IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 56, NO.

9, SEPTEMBER 2008

2875

Design of a Multiresonator Loaded Band-Rejected

Ultrawideband Planar Monopole Antenna With
Controllable Notched Bandwidth
Tzyh-Ghuang Ma, Member, IEEE, Ren-Ching Hua, and Chin-Feng Chou

AbstractWe present a new multiresonator loaded band-rejected planar monopole antenna for ultrawideband applications.
The proposed microstrip-fed antenna is composed of a flared
metal plate, a truncated ground plane, and two pairs of folded
strips. By applying the resonance nature of the folded strips
and the associated cross coupling effects, the proposed antenna
demonstrates bandstop-filter-like response with bandwidth controllability at the targeted rejection band. To illustrate the antenna
operating mechanism more clearly, an equivalent circuit model
consisting of RLC lumped resonators and J-inverters is discussed
and extracted. The antenna input admittance calculated with the
help of the equivalent circuit model agrees reasonably well with
the simulated one obtained by the full-wave simulator. The design
concept and the simulated and experimental results including
the return losses, radiation patterns, and gain responses versus
frequencies are carefully investigated throughout the paper. A
parametric study in terms of the bandwidth tunability of the
proposed design at the notched band is performed. The effects of
the finite size ground plane are studied in this paper as well.
Index TermsEquivalent circuits, monopole antennas, resonators, ultrawideband (UWB) antennas.

I. INTRODUCTION

LTRAWIDEBAND (UWB) radio, one of the core technologies in wireless personal area networks (WPANs),
has experienced a blooming growth in recent years [1][3].
For the moment, UWB chipsets with transmission rates up
to 480 Mbps have been developed and demonstrated. Due
to the inherently ultra-wide operating bandwidth from 3.1
to 10.6 GHz, the circuit components in an UWB radio face
quite different challenges. For example, an ultrawideband
antenna involves considerable extra design constraints. In such
a system the antenna behaves more like a bandpass filter in both
spatial and frequency domains. Any nonideal variation of the
antenna response will inevitably introduce signal distortion and
hence seriously deteriorate the overall performance. Various

Manuscript received September 10, 2007; revised January 2, 2008. Published

September 4, 2008 (projected). This work was supported by the National Science Council, R.O.C., under Grants 96-2221-E-011-007 and NCS 97-2221-E011-019-MY2.
T.-G. Ma is with the Department of Electrical Engineering, National Taiwan
University of Science and Technology, Taipei 10607, Taiwan, R.O.C. (e-mail:
tgma@ee.ntust.edu.tw).
R.-C. Hua was with the Department of Electrical Engineering, National
Taiwan University of Science and Technology, Taipei 10607, Taiwan R.O.C.
He is now with the R.O.C. Army, Taipei 112, Taiwan, R.O.C.
C.-F. Chou was with the Department of Electrical Engineering, National
Taiwan University of Science and Technology, Taipei 10607, Taiwan R.O.C.
He is now with Tecom Co. Ltd., Taiwan 300, R.O.C.
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TAP.2008.928778

researchers have been devoted themselves to investigating the

descriptions, analyses, and optimizations of ultrawideband antennas in either time domain or frequency domain [4][9], and
novel antenna designs have been successfully demonstrated in
the literatures as well [10][21]. Among those newly proposed
ultrawideband antennas, the planar monopole antennas should
be the most fascinating candidate for future applications due to
their remarkably compact size and stable radiation characteristics [12][14].
It is noted that, however, due to the coexistence of UWB and
WLAN radios at the frequency band of 56 GHz, ultrawideband
antennas with filtering properties are urgently required not only
to mitigate the devastating interference but also to remove the requirement of an additional bandstop filter. Such antenna designs
are generally referred to as the band-rejected or band-notched
ultrawideband antennas [15][18], and can be readily accomplished by introducing an additional resonant structure to the
antenna body so as to shift the input impedance level to either a
very high or very low state. As the resonance occurs, the antenna
behaves more similar to a virtual-open or short circuit which
is capable of preventing the energy from radiating to the free
space. The common ways to introduce the additional resonant
structure including etching slots or slits on the antenna metal
plate [15], [16], placing parasitic strips in close proximity to
the radiating element [16], [17], applying folded strips as coupled line sections [18], and etc., The amount of gain suppression
in such designs varies from several decibels to 1020 decibels,
which is principally depended on the quality factor of the resonator applied. Keeping the band-rejected filtering properties
invariant over all spatial directions is also a crucial but rather
difficult issue in the design.
Despite the successful developments of band-notched ultrawideband antennas, it is observed that the notched bandwidth
with acceptable signal-rejection level is sometimes quite limited owing to the inherent narrowband nature of the resonator
introduced. In addition, the bandwidth controllability of most
recently proposed band-notched ultrawideband antennas is relatively limited as well. It is likely owing to the fact that in these
designs only a single resonator or a single pair of resonators is
loaded to the antenna structures. Accordingly, the notched bandwidth is simply inversely proportional to the quality factor of
the resonator installed. It therefore has limited tunability if one
also wants to keep the center frequency of the rejection band
unchanged [18].
To deal with this problem, recently we present a new design
concept which explores the possibility of introducing multiple
pairs of resonators to an ultrawideband antenna [19]. The newly

