Bowtie Patch Antennas and Simple Arrays for Wireless Indoor Communications Sener Uysal,

Transcription

Bowtie Patch Antennas and Simple Arrays for Wireless Indoor Communications Sener Uysal,
738
IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 47, NO. 6, JUNE 1999
Bowtie Patch Antennas and Simple Arrays
for Wireless Indoor Communications
Sener Uysal, Member, IEEE, Mook-Seng Leong, Senior Member, IEEE, and Chee Hong Ng
Abstract— Several bowtie patch-antenna configurations are
studied for their suitability for use in broad-band indoor wireless
communications. A microstrip bowtie antenna (MBA), based on
the design of equilateral triangular patches, is first designed
and tested. The same design is used to realize a coplanarwaveguide bowtie antenna (CPWBA) with finite ground plane.
The CPWBA is coax-fed from its apex and matched in C band. The resonant slot length of this antenna is around three
times that of the guided wavelength. The measured gain for the
CPWBA antenna is 8.1 dB. The same MBA is also used in the
realization of (1 2 2)-, (2 2 1)-, and (2 2 2)-element MBA arrays.
The MBA and resultant arrays use microstrip feed networks
matched to the input impedance at around 10 GHz. The gain
of the MBA is 6.2 dB; the gains for the arrays vary within 13.7
and 17.3 dB. The 2 : 1 voltage standing-wave ratio bandwidths
lie in the 9.7%–10.8% range for the realized antennas. The
radiation patterns can be optimized to fit the required diversity
for the specific indoor wireless communication environment; the
beamwidths are demonstrated to vary between 15 and 85 ,
which allow for multipath minimization and radiation diversity
within the premises, thereby providing interference-free illumination.
Index Terms— Bowtie antennas, microwave antennas, simple
array antennas, wide-band antennas, wireless indoor communications.
I. INTRODUCTION
I
NDOOR wireless radio communication allows easy setup,
reconfigurability, portability, and mobility for phones, terminals, and computers connected within the network. Transmission reliability can also be significantly improved as some
of the network failures are due to wired connections. However,
high bit-rate transmission is not easy to achieve with wireless
indoor communications as compared to the wired connections
due to the multipath propagation effect. The resulting delay
spread experienced at the receiver results in intersymbol
interference when the delay spread exceeds 1/10 of a symbol
period.
A higher bit rate can be achieved by employing signal
processing and architectural modification of the system components [1]. The latter approach involves reconfiguration of
transmitter–receiver separation or elimination of reflected rays.
The delay spread can be minimized in a number of ways,
as reported in the literature. Among those techniques are: 1)
using circular polarization instead of linear polarization with
directive antennas [2], [3]. The handedness of the circularly
Manuscript received December 1, 1998; revised February 1, 1999.
The authors are with the Department of Electrical Engineering, National
University of Singapore, Singapore 119260.
Publisher Item Identifier S 0018-9480(99)04281-7.
polarized reflected wave is reversed and employing similar
transmit and receive antennas will help to reject multipath
components after a single reflection. As channel degradation
is primarily caused by singly reflected waves, this approach
effectively suppresses the multipaths; 2) using a distributed
antenna system [4], in which case, the user is always close to
an antenna. This may lead to installation of a base-station
antenna to the ceiling of every room within the premises;
and 3) using directive antennas with more-or-less line-ofsight transmission. With narrower antenna beams, most of the
signal delay paths are eliminated. Driessen [5] has reported
a successful implementation of this approach by using directional antennas with 15 beamwidth to reduce the multipath
components.
A composite notch-wire antenna for polarization diversity
reception in an indoor base-station system has been reported
by Kuga et al. [6]. A detailed study of the hostile indoor
environment, path loss, and delay spread using the ray-tracing
method have also been reported in the literature [7], [8]. An
interesting paper by Obayashi and Zander [9] reports on the
body-shadowing model for indoor wireless communications
using ray-tracing and imaging methods.
