
HFSS Simulation
Experiment
-3
-4
2.8
3.2
3.6
Frequency (GHz)
0.2
4
0
2.8
3.2
3.6
Frequency (GHz)
Measured gain: wider bandwidth.

Measured antenna efficiency: 84-98% (3.1-3.9 GHz).
55
Simulated and measured radiation patterns

30 o
60 o
0 o
30 o
60 o
RHCP
90 o
-40
-30
-20
330 o
LHCP
-10
90 o
0 o
30 o
RHCP
300 o
60 o
270 o
-40
-30
-20
-10
dB
120 o
120 o
150 o
150 o
xz-plane
90 o
0 o
30 o
RHCP
60 o
60 o
90 o
-40
-30
-20
120 o
240 o
150 o
210 o
180 o
(+x)
xy-plane
3.4 GHz
330 o
LHCP
-10
dB
180 o
(-x)
30 o
0 o
30 o
RHCP
300 o
60 o
90 o 270 o
-40
-30
-20
-10
dB
120 o
120 o
150 o
150 o
xz-plane
90 o
dB
120 o
240 o
150 o
210 o
180 o
(-x)
180 o
(+x)
xy-plane
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).
56
V. Dualband & Wideband DRAs
(i) Rectangular DRA
Background
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.
59
Use of higher-order DRA

mode
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.
60
Design Formulas for Dual-Mode rectangular DRA

z
2d
2d
b
x-z
y
TE111
xx 2d
Electric field
Magnetic field
a
b
x-y
x-z
xx
Electric field
Magnetic field
y
TE112
a
b
x-z
x-y
a
Electric field
Magnetic field
x-y
y
TE113
The E-field should vanish on the PEC and the TE112 mode
cannot be excited properly.

The TE111 mode and TE113 mode are used in the dualmode design.
61
Formula Derivation
Ground plane
The wavenumbers kx1, x2 and

kz1, z2 can be written as follows:
3
2d
k z1
2d
kz 2
z
Rectangular DRA r
y
x
k x1 k x 2
62
From the DWM model, the frequencies f1, f2 are given by:
f1, 2
where
c
2 r
k x21, x 2 k y21, y 2 k z21, z 2
k y1, y 2 k12, 2 k x21, x 2 k z21, z 2
(*)
in which k1, 2 2 r f1, 2 / c are wavenumbers in the

dielectric, with c being the speed of light in vacuum.
Engineering Formulas for the DRA dimensions

10 .32
9k12 k 22
k 22
k12
10 .32
(3.96
f2
)
f1
b 0.65b1 0.35b2
where
4
3
2
f2
f2
f2
f2
3
d 0.1393 2.3209 11 .4422 23 .4984 18 .4437 10
f1
f1
f1
f1
(m)
b1, 2
2
k y1, y 2
tan 1
1 k1, 2
1
1
r k y1, y 2
64
Limit of frequency ratio f2/f1

From
10 .32
9k12 k 22
10 .32
(3.96
f2
)
f1
We have
9k12 k 22 d 0
3k1 > k2 or 3f1 > f2
giving
f2/f1 < 3
which is the theoretical limit that is not known before.
65
Error analysis
4
Error of f1 (%)
r 5
r 10
r 30
r 70
f1 kept constant at 2.4 GHz.

1
1.2
1.4
1.6
1.8
2.2
2.4
2.6
2.8
f2 / f 1
Compared with DWM results, errors of f1, f2 are both

less than 2.5% for 1 < f2/f12.8 , 5 r 70.
66
A. Example for Dual-band Rectangular DRA Design
Given: f1 = 3.47 GHz (WiMAX)

f2 = 5.2 GHz (WLAN), r=10
Using dual-mode
formulas
a = 20.8 mm, b = 10.5 mm, and d = 18.5 mm.