0018-926X/$25.00 2008 IEEE

2876

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 56, NO. 9, SEPTEMBER 2008

TABLE I
ELEMENT VALUES OF THE LUMPED EQUIVALENT CIRCUIT MODEL IN [20]

Fig. 1. A modification of the antenna in [20] with W = 1:08; L =

16:85; L = 11; W = 3; L = 7:5; W = 3, and W = 15. All units are in
millimeters. (a) Antenna configuration. (b) Equivalent circuit model.

proposed antenna in [19] demonstrates bandstop-filter-like response and good frequency selectivity at the rejection band. In
this paper, we will investigate the performance of this antenna
in greater detail. The antenna geometry and design concept will
be first discussed in Section II. An equivalent circuit model is
proposed to explain the operating mechanism of the proposed
design. In Section III, the antenna performance, bandwidth controllability, and ground plane effects will be investigated thoroughly. The element values of the equivalent circuit model will
be extracted as well. This paper is concluded with a brief summary in Section IV.
II. ANTENNA CONFIGURATION AND DESIGN CONCEPT
The design initially begins with the antenna shown in
Fig. 1(a), which is a minor modification of the antenna discussed in [20]. The behavior of the antenna in Fig. 1(a) is quite
similar to that of the antenna proposed in [18], and can be
represented by the same equivalent circuit model, as shown
in Fig. 1(b). In the circuit model the pair of folded strips in
the fork-shaped radiator can be represented by two lumped
parallel lossy RLC resonators. The flared metal plates between
the folded strips and the microstrip feed line behave similar to
two quarter-wavelength transformers, and can be modeled by a
pair of J-inverters. The wideband nature of the antenna, on the
other hand, can be readily accounted for by a constant radiation
with
mho. By implementing
conductance
the procedure given in [18], the element values of the equivalent
circuit model in Fig. 1(b) can be extracted and are tabulated

in Table I for easy reference. The extraction procedure will be

repeated in detail in Section III-D.
To explore the concept of multiresonator loaded bandnotched ultrawideband antenna, in this paper we discuss a
new antenna design by introducing an extra pair of folded
strips to the antenna body in Fig. 1(a). The proposed antenna
configuration is illustrated in Fig. 2. As shown in the figure,
the new design lies in the xy-plane with its normal direction
being parallel to the z axis. The antenna is symmetric about its
centerline, and consists of a microstrip feed line, a truncated
ground plane, a flared metal plate, and two pairs of folded strips.
It can be shown that the folded strip behaves like a parallel
RLC tank [18], and can be physically realized by connecting
the top and bottom metal foils by rows of via holes. The radius
of the via hole is 0.2 mm. The coupling between the metal
foils provides the necessary capacitive loading whereas the
strip itself contributes to the inductance value of the resonator.
For the sake of simplicity, in the following discussions the
two pairs of resonators are referred to as the exterior and the
interior resonators, and the subscripts e and i are used to
distinguish the parameters associated with the exterior and
interior resonators, respectively.
To explain the complex behavior of the proposed antenna,
here an equivalent circuit model is illustrated in Fig. 3. In the
following discussion in this paragraph, the subscripts are neglected for the sake of brevity. Referring to the circuit model,
each folded strip can be represented by a lumped parallel lossy
RLC resonator. The resonant frequency of the resonator can be
readily controlled by adjusting the equivalent inductance and
and
, of the resonator, and are
capacitance values, i.e.,
in turn determined by the geometric parameters of the folded
strip. As shown in Fig. 2, the area of the metal plate on the
, preliminary determines the value
bottom layer,
of the capacitive loading. The insets on the top layer, on the
other hand, principally act as an inductive loading of the radiating strip. It has the ability to fine-tune the inductance value
of the resonator. It should be aware that as the areas of the inby
) become larger, the amount of capacitive
sets (
loading decreases accordingly. This implies a nonlinear dependency between the areas of the inductive insets and the resonant
frequency of the lumped resonator. The equivalent resistance
, meanwhile, accounts for the conduction loss, the dielectric
loss, as well as the parasitic radiation loss of the folded strip, and
can be used to determine the quality factor of the resonator.
and
, are used
In the meantime, a pair of J-inverters,
in the equivalent circuit model to characterize the transformer-