Multiple-beam antennas can be used to illuminate multiple
receiving terminals in an indoor environment. Such a radiation
diversity can significantly reduce multipath transmission by
having narrower beamwidths. Body-shadowing effects can
also be significantly reduced or eliminated by employing
similar receiver antennas and using space diversity for the
indoor base station. A detailed study of bowtie patch antennas
and simple arrays using the electromagnetic simulator Ensemble (Boulder Microwave Technologies Inc., Boulder, CO)
and experiments are presented for broad-band wireless indoor
communications. A rigorous moment solution [10] exists for
the analysis of bowtie antennas, which takes a long time for
the complete solutions. Instead, a simple but very accurate,
technique will be used, which is briefly explained in Section II.
The design of a coplanar-waveguide bowtie antenna (CPWBA)
at -band (developed for a 5.5-GHz 20-Mb/s transmit/receive
radio modem for wireless indoor communications), 2-, and
4-element microstrip bowtie antenna (MBA) arrays with favorable radiation characteristics are presented in this paper.
II. DESIGN
OF THE
CPWBA
Generally, the features of coplanar waveguide (CPW) antennas are similar to those of microstrip antennas. Both of
them can be considered as open-circuited transmission-line
structures with cross-sectional expansion. By drawing an anal-
0018–9480/99$10.00  1999 IEEE
UYSAL et al.: BOWTIE PATCH ANTENNAS AND SIMPLE ARRAYS FOR WIRELESS INDOOR COMMUNICATIONS
739
Fig. 1. (a) Layout of CPWBA fed by a 50-
SMA coaxial feed on Taconic substrate of "r = 2:2 and h = 0:79 mm. (b) Measured return loss (0–20 GHz)
for CPWBA. (c) Measured radiation pattern for CPWBA in the = 0 plane (E ). (d) Measured radiation pattern for CPWBA in the = 0 plane (E ).
(e) Measured radiation pattern for CPWBA in the = 90 plane (E ). (f) Measured radiation pattern for CPWBA in the = 90 plane.
ogy between them, we are inspired to design a CPWBA
by using the same dimensions from the designed microstrip
bowtie [11], [12] in order to maintain the same aspect ratios.
The expected resonant frequency would be close to that of
the single microstrip equilateral half since CPW antennas are
typically 75% more compact than microstrip antennas.
A finite-ground CPWBA was designed using the equations
given for an equitriangular microstrip patch [13], [14]. The
resonant frequencies for -independent TM modes satisfying
perfect magnetic-wall boundary condition for the equilateral
triangular patch based on the cavity-model theory is given by
(1)
are the order of various resonant modes,
is
where
the relative dielectric constant of the substrate, and and
are the dimensions indicated on Fig. 1(a). The side length
can be modified to include the nonideal magnetic-wall effect,
as reported by Helszajn et al. [15] and Dahele et al. [16] by
, with as the conductor thickness. Taking the
fundamental design frequency
to be 5.5 GHz, the following
mm, bow
dimensions are computed: side width
mm, and an optimum value of
mm.
length
mm and ground width
mm were
A slot width
used. The design parameters are illustrated in Fig. 1(a). The
expected resonance of CPWBA with the design dimensions
is in the -band. This design is presented because of its
inherent simplicity of matching to the 50- coaxial surface
mount adapter (SMA) connector in the -band. The return
loss was measured from 0 to 20 GHz and is shown in
Fig. 1(b). The measured response shows the desired wide-band
characteristics. The values of interest are noted as the resonant
GHz, bandwidth
%, and a
frequency
return loss of 31.8 dB.
Recently, Cai et al. [17], [18] proposed a coplanar waveguide antenna structure with finite dimension of the outside
conductor ground plane. The antennas in [17] were fed by
740
IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 47, NO. 6, JUNE 1999
SMA’s. The antenna is excited across the slot with the quasiTEM wave as the fundamental propagation mode. It was
claimed that resonance would occur when the perimeter of
the antenna slots is approximately equal to one (or integral
number of) guided wavelength in the slot. Interestingly, the
slot length measured in the case of CPWBA is three times
that of the guided wavelength , which is calculated from
formulas given in [19]. The average width of the bowtie
mm and
mm was used to determine
between
, which turned out to be 45.45 mm.