67
Configuration of the dualband DRA

z
d
LS
Coupling
slot
W
L
Substrate rs
Ground plane
Wf
Rectangular
DRAr
Microstripline
W = 2.6 mm, L =10.6 mm, Ls=7.2 mm, Wf=1.94 mm,

h=0.762mm, rs= 2.93
68
Measured and simulated reflection coefficients

Reflection coefficient |S11| (dB)
0
-10
-20
HFSS Simulation
Measurement
-30
-40
3.2
3.6
4
4.4
4.8
Frequency (GHz)
5.2
5.6
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).
69
COMPARISON OF DESIGN, SIMULATED, AND MEASURED

RESONANCE FREQUENCIES OF TE111y AND TE113y MODES
Resonant
Mode
Measured
frequency
(GHz)
Design
frequency
Simulated HFSS
frequency
f1,2
(GHz)
Error
(%)
fHFSS
(GHz)
Error
(%)
TE111y
3.40
3.47
2.05
3.47
2.05
TE113y
5.18
5.30
2.32
5.24
1.15
70
Measured and simulated radiation patterns

Co-pol
30 o
0o
30 o
90 o
-40
0o
30 o
60 o
-30
-20
-10
60 o
60 o
90 o 90 o
-40
-30
-20
-10
dB
120 o
120 o
90 o
dB
120
150 o
150 o
Co-pol
Cross-pol
Cross-pol
60 o
30 o
120 o
150 o
150 o
180 o
180 o
-y
-x
+x
E-plane (x-z plane)
+y
H-plane (y-z plane)

Measurement
Simulation
(a)
TE111y 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.
71

Co-pol
30 o
0o
30 o
30 o
60 o
60 o
60 o
-40
30 o
Cross-pol
Cross-pol
90 o
0o
-30
-20
-10
60 o
90 o 90 o
-40
-30
-20
-10
120 o
90 o
120 o
120 o
150 o
150 o
dB
dB
120 o
Co-pol
150 o
150 o
180 o
180 o
-y
-x
+x
E-plane (x-z plane)
+y
H-plane (y-z plane)

Measurement
Simulation
(b)
TE113y 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.
72
Measured antenna gain

Gain (dBi)
10
-5
3.5
4.5
5
Frequency (GHz)
5.5
TE111y mode: Maximum gain of 4.02 dBi at 3.48 GHz.

TE113y mode: Maximum gain of 7.52 dBi at 5.13 GHz.
Electrically larger antenna has a higher antenna gain.
73
B. Example for Wideband DRA Design
Given: f1 = 1.98 GHz (PCS)

f2 = 2.48 GHz (WLAN), r=10
Using formulas for
dual-mode
rectangular DRA
a = 30.7 mm, b = 24.7 mm, and d = 47.7 mm.
74
Configuration of the wideband DRA

z
Ground plane
a
d
Rectangular
DRA r
Conducting
feeding strip
Coaxial
aperture
l = 17 mm, W = 1 mm
75

Reflection coefficient |S11| (dB)
0
-10
-20
-30
HFSS Simulation
Measurement
-40
1.5
2
2.5
Frequency (GHz)
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)
76

Co-pol
30 o
0o
30 o
30 o
Cross-pol
60 o
-40
30 o
-30
-20
-10
60 o
60 o
90 o 90 o
-40
-30
-20
-10
dB
120 o
120 o
90 o
dB
120 o
120o
150 o
150 o
Co-pol
Cross-pol
60 o
90 o
0o
150 o
150 o
180 o
180 o
-y
-x
+x
E-plane (x-z plane)
+y
H-plane (y-z plane)

Measurement
Simulation
(a)
Measured (2.16 GHz), simulated (2.11 GHz).

Co-polarized fields > cross-polarized fields
by more than 20 dB in the boresight direction.
77

Co-pol -30 o
0o
30 o
-30 o
Cross-pol
60 o
-40
30 o
-30
-20
-10
-60 o
60 o
90 o -90 o
-40
-30
-20
-10
dB
90 o
dB
120 o
-120o
120 o
150 o
-150o
Co-pol
Cross-pol
-60o
-90 o
0o
150 o
-150 o
180 o
180 o
-y
-x
+x
E-plane (x-z plane)
+y
H-plane (y-z plane)

Measurement
Simulation
(b)
Measured (2.41 GHz), simulated (2.46 GHz).