MA et al.: DESIGN OF A MULTIRESONATOR LOADED BAND-REJECTED UWB PLANAR MONOPOLE ANTENNA

2877

Fig. 4. Photograph of the proposed antenna. (Units: cm).

Fig. 2. Proposed antenna configuration. (a) Oblique view. (b) Top view.

Fig. 3. Equivalent circuit model of the proposed antenna.

like behaviors of the flared metal plates. As shown in Fig. 2,

the tapered profiles of the flared metal plates between the interior and exterior folded strips are described by equations of arcs.
Similar to the antenna in [18], the flared metal plates behave like
two quarter-wavelength impedance transformers. The parasitic
coupling between the interior resonator and its nearby exterior
resonator can be accounted for by these J-inverters as well. On
the other hand, the cross coupling between the pair of interior
resonators, which are in close proximity to each other, can be
. The coupling coeffimodeled by an additional J-inverter
cient can be calculated using the theory of coupled resonator
circuit in ([22], Chap 8), and will be introduced in Section III.D.
is again used to account
Finally, the radiation conductance
for the wideband nature of the proposed design, and the microstrip feed line can be represented by a transmission line with
and propagation constant . For
a characteristic impedance

further illustration, the element values of the equivalent circuit

model in Fig. 3 will be extracted and discussed in greater detail
in Section III-D.
Following the discussions, the antenna was first optimized by
the full wave simulation solver HFSS version 9.2, and then fabricated on a 20-mil Rogers 4003 substrate with a dielectric constant of 3.38 and loss tangent of 0.0027. A photograph of the
fabricated antenna is shown in Fig. 4. The overall dimension of
the antenna is 34 by 34 mm . The parameters associated with
mm,
mm,
two pairs of folded strips are
mm, and
mm, and the parameters related to the capacitive and inductive loadings of the interior and
mm,
exterior resonators are given by
mm,
mm,
mm, and
mm,
mm,
mm,
mm, respectively. The distance between the interior and exterior resonators
is 8.8 mm, and the distance between the two interior resis 1.4 mm. The feeding microstrip line, which is
onators
designed to have a line impedance of 50 ohms, has a length
of 16.5 mm and a width
of 1.08 mm. The feeding gap
between the truncated ground plane and the flared metal plate is
equal to 0.2 mm. To maximize the antenna impedance bandwidth, the tapered profile of the flared metal plate is described
by an arc with a radius of 33.8 mm and a sector angle of 26.2 degrees. The truncated ground plane occupies an area of 16.5 by 34
mm . The truncated ground plane, whose size is comparable to
the radiator, has significant effects on the antenna performance
and also contributes to the radiation. More studies with respect
to the finite size ground plane will be discussed in Section III-C.
III. EXPERIMENTAL RESULTS AND DISCUSSIONS
A. Return Losses, Patterns and Gain Responses
First of all, the antenna performances in terms of the return
loss, radiation patterns, and gain responses versus frequencies
are investigated in this subsection. The simulated and measured
antenna return losses are illustrated in Fig. 5. Owing to the
strong reactive near-field coupling of the proposed multiresonator loaded antenna, it is observed through measurement that

2878

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 56, NO. 9, SEPTEMBER 2008

Fig. 6. Measured radiation patterns in the yz-plane at (a) 4 GHz and (b) 8 GHz.