The radiation patterns measured showed similar properties
to those of the microstrip bowtie and the difference occurred
only in and below the direction of the antenna plane (endfire
direction), which was caused by the limited outer ground-plane
and
dimensions. The measured radiation patterns in the
90 plane are given in Fig. 1(c)–(f).
plane, both the
and
components
In the
are present and the radiation pattern for both components are
symmetrical about the 0 and 180 axis. For , a radiation
pattern shaped like a butterfly was observed. There are two
major lobes recorded above the 90 axis and two minor
)
lobes below it. A major dip in the broadside direction (
separates the two major lobes with peak power being measured
at 38.9 . The positions of minimum power were measured in
the direction of the antenna plane (endfire direction). The
component measured its maximum power ratio at broadside
with a half-power beamwidth of 38.3 . Besides the major lobe,
four almost similar minor lobes were observed: two side lobes
in the endfire direction and two minor lobes below the 90
axis.
plane, the measured
antenna pattern is
In the
almost symmetrical about the 0 and 180 axis with peak
power recorded at broadside and positions of minimum power
at 127.8 . A 3-dB beamwidth of 41.9 was measured in the
major lobe. A significant back lobe was also observed. For
, a “butterfly” radiation pattern was recorded. There is a
major dip at broadside with the positions of minimum power
at 111.6 . The positions of peak power were observed at
50.2 . The “butterfly” pattern obtained in this case is similar
component in the
plane. The
to that for the
difference lies in the average value of the power ratio measured
above the 90 axis, which is higher in this case compared
component mentioned earlier, i.e., the major lobes
to the
are broader. Comparatively, the two minor lobes below the
90 axis are less pronounced than their counterparts in the
plane.
III. DESIGN OF 1
2
AND
2
1 MBA ARRAYS
The arrays are simulated in Ensemble and fabricated on
,
mm) substrates (supported
Taconic (
by an aluminum block of the same size as the substrate)
response were
for measurement. The simulations for
done from 9.5 to 10.5 GHz over 51 frequency points to
give emphasis to the bandwidth over the resonant frequency,
whereas the measurements were made over the -band on
the HP8510C network analyzer.
2
Fig. 2. (a) Layout of 1
2 MBA array with spacing /2 (all dimensions
are in millimeters). (b) Simulated and measured return loss. (c) Simulated
radiation pattern at 9.88 GHz (
E component and
E component).
(d) Measured radiation pattern for E component at 9.88 GHz. (e) Measured
radiation pattern for E component at 9.88 GHz.
A. 1
1=
1=
2 MBA Array
2 MBA array with halfThe design layout for the 1
wavelength spacing is shown in Fig. 2(a). The measured return
loss for this array is compared with the Ensemble simulated
response in Fig. 2(b). From this figure, we can see that the
response are
shapes of both measured and simulated
similar. The resonant frequency of the simulated response is
at 9.92 GHz with a return loss of 39.6 dB. This is close to the
measured resonant frequency of 9.88 GHz having a return loss
of 30.8 dB. Another dip for both the simulated and measured
response is seen at 10.36 GHz ( 24.2 dB) and 10.26 GHz
( 28.3 dB), respectively. The measured bandwidth of 9.92%
is wider than the simulated 9.27%.
Similarly, the radiation pattern for this design was simulated
in Ensemble over 9.5–10.5 GHz. The simulated and measured
UYSAL et al.: BOWTIE PATCH ANTENNAS AND SIMPLE ARRAYS FOR WIRELESS INDOOR COMMUNICATIONS
741
measured return loss for this array is compared with the
Ensemble simulated response in Fig. 3(b). For this design, the
simulated and measured response agree well. The simulation
predicts a resonant frequency of 10.25 GHz with 39.8-dB
return loss. The measured resonant frequency is, however,
at 9.97 GHz, having a 35.1-dB return loss. The measured
response sees another dip of 29.8 dB at 10.2 GHz. The
measured bandwidth is 8.99%, which agrees well with the
simulated 9.2%.
The simulated and measured radiation patterns for the
and
components at 9.97 GHz are given in Fig. 3(c)–(e).
Both simulated and measured patterns show symmetry about
and
. At 9.97 GHz,
the 0 and 180 axis for
experiences a major dip at
and has multiple lobes at
component sees
broadside with no distinct main beam. The
and has a 3-dB beamwidth
a main beam directed at
of 18.9 . The sidelobes at 38.25 are, however, high, with
its level at about 3 dB below the main one.
