1.

423
STUDY AND DESIGN OF COMPACT WIDEBAND MICROSTRIP ANTENNAS
F. Carrez, J . Vindevoghel
Institut d'Electronique et de Microelectronique du Nord (I.E.M.N), France
INTRODUCTION
'I'he emergence of wireless communication systems and
iradar applications requires compact antenna structures easy
to manufacture. Microstrip antennas are suitable for these
applications because they have many advantages: high
performances, thin profile, low cost. However, applications
of microstrip antennas are limited mainly because of their
inarrow bandwidth. Furthermore, the conventional array
antennas associated with the feeder network require a large
:jize and suffer fi-omunwanted mutual coupling between
he feeder network and the resonators which reduces the
efficiency of the array antenna. In this paper, we propose a
design of linear and 2D array antennas which leads to a
!sizereduction and high performances. A short description
of the analysis method is exposed and an emphasis on the
experimental results of these antennas operating in the X-
lband are presented to validate this design. Moreover, the
techniques employed for this design are well-adapted for
the monolithic conception of millimetre waves band active
antennas.
ANTENNA CONFIGURATIONS
'Theantennas presented in this study are shown in Fig. 1.
'They have a two-layer structure and consist in a n,X ny
of rectangular patches of width W (radiation edge) in
ihe x direction and of length L in the y direction. The
]patches working in the resonant mode TMo, are etched on a
low-loss dielectric substrate with a relative permittivity
,~,,=2.35 and a thickness hl=0.762 mm. To avoid any
interaction between the resonators and the microstrip
feedline, this last is built on a high permittivity substrate
.with a relative permittivity and a thickness h2 of 10.8
;and 0.635 mm, respectively, on the opposite side of the
,antenna. Contrary to the conventional array antennas which
(avoid the coupling between the resonators, the patches are
:strongly coupled in the E-plane or in the H-plane in the
(case of the linear array antennas and are referred to EPLA
,and HPLA. The two dimensional array antenna (2DA)
(consists of rectangular patches which are strongly coupled
inthe different planes of the antenna. To feed the patches
two techniques are used: the driving patch is fed by
electromagnetic coupling to a microstrip line through a
slot-aperture in the common ground plane. This technique
of feeding commonly called Aperture Coupled Patch
Antenna (ACPA), presents many advantages already
detailed in the litterature (1) and are not mentioned hereby.
This driving patch is loaded by parasitic ones and form the
array of the antenna. These parasitic patches are
electromagnetically influence-coupled and involve an
increase in the bandwidth as well as in the directivity.
These techniques were also employed to design a compact
two-port microstrip antenna. Its description and
performances are reported in (2).
ANALYSIS METHOD
The first step of the analysis consists of the modelling of
the isolated aperture coupled driving patch. The
rectangular microstrip resonator is probably the most
commonly used microstrip antenna and several methods
are available far its design. The aperture is one of the most
important part in the structure of the array antennas
described above. The dimensions of the aperture must be
wide enough to provide a sufficient coupling between the
resonator and the transmission line but small enough to
avoid a perturbition of the fields inside the resonant cavity.
Furthermore, the size of the aperture has also an influence
on the resonant fi-equency of the resonator and can be
reduced within about 20%. In our study a transmission line
method was employed to design the ACPA. The
formulations ofthis method have already been detailed in
(3) and good jpredictions were obtained with a gain of
computing time compared to a full wave approach.
The next step deals with the modelling of mutual coupling.
To perform the calculation of the mutual impedance, the
following procedure was used. We have first calculated the
electric field inside the resonant cavity inthe case of the
parasitic patch and the aperture-coupled driving patch.
Then, using the equivalence principle, the equivalent
magnetic current M at the boundary cavity and the
magnetic field H were determined. Finally, the mutual
impedance was calculated fkom the reaction theorem
introduced by Rumsey (4):
z, =- I f H, . G, . d l
where C, is the contour of the antenna J, li and 'j are the
currents supplied to patches I and J, Hi is the magnetic
field of patch 1 and Mj is the equivalent magnetic current
source of patch J. A good accuracy was obtained even for
strong coupling (>-lo dI3) until the inner field distribution
of the resonator is not disturbed.
I l . I , c,
10th International Conference on Antennas and Propagation, 14-17 April 1997, Conference Publication No. 436 0 IEE 1997
1.424
The final step concerns the modelling of the array antenna.
It can be simply made by considering the antenna as a M-
port network and the procedure described in ( 5) was
followed. This method which accounts for all the coupling
effects, enables us to compute the input impedance of the
antenna and the induced currents in every resonator.
Therefore, the radiation patterns can bealso determined.
This analysis method employing a transmission line model
and matrix algebra provides ease of computing ; good
agreements were obtained between the simulated and the
measured datawith a gain of computing time compared to
a full wave approach.