Fig. 5. Simulated and measured return losses of the proposed antenna.

the antenna return loss is very sensitive to the SMA connector

attached as well as to the truncated ground plane used. To tackle
this problem and to improve the measuring accuracy, in this
work the antenna return losses were measured by a Cascade
Microtech MTF26 microstrip test fixture along with an Agilent PNA-L 5230 A at National Taiwan University. Several absorbers were placed around the test fixture to minimize the undesired coupling. Referring to Fig. 5, reasonable agreement between the simulated and measured return losses can be readily
observed in the frequency band of concern, and the discrepancy
is likely a result of the parasitic coupling between the antenna
and the nearby metallic objects. The slight misalignment between the top and bottom capacitive plates as well as the tolerance of the substrate used in the design also contribute to the
measurement uncertainty since the value of the capacitance may
be slightly changed due to these nonideal effects. In addition, it
is shown in the figure that the rejection band demonstrates two
resonance peaks, which are at 5.35 and 5.75 GHz, respectively.
The maximum signal rejection levels read from the measured results are 3.61 and 3.97 dB. The simulated signal rejection levels,
on the other hand, are slightly higher than the measured ones. It
can be readily accounted for by the parasitic coupling that does
not take into account in the simulation. And the most important
of all, it is noted that at the notched frequencies the roll-off rate
of the return loss of the proposed design is significantly sharper
than that of a conventional single resonator loaded ultrawideband antenna. As a result, the measured data preliminarily sustains the design concept proposed in this work.
The antenna radiation patterns were measured in a
7 3.2 3 m anechoic chamber at National Taiwan University of Science and Technology. The measurement was
performed by an Agilent E8362B network analyzer along with
NSI 2000 far-field measurement software. In the measurement
the connecting cable attached to the antenna, as illustrated
in the insets of Figs. 6 and 7, was carefully surrounded by
electromagnetic wave absorbers to reduce the multireflection
interference. In addition, to minimize the parasitic coupling the
antenna was fed directly from the connecting cable without a
SMA adapter. The simulated and measured radiation patterns

Fig. 7. Measured radiation patterns in the xz-plane at (a) 4 GHz and (b) 8 GHz.

in the yz- (H-) and xz- (E-) planes at the center frequencies, 4
and 8 GHz, of the UWB low and high bands are illustrated in
Figs. 6 and 7, respectively. Here the UWB low band is referred
to as the mandatory band from 3.1 to 5.1 GHz whereas the
UWB high band stands for the optional band from 5.85 to
10.6 GHz. The agreement between the simulated and measured
radiation patterns is very excellent in the yz-plane and remains
acceptable in the xz-plane. The discrepancy in the xz-plane can
be mostly attributed to the interference of the bended coaxial
cable used for feeding purpose in the measuring arrangement,
as illustrated in the inset of Fig. 7. Referring to the figures,
the radiation pattern in the yz-plane is nearly omnidirectional
at the UWB low band. As the operating frequency increases,
it becomes similar to the shape of a four-leaved clover. On
the contrary, the radiation patterns in the xz-plane remain
roughly a dumbbell-like shape over the frequency band of
interest. Referring to Figs. 6 and 7, in the UWB mandatory
band the cross-polarized electric field, or equivalently the
component, is generally much lower than the co-polarized one
component). As the operating frequency increases, the
(
cross polarization level rises considerably. Although not shown
here for simplicity, it should be aware that at the upper edge
of the UWB high band the cross-polarized component may be
comparable to the co-polarized one in the yz-plane.
The antenna was further examined by means of the gain responses versus frequencies from 3 to 11 GHz [21]. Fig. 8(a)
illustrates the measured gain responses in the yz-plane at
, and
. The gain responses in the same plane
, and
are unveiled in Fig. 8(b).
as
The required masked spectrum at the 5-GHz WLAN band is

MA et al.: DESIGN OF A MULTIRESONATOR LOADED BAND-REJECTED UWB PLANAR MONOPOLE ANTENNA

2879

also depicted in the figures for comparison purpose. The experimental results manifest that the proposed design demonstrates
bandstop-filter-like response at the targeted rejection band. As
shown in the figures, the gain responses fluctuate back and forth
between 0 to 2 dBi at the UWB low band, and climb up to 5 dBi
. In
at the UWB high band in the azimuth plane, i.e.,
addition, a rejection band with good frequency selectivity can
be readily observed at the 5-GHz WLAN band with two notch
peaks. At the center frequency of the rejection band, the gain
.
suppression can be more than 20 dB, for example, at
Despite the fact that the band-notched characteristics remain acceptable in most of the reception directions, it is noted that the
rejection band may be split into two individual ones with less
.
frequency selectivity at certain reception angles, e.g.,
It is likely a result of the slight shifts in the center frequencies
of the notch peaks with respect to the rotating angles. Further
investigation with the help of the multilayered printed circuit
broad fabrication process is currently underway to mitigate the
spatial dependencies of the gain responses as well as to improve
the signal rejection level at the notched frequencies. The result
will be discussed in a future publication.