C. 2
2
Fig. 3. (a) Layout of 1 2 MBA array with spacing 1 mm (all dimensions
are in millimeters). (b) Simulated and measured return loss. (c) Simulated
radiation pattern at 9.97 GHz (
E component and
E component).
(d) Measured radiation pattern for E component at 9.97 GHz. (e) Measured
radiation pattern for E component at 9.97 GHz.
1=
1=
and
components of the radiation patterns at the resonant
frequency of 9.88 GHz is as shown in Fig. 2(c)–(e). As seen
from the plots, the simulated radiation patterns agree well
with the measured patterns. Both the simulated and measured
patterns show symmetry about the 0 and 180 axis for
the - and -planes. At the resonant frequency of 9.88
component experiences a major dip at
,
GHz, the
component has
with multiple sidelobes at broadside. The
distinct lobes where the main beam is comparable to the
and a
sidelobes. The main lobe has its peak power at
.
beamwidth of 13.5 . The sidelobes are at
B. Compact 1
2 MBA Array
A compact array is obtained by reducing the element
spacing to 1 mm [its layout is shown in Fig. 3(a)]. The
1 MBA Array
response for the
Both the simulated and measured
1 MBA array, illustrated in Fig. 4(a), are as shown
2
in Fig. 4(b). The simulated response agrees well with the
measurements. The resonant dips are at 10.25 GHz ( 29.7 dB)
for the simulated response and 10.22 GHz ( 39.9 dB) for the
measured response. The measured bandwidth is 10.72%.
The simulated and measured radiation patterns for the
and
components at 10.22 GHz are given in Fig. 4(c)–(e).
Both simulated and measured patterns are relatively symmetric
about the 0 and 180 axis. At resonant frequency of
records a dip at
with maximum
10.22 GHz,
and
. The sidelobes
power ratios at
are 6 dB or more below the maximum power
at
displays a broad beam, recording a maximum power
ratio.
and a 3-dB beamwidth of 83.7 . Slight
ratio at
asymmetry is observed for 0 –180 due to interference of the
feeding network on one side of the circuit.
IV. DESIGN
OF
2
2 MBA ARRAY
Using the same MBA, the 2
2 array is realized for an
(15 mm). The layout for this design is
equal spacing of
as shown in Fig. 5(a). As this array is too large for simulation
in Ensemble, it is thus fabricated on Taconic (
mm) and measured. Measurement was made over the
-band on an HP8510C network analyzer. The measured
response is shown in Fig. 5(b). The resonant frequency is at
9.8 GHz with a return loss of 27.4 dB. The bandwidth of this
design is 10.7%.
and
comThe measured radiation patterns for the
ponents at 9.8 GHz are as illustrated in Fig. 5(c) and (d).
is relatively symmetric, whereas
is symmetric about the
0 and 180 axis. At 9.8 GHz, a major dip at broadside
) is observed for
.
also has multiple lobes
(
sees three distinct lobes and
and no distinct main beam.
. The main
the maximum power ratio is recorded at
742
IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 47, NO. 6, JUNE 1999
2
Fig. 4. (a) Layout of 2
1 MBA array with spacing =2 (all dimensions are in millimeters). (b) Simulated and measured S11 . (c) Simulated radiation
pattern at 10.22 GHz. (d) Measured radiation pattern for E component at 10.22 GHz. (e) Measured radiation pattern for E component at 10.22 GHz.
2
Fig. 5. (a) Layout of 2
2 MBA array with spacing =2 in horizontal and vertical directions (all dimensions are in millimeters). (b) Measured return loss
from 8 to 12 GHz. (c) Measured radiation pattern for E component at 9.8 GHz. (d) Measured radiation pattern for E component at 9.8 GHz.