Co-polarized fields > cross-polarized fields by more
than 20 dB in the boresight direction.
78
Measured antenna gain

Gain (dBi)
8
0
1.6
1.8
2.2
2.4
Frequency (GHz)
2.6
2.8
The maximum gain of 6.98 dBi at 2.47GHz.

TE113y -mode gain > TE111y -mode gain.
79
(ii) Cylindrical DRA
Resonance frequency of the HEMmnr mode of the cylindrical DRA
Ground plane
r
Cylindrical
DRA
k2i k zi2 r k02i
(1)
i = 1, 2 for f1, f2
f1 : HEM111 mode frequency

f2 : HEM113 mode frequency
ki & kzi :dielectric wavenumbers along the & z directions

k0i = 2fi/c : wavenumber in air
81
Resonance frequency of the HEMmnr mode of the cylindrical DRA

For k:
Infinite
dielectric rod
m 2 (k i 2 k i '2 )(k i 2 r k i '2 )

(k i k i ' ) 4 a 2
(2)
where
r J m ' ki a 1 K m ' ki ' a

ki J m (ki a) ki ' Km ki ' a
1 J m ' k i a 1 K m ' k i ' a
k i J m (k i a) k i ' K m k i ' a
k i ' ( r 1) k02i k i 2
(3)
is the radial wavenumber outside the

dielectric rod
Jm(x) : Bessel function of the first kind
Km(x): modified Bessel function of the second kind.
D. Kajfez and P. Guillon, Dielectric resonators, Norwood, MA, Artech House, Inc., 1986.
82
Resonance frequency of cylindrical DRA

For kz: approximated by the
TM01-mode wavenumber
2
2
hk zi
1 r ( r 1) k0i k zi
tan
pi
k zi
Infinite dielectric
z
slab
(i = 1, 2 for f1, f2)
(4)
where p1 = 1 and p2 = 3
correspond to the HEM111 and
HEM113 modes, respectively.
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 Computer83
Aided Engineering, vol. 4, no. 3, pp 230-247, 1994.
Design formula of ratio h/a for given f1, f2, and r

z
f1 : HEM111 mode frequency (lower band)

f2 : HEM113 mode frequency (upper band)
Using the covariance matrix adaptation
evolutionary strategy again,
Ground plane
r
Cylindrical
DRA
h ES
1
Ai
4 4i Bi f 2
Di
a r
i 1 r
e f1 C
A1 B1
A B
2
2
A3 B3
A4 B4
C1
C2
C3
C4
D1 Es
D2 0
D3 0
D4 0
(1)
489.7 0.234 0.937 34800 116500

680.3 625.2 4.402 3682 .7
0
0
36.15 1.511 4.713 160.2
3.982
1.996
0
19.23 1.162
84
Design formula of radius a

Radius a can be found by inserting h/a into (2) below:
E
1
A
S
i
a
i (2)
4
4 i Bi h
2 r f1 r
i 1 r
e a C
A1
A
2
A3
A4
B1
C1
D1
B2 C2
D2
B3 C3
D3
B4 C4
D4
Es
0
1.751 0.00152 3107 .8 10932

1.109
0.0571 0.005 0.9973 304.1
0
0.0368 0.9764 17.814

0
0.152
5.659
6.114
0.057
0
4.429
After a is found, h can be determined from h/a.