RESULTS
Using the analysis methods described hereabove, the
proposed linear and 2D array antennas have been studied
and realized to operate in the X-band. The dimensions of
the resonators are taken as L=8.5 mmand Wl 1.3 mmand
their resonant frequency is 10.2 GHz. The dimensions of
the slot-aperture centred under the central resonator are
W,=0.5 mm and L,=4.0 mm. The feedline has a
characteristic impedance of 50 SZ and the length L, of the
stub is adjusted to obtain a good matching for the different
antennas.
We have first investigated the effects of the coupling on the
bandwidth in the case of linear array antennas. Fig. 2
shows the simulated results of the linear array antenna with
resonators coupled in the H-plane for different gap widths
S,. Wecan note that the coupling in the H-plane tends to
increase the bandwidth of the antenna towards frequency
lower than the resonant fiequency of the isolated ACPA
with a shift of the 50 Q input impedance. The shift of the
50 R input impedance and the increase of the bandwidth
are due to the influence of the two additional resonators
which involve a variation of the input impedance of the
central resonator. We obtain a maximumbandwidth of 12
YO for a VSWR 5 2:l. The measured results presented in
Fig. 3 confirm this observation except in the case of a gap
width S, =0.1 111111. In this case, the distance between the
non-radiating edges of the resonators is very small and
consequently, the mutual interaction disturbs the inner field
distribution and leads to a modification of the radiating
properties of the resonators. The same observations can be
made in the case of resonators coupled in the E-plane, as
seen in Fig. 4. When the distance between the radiating
edges of the resonators is not too small (S,=l.O mm), the
bandwidth of the antenna also increases towards the lower
fiequencies (14 % for a VSWR <2:1), whereas a high
coupling between the resonators (Sy=O.l mm) implies a
shift of the resonant fiequency without an increase of the
bandwidth. In this case, the distance between the radiating
edges is smaller than the extension of the fiinging field of
the resonators. Therefore, the radiating edges at the open-
circuited ends of the patch are altered and imply this shift
of the resonant frequency of the array antenna.
A two dimensional array antenna with resonators gap-
coupled in the two planes of the ACPA have also been
investigated. The distance between the resonators gap-
coupled in the E-plane has been kept wide enough to avoid
any perturbation of the fi gi ng field and was set to Sy=l .O
mm. Fig. 5 shows the evolution of the VSWR variations
for different gap widths S,. Here again, an increase of the
bandwidth is obtained when the distance between the
resonators is decreased. As expected, the bandwidth is
wider than the bandwidth of the linear array antennas with
the same distance between the resonators. The different
resonant fiequencies of the resonators in the E and H-
planes due to the mutual interactions cause a higher
increase in the bandwidth of the array antenna compared to
a linear array with resonators coupled in only one plane. A
maximumbandwidth of 1.5 GHz was measured (VSWR L.
2: 1) for S,=O. 1 mm.
The influence of the coupling between the resonators on
the radiation patterns has been also studied. The radiation
patterns were measured at frequency where the gain is
maximum.
Fig. 6 presents the radiation pattem in the E-plane of the
linear array with resonators coupled in the E-plane. For a
gap width S, of 1 .O mm, the directivity of the array antenna
increases and a gain of 8.6 dE3is obtained at F=10.0 GHz.
For a gap width S,=O.l mm, a gain of 9.2 dB is measured
at F=l 1.0 GHz and side lobes appear with a level of -8.5
dE3. In the case of the linear array antenna with resonators
coupled in the H-plane, wecan also note an increase of the
directivity without side lobes as shown in Fig. 7. A gain of
8.2 dB and 7.7 dB was measured at F=10.0 GHz for a gap
width S, of 0. lmm and 1.6 mm, respectively.
The radiation patterns of the 2DA have been measured and
Fig. 8 presents the H-plane radiation pattem. In this case
again, the directivity of the array antenna increases when
the distance between the resonators decreases and a
maximumgain of 1 1.8 dB is obtained for the gap widths S,
and S, of 1.0 mm and 0.2 mm, respectively. The E-plane
radiation pattem is not reported here, but it was found that
the directivity slightly increases without side lobes and the
variation of the coupling in the H-plane does not affect the
E-plane radiation pattem.