UYSAL et al.: BOWTIE PATCH ANTENNAS AND SIMPLE ARRAYS FOR WIRELESS INDOOR COMMUNICATIONS
743
2
Fig. 6. (a) Layout of 1
2 MBA array with arbitrary spacing x. (b)
Simulated radiation pattern (simulations are at 10 GHz) for spacing =8
mm. (c) Simulated radiation pattern for spacing =4 mm. (d) Simulated
radiation pattern for spacing 3=8 mm. (e) Simulated radiation pattern for
spacing 5=8 mm. (f) Simulated radiation pattern for spacing 3=4 mm. (g)
Simulated radiation pattern for spacing 7=8 mm. (h) Simulated radiation
pattern for spacing mm.
lobe has a beamwidth of 14.85 . The sidelobes are high, only
.
about 2.5 dB below the main beam located at
V. 1 2- AND 2 1-ELEMENT MBA ARRAYS WITH
HORIZONTAL AND VERTICAL VARIATION IN SPACING
The radiation pattern can be optimized by adjusting the
configuration, mainly by changing the spacing between the
elements. This is investigated in this section by using a linear
1 2 MBA array with different element spacing and an offset
2 1 MBA array. For designs with three or more elements, the
PC used with 40-MB random access memory (RAM) running
at 200 MHz was not able to support the simulation.
A. Linear 1
2 MBA Arrays with Various Element Spacing
to in steps
The 1 2 MBA arrays for spacing of
are designed as shown in Fig. 6(a). The corporate
of
feed network is used as the array feed technique. The designs
and
were simulated in Ensemble at 10 GHz for both the
amplitude is found to
radiation patterns. The simulated
vary between 18 and 38 dB; the 2 : 1 VSWR bandwidth is
found to vary very little from 10%. The simulated radiation
patterns are given in Fig. 6(b)–(h).
2
Fig. 7. (a) 2
1 MBA array configuration used for the simulations at
10 GHz with horizontal spacing x and vertical spacing y (all dimensions
are in millimeters). (b) Radiation pattern for narrow x and y . (c) Radiation
pattern for x = =2 by y = =2. (d) Radiation pattern for x = 1=2 u by
narrow y . (e) Radiation pattern for x = u by narrow y . (f) Radiation pattern
for x = =4 by narrow y . (g) Radiation pattern for x = =2 by narrow
y . (h) Radiation pattern for narrow x by y = =4. (i) Radiation pattern for
narrow x by y = =2.
B. Offset 2 1 MBA Arrays with Various Vertical
and Horizontal Element Spacing
and a
The MBA array for a horizontal spacing of
vertical spacing of is designed as shown in Fig. 7(a). The
array feed network was not designed, as the objective is
to optimize the radiation pattern. Here, a probe feed is used
at each of the MBA elements to simulate the structure. The
designs were simulated in Ensemble at 10 GHz for both of
and radiation patterns. The 2 : 1 VSWR bandwidths
the
(not shown here) are found to vary between 10.5% and
11.1%; the radiation patterns are as shown in Fig. 7(b)–(i)
for different and combinations.
744
IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 47, NO. 6, JUNE 1999
VI. DISCUSSIONS
The antenna feed networks for the arrays were designed
using MDS by first generating a file for the input impedance
of the antenna and then matching this impedance to 50 using
a microstrip impedance transformer technique. This approach
of matching the antenna input impedance is sufficient for up to
around 10% bandwidth; for wider bandwidths more sections
are needed, which may lead to excessive microstrip losses
in the feed network and a significant increase in size and
radiation from the feed network. Stub matching can be used to
obtain wider bandwidths; alternatively, a double-layered feed
(electromagnetically coupled) [20] or aperture coupling [21]
can be considered.
The antenna gain was calculated by calculating the losses in
the measurement setup used to measure the radiation pattern. A
5–12-GHz anechoic chamber was used for the antenna pattern
construction. The transmitter antenna for the MBA and MBAarray measurements was a microwave horn with 16.35-dB
gain and 50% efficiency at 10 GHz. The transmitter for the
CPWBA was, again, a horn antenna with 14-dB gain and
50% efficiency at 5.5 GHz. The measurements were carried
out using an HP8530A microwave receiver connected to the
antennas via two 12.7-m-long cables with 3.8- and 8-dB loss
port of the receiver
at 5.5 and 10 GHz, respectively. The
was fed by a signal power sampled from the source power
by using a 10-dB directional coupler (has 1-dB maximum
insertion loss, including the coupled loss); the input power to
the directional coupler was 10 dBm. The port of the receiver
is connected to the antenna under test via the 12.7-m-long
cable. The antenna under test was oriented such that its apex
plane was in the
was directly facing the horn and the
azimuth.