Maximum error of a: 2.1% for 1 h/a 3.5, 9 r 27
Maximum error of h: 3.0% for 1.28 h/a 1.85, 9 r 27
85
A. Example for dualband cylindrical DRA design
Given: f1 = 1.71 GHz (DCS:1.71- 1.88 GHz )

f2 = 2.4 GHz (WLAN:2.4 - 2.48 GHz ),
r=9.4
Using formulas
(1) & (2)
a = 17.9 mm & h = 42.5 mm
86
Configuration of the dualband LP DRA

Excitation strip
Ground plane
x
Cylindrical
DRA r
Cylindrical
DRAr
a
Ws
Excitation
strip
Via
Feedline
Matching slot
Top view
Ground plane
Via
Aperture
Wf
Aperture
for via
Ds
Ls
Feedline
Matching slot
Side view
a = 18.7 mm, h = 42.5 mm, 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
87
Measured and Simulated Reflection coefficients

Reflection Coefficient |S11| (dB)
0
-10
-20
HFSS Simulation
Measurement
-30
1.6
1.8
2.2
2.4
2.6
Frequency (GHz)
Lower band impedance bandwidth: 15.5% (1.70-2.00 GHz)
Upper band impedance bandwidth: 3.7% (2.39-2.48 GHz)
88
30 o
0 o
90 o
-40
-30
30 o
0 o
60 o
-20
-10
60 o
60 o
90 o 90 o
-40
-30
-20
-10
dB
120 o
120 o
90 o
0 o
Cross-pol
60 o
90 o
-40
Col-pol
30 o
-30
60 o
-20
-10
120
120 o
150 o
150 o
180 o
-y
+y
H-plane (y-z plane)

Measurement
Simulation
(a)
0 o
30 o
Cross-pol
60 o
60 o
90 o 90 o
-40
-30
-20
-10
120 o
120 o
120
150 o
150 o
90 o
dB
120 o
150 o
150 o
180 o
180 o
-x
+x
E-plane (x-z plane)
30 o
dB
dB
150 o
150 o
30 o
30 o
Cross-pol
Cross-pol
60 o
Col-pol
30 o
180 o
-y
-x
+x
E-plane (x-z plane)
+y
H-plane (y-z plane)

Measurement
Simulation
(b)
HEM111 mode: measured (1.8 GHz), simulated (1.8 GHz)

Co-polarized fields > cross-polarized fields by more than 20 dB in the
89
boresight direction.
Measured and simulated gain

Lower band gain (dBi)
6
Upper band gain (dBi)

1010
88
66
44
22
00
2.3
2.3
2.35
2.35
HFSS Simulation
Measurement
0
1.6
1.65
1.7
2.4
2.4
2.45
2.45
2.5
2.5
2.55
2.55
Frequency (GHz)
1.75
1.8
1.85
1.9
Frequency (GHz)
HEM111 mode: Maximum measured gain of ~6 dBi (1.75 GHz)

HEM113 mode: Maximum measured gain of ~ 8 dBi (2.43 GHz)
90
Dualband CP DRA
Ground
plane
Excitation strip
w
Dualband
Cylindrical quadrature To
grounding
coupler
DRA r
via
W
Cylindrical
DRAr
Matching a
slot
Excitation
strip
Via W3
To grounding
via
DS
L2
L1
W2
Ground plane
W0
W1
Isolation
port
L3
Input port
Top view
Aperture
for via
LS
Via
Feedline
Matching slot
Side view
a = 18.7 mm, h = 42.5 mm, r = 9.4, l = 12.5 mm, w = 1 mm, Ls = 21 mm, Ws = 1.5 mm, Ds =
12.75 mm, L1 = 26.9 mm, L2 =26.5 mm, L3 = 56.65 mm, W0 = 4.66 mm, W1 = 7.3 mm, W2 =
4.44 mm, and W3 = 0.46 mm.
91

0
-10
-20
HFSS Simulation
Measurement
-30
-40
1.6
1.8
2.2
2.4
Frequency (GHz)
Lower band bandwidth:18.9% (1.58-1.91 GHz).
Upper band bandwidth:7.8% (2.33-2.52 GHz).
2.6
Measured and simulated axial ratios (ARs)