B. Bandwidth Controllability
As illustrated in Fig. 2, by slightly adjusting the lengths of
the metal plates printed on the bottom layer of the substrate,
the amount of capacitive loading of the exterior and interior
resonators can be made somewhat unequal. Together with the
cross coupling between the resonators, the two pairs of asynchronous tuned coupled resonators provide the additional flexibility in controlling the bandwidth of the rejection band. In
this subsection, the bandwidth controllability of the proposed
band-notched ultrawideband antenna is studied in terms of the
geometric parameters of the resonators loaded to the antenna
body. It is observed through both simulation and measurement
that the bandwidth as well as the center frequency of the rejection band can be effectively controlled by adjusting the geometric parameters of the exterior resonators. The interior resonators, on the contrary, have less influence on the bandwidth
controllability. Fig. 9 summarizes the measured antenna return
losses versus frequencies for five geometric parameters related
to the exterior resonators. Here the return losses are only shown
from 3 to 8 GHz for the sake of easy comparison. As shown
in the figures, the parameters are the length and width of the
and
, the
metal plates for capacitive loading, i.e.,
length and depth of the insets for inductive loading, i.e.,
and
, and the distance between the interior and exterior
. As illustrated in Fig. 9(a) and (c), the lengths
resonators,
of the capacitive plates as well as the inductive insets preliminarily determine the center frequency of the notched band. The
notched bandwidth, on the other hand, remains roughly unalor
varies. On the contrary, the widths of the
tered as
metal plates or the depths of the insets can be used to adjust the
amount of the capacitive or inductive loading of the folded strip,
and hence have significant controllability on the bandwidth of
the rejection band. Referring to Fig. 9(b) and (d), as the amount
of inductive or capacitive loading increases (or equivalently as

Fig. 8. Measured gain responses of the proposed antenna at (a) 

0 ; 90 ; 180 ; 90 and (b)  = 45 ; 135 ; 135 ; 45 .

or
increases), the bandwidth of rejection band increases proportionally. In addition, the capacitive loading has
a more significant controllability on the bandwidth than the inductive loading does. It is likely a result of the nonlinear dependency between the inductive loading and the resonant frequency of the resonator, as mentioned in the previous section.
Finally, it is noted from Fig. 9(e) that the two notch peaks in the
rejection band can be effectively relocated by simply adjusting
the distance between the interior and exterior resonators, i.e.,
. It is because that
has the ability to fine-tune the
transformer-like behavior of the flared metal plate as well as
the amount of parasitic near-field coupling between the two resonators. Accordingly, it determines the values of the J-inverters,
and
, in Fig. 3. Moreover, it is also worthwhile to mention that by tuning the distance between the exterior and interior resonators, the proposed antenna can demonstrate two individual rejection bands with less frequency selectivity. The result, however, is not of the most concern in the current design,
and will not be shown here for simplicity.

2880

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 56, NO. 9, SEPTEMBER 2008

Fig. 10. The simulated and measure return losses of the antenna design for
= 44 mm.
compensating the effects of the ground plane with W

nonideal effects related to the variations of the ground plane on

the rejection characteristics can be simply compensated by applying the tuning parameters discussed in the previous subsection.
D. Extraction of the Equivalent Circuit Model
Fig. 9. Bandwidth controllability of the proposed antenna versus (a) the length
and (b) width of the capacitive plate, (c) the length and (d) depth of the inductive
inset, and (e) the distance between the exterior and interior resonators.

C. The Effects of Ground Plane

It has been mentioned in Section II that the proposed antenna
is very sensitive to the size of the truncated ground plane used
in the design. Accordingly, it is necessary to discuss the effects of the ground plane in greater detail. With the help of the
full-wave simulator HFSS, we observe that the behavior of the
rejection band can be dramatically affected by the variations of
the size of the truncated ground plane. For demonstration purpose, here an example is illustrated in this subsection with the
being widened by 10 mm.
width of the ground plane
It is observed through simulation that as the size of the ground
plane becomes larger the notch peaks may be further apart and
tend to split into two individual rejection bands. To tackle this
problem, the antenna parameters for bandwidth controllability
in Section III.B can be applied to fine-tune the antenna response.
By relocating the exterior pair of folded strips with a distance
mm and suitably adjusting the tapered profile
of
of the flared metal plate, well-behaved band-notched response
with good frequency selectivity at the 5 GHz-WLAN band can
be readily restored. The simulated and measure return losses of
the re-tuned antenna design are depicted in Fig. 10. Fairly good
agreement between the results can be again observed. Referring
to the measured data, the impedance matching at the UWB high
band can be substantially improved at the expanse of a slight
degradation of the impedance matching at the UWB mandatory
band. With this illustration, we can therefore conclude that the