The antenna gains were calculated using [22]
(2)
is the received power at the receiver antenna
where
is the input power and
(antenna under test) output,
is the gain of the transmit antenna (horn antenna), is the
free-space wavelength, and is the distance between the two
antennas. The calculated gains using the measurements with
m were: 6.2 dB for the MBA, 8.1 dB for the CPWBA,
13.71 dB for the compact 1 2 MBA array, 14.5 dB for the
-element spacing, 15.3 dB for the
1 2 MBA array with
2 1 MBA array, and 17.3 dB for the 2 2 MBA array.
VII. CONCLUSIONS
Several bowtie patch antenna configurations have been
studied for their suitability with respect to broad-band operation, adjustable beamwidth, and multiple-beam (radiation
diversity) capability for use in broad-band indoor wireless
communications.
We have demonstrated a design for a finite ground CPWBA
with favorable features. The presented antenna is compact,
allows easy integration with active devices, has wide bandwidth (10.6%), and has good radiation characteristics. The
CPWBA is coax fed from its apex and matched in -band. The
resonant slot length of this antenna is around three times that
of the guided wavelength. The measured gain for the CPWBA
antenna is 8.1 dB.
Corporate microstrip feed networks matched to the antenna
input impedance at around 10 GHz were used in the realization
of (1
2)-, (2
1)-, and (2
2)-element MBA arrays.
The gains for the arrays vary within 13.7 dB (1 2-element)
2-element). The 2 : 1 VSWR bandwidths
and 17.3 dB (2
lie in the 9.7%–10.8% range for the realized antennas. The
beamwidths are demonstrated to vary between 15 and 85 ,
which allow for multipath minimization. It was further demonstrated by simulations in this paper that the radiation pattern
can be further optimized by adjusting the configuration, mainly
the spacing between the elements. This can be achieved either
2)-element array or by an offset design of a
by a (1
(2 1)-element array. The simulations indicate that the latter
is more applicable, as it can provide higher gains and slightly
increased bandwidths.
REFERENCES
[1] L. J. Cimini, “Performance studies for high-speed indoor wireless
communications,” Wireless Personal Commun., vol. 2, no. 1–2, pp.
67–84, 1995.
[2] T. Manabe, Y. Miura, and T. Ihara, “Effects of antenna directivity and
polarization on indoor multipath propagation characteristics at 60 GHz,”
IEEE J. Select. Areas Commun., vol. 14, pp. 441–448, Apr. 1996.
[3] T. Manabe, K. Sato, H. Masuwaza, K. Taira, T. Ihara, Y. Kasashima, and
K. Yamaki, “Polarization dependence of multipath propagation and highspeed transmission characteristics of indoor millimeter-wave channel at
60 GHz,” IEEE Trans. Veh. Technol., vol. 44, pp. 268–273, May 1995.
[4] A. A. M. Saleh, A. J. Rustako, and R. S. Roman, “Distributed antennas
for indoor radio communications,” IEEE Trans. Commun., vol. COM-35,
pp. 1245–1251, Dec. 1987.
[5] P. F. Driessen, “Gigabit indoor wireless systems with directional antennas,” IEEE Trans. Commun., vol. 44, pp. 1034–1043, Aug. 1996.
[6] N. Kuga, H. Arai, and N. Goto, “A notch-wire composite antenna for
polarization diversity reception,” IEEE Trans. Antennas Propagat., vol.
46, pp. 902–906, June 1998.
[7] C.-F. Yang, B.-C. Wu, and C.-J. Ko, “A ray-tracing method for modeling indoor wave propagation and penetration,” IEEE Trans. Antennas
Propagat., vol. 46, pp. 907–919, June 1998.
[8] S. Y. Seidel and T. S. Rappaport, “A ray tracing technique to predict
path loss and delay spread inside buildings,” in Conf. Rec. IEEE
GLOBECOM, Orlando, FL, Dec. 1992, pp. 649–653.
[9] S. Obayashi and J. Zander, “A body-shadowing model for indoor radio
communication environments,” IEEE Trans. Antennas Propagat., vol.
46, pp. 920–927, June 1998.
[10] R. C. Compton, Z. Popovic, G. M. Rebeiz, D. B. Rutledge, R. C.
McPhedran, and P. P. Tong, “Bowtie antennas on a dielectric half-space:
Theory and experiment,” IEEE Trans. Antennas Propagat., vol. AP-35,
pp. 622–630, June 1987.