10
Lower band AR (dB)

88
Upper band AR (dB)
66
44
22
00
2.3
2.3
2.5
2.5
2.4
2.4
Frequency (GHz)
2
0
1.6
HFSS Simulation
Measurement
1.65
1.7
1.75
1.8
1.85
1.9
Frequency (GHz)
Lower band AR bandwidth: 12.4% (1.67-1.89 GHz)
Upper band AR bandwidth: 7.4% (2.34-2.52GHz)
93
30 o
0 o
LHCP
30 o
30 o
60 o
-40
-30
-20
-10
60 o
90 o 90 o
-40
-30
120 o
120 o
0 o
LHCP
30 o
-20
-10
90 o
60 o
90 o
-40
-30
150 o
120 o
150 o
150 o
180 o
180 o
-y
-x
+x
E-plane (x-z plane)
+y
(a)
30 o
-20
-10
60 o
60 o
90 o 90 o
-40
-30
-20
-10
dB
120 o
120 o
120
150 o
150 o
90 o
dB
120 o
150 o
150 o
180 o
H-plane (y-z plane)

Measurement
Simulation
0 o
RHCP
60 o
dB
120
30 o
RHCP
60 o
dB
150 o
30 o
30 o
RHCP
RHCP
60 o
90 o
0 o
180 o
-y
-x
+x
E-plane (x-z plane)
+y
H-plane (y-z plane)

Measurement
Simulation
(b)

LHCP fields > RHCP fields by ~20 dB in the boresight direction.
94
B. Example for wideband cylidnrical DRA design
Given: f1 = 2.90 GHz, f2 = 3.72 GHz, r= 9.4

Using formula
(5) & (6)
a = 10.3 mm & h = 34.3 mm
95
Wideband LP cylindrical DRA

Configuration
Cylindrical
DRA r
Ground plane
-10
-20
Conducting
feeding strip
-30
Coaxial
aperture
y
x
HFSS Simulation
Measurement
-40
2.8
3.2
3.4
3.6
3.8
4.2
Frequency (GHz)
a = 10.3 mm, h = 34.3 mm, r = 9.4,

l = 12 mm, and w = 1 mm.
Good agreement
Measured impedance bandwidth:
23.5% (3-3.8 GHz)
96
Measured and simulated gain

Antenna Gain (dBi)
12
10
8
6
4
HFSS Simulation
Measurement
2
0
3.2
3.4
3.6
3.8
Frequency (GHz)

97
Wideband CP cylindrical DRA

x
Excitation strip
w
Excitation strip
Cylindrical
Ground
DRAr
plane
Excitation
strip
via
Wideband
quadrature
coupler
Cylindrical
DRAr
via
Aperture
for via
L1
L1
W0
Ground plane
W1
W0
Input port
Top view
Isolation
port
Via
Wideband quadrature coupler
Side view
a = 10.3 mm, h = 34.3 mm, r = 9.4, l = 11.5 mm, w = 1 mm,

L1 = 14.67 mm, W0 = 1.94 mm, and W1 = 3.21 mm.
98
Wideband CP DRA
Axial ratio
Axial ratio (dB)
-10
HFSS Simulation
Measurement
4
-20
HFSS Simulation
Measurement
-30
3
3.2
3.4
3.6
3.8
Frequency (GHz)
Measured impedance bandwidth:

25.5% (3.04-3.93 GHz).
3.2
3.4
3.6
Frequency (GHz)
3.8
Measured 3-dB AR bandwidth :

24.7% (3.05-3.91 GHz).
99
VI. Dualfunction DRAs
Advantage
System size and cost can be reduced by
using dualfunction DRAs.
Additional functions
- Packaging cover
- Oscillator
101
Packaging Cover
102
Conventional
Dielectric resonator
antenna/oscillator
load and packaging
cover
Proposal
Aperture
Metallic supports
for grounding
Microstrip
feedline
Transistor (other
Metal ground
RF components
not shown)
Power supply
Front view
103
Antenna Configuration
Dielectric resonator antenna
and packaging cover
Aperture
z
Metallic supports
for grounding
Resonant frequency
f0 = 2.4GHz
Metal ground
Feedline and the