An analysis of the radiation pattem of the 2DA has been
also performed for different fiequencies included in the
bandwidth. Fig. 9 shows the H-plane radiation pattem of
the array antenna with gap widths of S,=0.4 mm and
S,=l.O mmat F=9.25 GHz, F=10.0 GHz and F=10.25
GHz. The directivity increases and side lobes appear when
the fkequency increases until F=10.25 GHz and then the
directivity decrease. The presence of the side lobes at
fi-equencies close to 10.2 GHz are due to the different
resonant frequencies of the resonators. Although the
resonators have the same dimensions, the resonant
fiequency of the central resonator is lower than the lateral
ones. The resonant fkequency of the lateral resonators is
10.2 GHz whereas the resonant fiequency of the central
aperture coupled resonator is 10.0 GHz. As exposed
before, the feeding structure tends to decrease the resonant
10th International Conference on Antennas and Propagation, 14-17 April 1997, Conference Publication No. 436 0 IEE 1997
1.425
frequency of the central resonator. Therefore, when the
6requency increases, the influence of the lateral resonators
becomes greater and involves the presence of side lobes in
the radiation pattern.
The crosspolar measurements are not reported here but
have been found to be lower than -25 dB for each antenna.
Compact linear and 2D array microstrip antennas have
been studied and developed. These antennas have a two-
layer structure and the radiating array consists of
electromagnetically coupled resonators. The techniques
used for the design provide a wide bandwidth, an increase
of the directivity, a high efficiency with a small size
compared to conventional may antemas. The design of
these compact antennas is also well-suited for the
monolithic fabrication of microwave active antennas used
fix short range communication or
require some bandwidth. Moreover, ckcular polarizaliorms
Cim also be obtained by employing square resonators and a
cross-aperture feeding structure.
Pozar D. M., 1985, "Microstrip antenna aperture-
coupled to a microstrip line", Electron. Lett., =, 49-50
Carrez F., Vindevoghel J ., 1996, "Compact TWQ pori
microstrip antenna", Electron. Let t . , 32, 1337-133
Himdi M., Daniel 9. P., Terret C., 1989, "Transmission
line analysis of aperture-coupled microstrip antenna",
Electron. Lett., 25, pp. 1229-1230
Rumsey V. H., 1954, "Reaction concept in
eiectromagnetic theory", Phvs. Rev., 94, pp. 1483-
1491
James J . K. and Hall P. S. (Editors), 1989, "Handbook
of Microstrip Antenna", Peter Perewus Ltd, Vol. 1,
London, UK, pp. 434-436
a. Explosed view of the linear array antenna with
resonators coupled in the E-plane (EPLA).
microstripline
b. Linear array antenna with resonators coupled in the H-
plane (WLA).
E-ulane
micros triplin e
c. 2D array antenna with resonators (2DA).
Fig. 1 : Geometry of the studied array antennas.
10th International Conference on Antennas and Propagation, 14-17 April 1997, Conference Publication No. 436 0 IEE 1997
5
4
g 3 - Sx4.1 mm
k.
- Sx4.2 mm
2 - Sx4.4 mm
8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0
frequency (GHz)
Fig. 2: Simulated VSWR variations of HPLA for
different gap widths S,.
5
4
a
k
E 3
2
I
- Sx4. l mm
- Sx=1.6 mm
I
8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0
frequency (GHz)
Fig. 3: Measured VSWR variations of HPLA for
different gap widths S,.
5
4
a
E 3
2
1
8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0
frequency (GHz)
Fig. 4: Measured VSWR variations of EPLA for
different gap widths S,.
5
4
a
5 3
- Sx=O.l mm
+Sx=0.2 mm
- Sx=0.4 mm
2
1
8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0
frequency (GHz)
Fig. 5: Measured VSWR variations of the 2DA for
different gap widths S,, Sy=l .O mm.
1.426
0
g -5
v
L
E
8 -10
U
I
.-
id
-
2 -15
- ACPA - Sy-O.1 mm -- Syl . 0 mm
-20
-90 -60 -30 0 30 60 90
angle (deg.)
Fig. 6: Measured E-plane radiation pattem of EPLA
for different gap widths S,.
0
-5
-10
-15
6
aJ
W
- Sx4.l mm
- Sx=1.6 mm
.-
; -20
-
0
-25
-30 1
-90 -60 -30 0 30 60 90
angle (deg.)
Fig. 7: Measured H-plane radiation pattern of HPLA
for different gap widths S,.
0
h
E9 -5
t
L
E
.- P
e -15
-10
id
-
- Sx4.l mm
- Sx4.2 mm
- Sx4.4 mm
\,
-90 -60 -30 0 30 60 90
angle (deg.)
Fig. 8: Measured H-plane radiation pattern of 2DA for
different gap widths S,, Sy=l .O mm.
0
-5
-10
E-15
is
E
Y
; -20
.-
-
W
-25
- Fz9.75 GHz
- F=lO.OO GHz
+F=10.25 GWz
-30 - , a , -
-90 -60 -30 0 30 60 90
angle (deg.)
Fig. 9: Measured H-plane radiation pattern of the 2DA
at different frequencies. S,,=0.4 mm, S=l.O
mm.
10th Intemationali Conference on Antennas and Propagation, 14-17 April 1997, Conference Publication No. 436 0 IEE 1997