[11] K. W. Loi, S. Uysal, and M. S. Leong, “Novel design of a wide-band
microstrip bowtie patch antenna,” Proc. Inst. Elect. Eng., vol. 145, pt.
H, pp. 137–140, Apr. 1998.
[12] C. H. Ng, S. Uysal, and M. S. Leong, “Microstrip bowtie patch antenna
for wireless indoor communications,” presented at IEEE RAWCON’98,
Boulder, CO, Aug. 1998.
[13] S. A. Schelkunoff, Electromagnetic Waves. New York: Van Nostrand,
1943, ch. 10.
[14] J. R. James and P. S. Hall, Handbook of Microstrip Antennas, vol. 1.
Stevenage, U.K.: Peregrinus, 1989, pp. 151–169.
[15] J. Helszajn and D. S. James, “Planar triangular resonators with magnetic
walls,” IEEE Trans. Microwave Theory Tech., vol. MTT-26, pp. 95–100,
Jan. 1978.
[16] J. S. Dahele and K. F. Lee, “Experimental study of the triangular
microstrip antenna,” in IEEE AP-S Int. Symp. Dig., 1994, pp. 283–286.
[17] M. Cai, P. S. Kooi, and M. S. Leong, “A compact slot loop antenna,”
Microwave Opt. Technol. Lett., pp. 292–294, Apr. 1993.
[18]
, “A novel type of compact coplanar waveguide antenna,” Microwave Opt. Technol. Lett., pp. 349–351, May 1993.
UYSAL et al.: BOWTIE PATCH ANTENNAS AND SIMPLE ARRAYS FOR WIRELESS INDOOR COMMUNICATIONS
[19] B. C. Wadell, Transmission Line Design Handbook. Norwood, MA:
Artech House, 1991, p. 87.
[20] P. B. Katehi, N. G. Alexopoulos, and I. Y. Hsia, “A bandwidth
enhancement method for microstrip antennas,” IEEE Trans. Antennas
Propagat., vol. AP-35, pp. 51–62, Jan. 1987.
[21] P. Bhartia, K. V. S. Rao, and R. S. Tomar, Millimeter-Wave Microstrip
and Printed Circuit Antennas. Norwood, MA: Artech House, 1991.
[22] J.-F. Zurcher and F. E. Gardiol, Broadband Patch Antennas. Norwood,
MA: Artech House, 1995.
Sener Uysal (S’88–M’89) received the B.Eng. degree (with high honors) in electrical engineering
from Eastern Mediterranean University, Famagusta,
Cyprus, in 1984, and the M.Sc. degree in digital
electronics and Ph.D. degree in microwaves from
King’s College London, University of London, London, U.K., in 1986 and 1990, respectively.
In 1992, he joined the Electrical Engineering
Department, National University of Singapore, Singapore, as a Lecturer and, in 1995, was promoted to
a Senior Lecturer. His previous appointment was at
King’s College London, as a Post-Doctoral Research Fellow, where he worked
on microwave and millimeter-wave techniques involving passive circuits.
He has published over 40 technical papers in international symposia and
journals, and has authored Nonuniform Line Microstrip Directional Couplers
and Filters(Norwood, MA: Artech House, 1993). He has invented bandpasstype microstrip directional couplers and filters. His main areas of research
are in coupled lines, filters, and their applications in microwave integrated
circuits (MIC’s).
745
Mook-Seng Leong (M’75–SM’98) received the
B.Sc. degree in electrical engineering (with first
class honors) and the Ph.D. degree in microwave
engineering from the University of London, London,
U.K., in 1968 and 1971, respectively.
He is currently a Professor of electrical
engineering at the National University of Singapore,
Singapore. His main research interests include
antenna and waveguide boundary-value problems.
He is an Editorial Board member for Microwave and
Optical Technology Letters and Wireless Personal
Communications.
Dr. Leong is a member of the Massachusetts Institute of Technology
based Electromagnetics Academy and is a Fellow of the Institution of
Electrical Engineers (IEE), U.K. He received the 1996 Defence Science
Organization (DSO) Research and Development Award presented by DSO
National Laboratories, Singapore.
Chee Hong Ng, photograph and biography not available at the time of
publication.