RF/MIC circuits
Coaxial line
Side view
L
a
Aperture
b
y
Microstrip
feedline
Coaxial
Top view
Parameters:
Hollow DRA:
L=30mm, W=29mm,
H=15mm, & r =12
Metallic Cavity:
a = 15mm, b = 21.6mm, h = 5mm
Top face : Duroid r =2.94
thickness 0.762mm
Aperture: 0.2063 e
104
Design Procedure (Simulation):

Step 1
Step 2
Use the DWM to design

a solid rectangular DRA
at 2.4-GHz fundamental
TE111 Mode.
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 f0 )
z
x
105
Experimental Verification:
- Hard-clad foam (r 1) is used to form the
container.
- ECCOSTOCK HiK Powder of r =12 is used as
the dielectric material.
106
Return Loss and Input Impedance

(Passive hollow RDRA with a metallic cavity)
0
-20
80
Input Impedance (ohm)
Return Loss (dB)
-10
-30
Theory
-40
Experiment
-50
60
40
20
0
-20
-40
-60
2
2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9
Frequency (GHz)
-60
2
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
Frequency (GHz)
Good agreement.
Bandwidth ~ 5.6%.
Measured resonance frequency: 2.42GHz (error < 0.83%)
107
Radiation Patterns
(Passive hollow DRA with a metallic cavity)
co-pol
co-pol
0
0
30
-30
-60
-30
30
cross-pol
cross-pol
60
-90
-40
-30
-20
-120
-10
0 90
120
-150
-60
-90
-40
-30
-20
-120
150
-10
0 90
120
-150
180
x-z plane
60
150
180
Theory
y-z plane
Experiment
(a)
(b)
Broadside TE111y mode is observed.

Co-polarized fields generally stronger than the crosspolarized fields by 20dB in the boresight direction. 108
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 50 at the input.
A small hole is drilled on the ground plane to supply the DC bias to the LNA.
Return Loss (dB)
-10
-20
DRA
DRA
DRA (passive)
-30
A
Return
Loss
Receiver
-40
2
2.1
DRA (active, receiver)
Return
Loss
DRA (active, transmitter)
Transmitter
2.2
2.3
2.4
2.5
2.6
Frequency (GHz)
2.7
2.8
2.9
3
109
Amplified Radiation Pattern

0
-10
Amplitude (dB)
-20
-30
-40
co-pol (passive)
-50
co-pol (active)
-60
-70
-80
-180
cross-pol (passive)
cross-pol (active)
-120
-60
60
120
180
Observ ation angle (degree)
Compared to the passive DRA, the active DRA has a gain of

7 - 12dB across the observation angle from -90o to 90o.
The gain is less than the specification due to unavoidable
impedance variations and imperfections in the measurement.
110
Dielectric Resonator Antenna

Oscillator (DRAO)
Methodology
The DRA is used as the oscillator load,
named as DRAO.
The reflection amplifier method is used to
design the antenna oscillator.
112
DRAO Schematic Diagram

Z =R +X
L
Lm
Transistor
jX
Z =R +X
in
in
in
DRA
- Oscillate condition: XL+Xin=0 & RL<|Rin|

- DRA first replaced by a 50 load at 1.85GHz.
113
Antenna Configuration:
Dielectric resonator
antenna and
oscillating load
Resonance frequency
fo = 1.85GHz at TE111y
x
Ground
H
d
Substrate
Microstrip
feedline
Transistor (other
RF components
not shown)
Aperture
Side view
Transistor (other
RF components
not shown)
L
Wa
Microstrip
feedline
Aperture
w
La
Lm
Ls
Top view
Parameters:
DRA
L=52.2mm,
W=42.4mm,
H=26.1mm,
r = 6.
Aperture
La = 0.3561e, Wa = 2mm
Ls = 9.5 mm, Lm = 40 mm.
Duroid substrate
rs=2.94, d=0.762mm
114
Return Loss and Input Impedance

Return Loss (dB)
0
-10
-20
Input Impedance (ohms)
80
-30
60
40
20
-40
0
-20
-50
-60
1.6
HFSS Simulation
Experiment
1.7
1.8
-40
-60
-80
1.6
1.9
Frequency (GHz)
1.7
1.8
1.9
Frequency (GHz)
2.1
2.1
2.2
Good agreement.
Bandwidth ~ 22.14%.
Resonance frequency: Measured 1.86GHz
Simulated 1.83GHz (1.5% error).
115
Spectrum of the Free-running DRAO
Transmitting power Pt = 16.4dBm

DC-RF efficiency: ~ 13% (2-25% in the literature).
Phase noise: 103dBc/Hz at 5MHz offset
Second harmonic < fundamental by 22dB
116
Radiation Pattern
co-pol
co-pol
-30
30
-30
cross-pol
cross-pol
-60
-60
60
-90
-40
30
-30
-20
-10
120
-120
150
-150
90
60
-90
-40
-30
-20
-10
90
120
-120
150
-150
180
180
x-z plane
y-z plane
HFSS Simulation
Solid DRAO (measur.)
Passive Solid DRA (measur.)
(a)
(b)
Broadside TE111y is observed.

Co-polarized fields are generally 20dB stronger
than the cross-polarized fields in the boresight direction.
117
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?
Yes!
It leads to probably the most beautiful antenna in
the world .
118
Glass-Swan DRA
Distinguished Lecture
Transparent antennas: From 2D to 3D

119
Conclusion
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.
120
121
Q&A
122

Development of Dielectric Resonator Antenna 2

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Development of Dielectric Resonator Antenna 2

Uploaded by

Copyright:

Available Formats

Development of Dielectric

Resonator Antenna (DRA)

What is Dielectric Resonator Antenna (DRA) ?

The DRA is an antenna that makes use of a radiating mode

Advantages of the DRA

Corporate feedline for DRA array

Slot-fed DRA array using corporate

Proposed Antenna Geometry

Waveguide Model (DWM)

Resonant frequency of TEmnl(y) mode

Time-Domain (FDTD) method

Source model and extraction of S parameters

Source occupies only one grid

total time steps : 10000

Input Impedance (ohm)

Comparison between Theory and Measurement

Field Distribution --- TE111y

Imaged DRA (gound plane removed)

With gound plane

Field Distribution --- TE112y

Imaged DRA (gound plane removed)

Field Distribution --- TE113y

Imaged DRA (gound plane removed)

With gound plane

-40 -30 -20-10 0

Broadside radiation patterns are observed.

III. Circularly Polarized Design

Proposed Antenna Geometry

Conducting Parasitic Patch)

Axial Ratio in the boresight direction

Axial Ratio (dB)

3-dB AR bandwidth is ~ 2.7%, which is a typical value for

The H field of the DRA without and with parasitic

With parasitic strip - CP field

Radiation Patterns (f = 3.4GHz, )

A broadside radiation mode is observed.

Effects of feeding strip length l1

Axial Ratio (dB)

Input impedance changes substantially with l1.

II. Frequency Tuning Technique

Frequency Tuning Technique

The slot-coupled DRA with a conducting loading

Hemispherical DRA: radius a = 12.5 mm, dielectric constant r = 9.5.

Calculated and measured return losses

Calculated and measured radiation patterns

3.25 GHz ( = 26.38o and Lt = 4.42 mm)

The effect of loading

3.58 GHz ( = 52.8o and Lt = 13.6 mm)

Stub Length Lt (mm)

Calculated and Lt for having a good return loss

Tuning frequency fr (GHz)

The resonant frequency can be tuned by varying and Lt

Impedance Bandwidth (%)

The bandwidth decreases after a loading cap is added.

Frequency Tuning Technique

The annular-slot-excited cavity-backed DRA

(a) Side view

(b) Top view

IV. Omnidirectional Circularly

Advantages of omnidirectional CP antenna

CP DRAs concentrated on broadside-mode designs only.

Slotted omnidirectional CP DRA

Dielectric cube with oblique slots (polarizer) fabricated on

Dielectric block with the wave