In this subsection, the element values of the antenna equivalent circuit model in Fig. 3 are extracted, and the calculated
antenna input admittance using the equivalent circuit model is
compared to that achieved by the HFSS simulator. The extraction procedure initially begins with the extraction of the element
values of each of the individual exterior and interior resonators
[18]. Referring to Fig. 11(a), the one-port network is consisted
of a single resonator and a finite-size ground plane. A delta
source is used as the excitation. With the help of a full-wave simulator, the one-port network parameter can be easily achieved
and transformed into the lumped equivalent parallel resonant
circuit with
(1)
(2)
(3)
Here, FBW represents the fractional bandwidth with the magnitude of the input impedance of the one-port network dropping to 0.707 of its peak value. Q is the quality factor, and
is the center radian frequency. Noted that
corresponds to
the real part of the input impedance of the one-port network at
the center frequency . Fig. 12 compares the simulated input
impedance of one of the exterior resonators using HFSS to the
calculated one with the extracted lumped RLC elements. Fairly
good agreement between the results can be readily observed.
It is well known that the input admittance of a lumped parallel
resonator is given by
(4)

MA et al.: DESIGN OF A MULTIRESONATOR LOADED BAND-REJECTED UWB PLANAR MONOPOLE ANTENNA

2881

Fig. 11. (a) Circuit layout for extracting the lumped equivalent circuit model
of a single resonator. (b) Circuit layout for extracting the coupling coefficient
of a pair of resonators.

Accordingly, the admittance at the input of the J-inverter

can be expressed as

(5)
,
To calculate the admittance at the antenna feed point, i.e.,
the equivalent circuit in Fig. 3 is first divided into two half circuits with respect to the centerline of the antenna, as illustrated
can be represented by the
in Fig. 13(a). Here the -inverter
cascade of two identical two-port networks whose ABCD matrices are described by [22, Ch. 10]

Fig. 12. Comparisons of the input impedance of the exterior resonator simulated using HFSS to that calculated by the equivalent parallel RLC resonator.
(a) Resistance and (b) Reactance.

In (7), the admittance looking into the input port of the two-port
network in (6) with the other port being open-circuited is equal
[22, Ch. 10]. It can be readily verified with the help
to
of (6).
The admittance at the antenna feed point is therefore

(6)
It is routine to verify that the consecutive multiplication of (6)
results in the ABCD matrix of an ideal J-inverter. From Fig. 2,
it is evident that the feed scheme of the proposed design can
only support even mode operation. It in turn implies that the two
half circuits of the equivalent circuit in Fig. 13(a) are always at
equal potential. Accordingly, the antenna input admittance can
be calculated by replacing the plane of symmetry in Fig. 13(a)
with a magnetic wall, and find the even mode admittance of the
half circuit in Fig. 13(b) as

(7)

(8)
and the input antenna admittance including the effects of the
microstrip line can be expressed as

(9)
, the coupling coTo quantitatively extract the J-inverter
of the interior resonator pair can be first calcuefficient
lated with the help of the formulations for a standard coupled
resonator circuit [22, Ch. 8]. As shown in Fig. 11(b), the pair of

2882

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 56, NO. 9, SEPTEMBER 2008

Fig. 13. Derivation of the antenna input admittance. (a) Equivalent circuit
model with the plane of symmetry. (b) The half circuit.

coupled resonators is excited by lumped ports, and the -parameters of this two-port network can be obtained using HFSS simulator. The natural resonant frequencies, i.e., and , of the
network can be readily observed by recording the two resonant
response of this two-port netpeaks in the transmission
work. The coupling coefficient between the coupled resonators
is therefore
(10)
Fig. 14. Comparisons of the input admittance of the proposed antenna simulated using HFSS to that calculated by the equivalent circuit model in Fig. 3. (a)
Conductance. (b) Susceptance.

and the J-inverter for representing the cross coupling is

(11)
(12)
Here, is the slope parameter of the resonator . The final parameters,
and
, do not have simple extraction method
since this pair of inverters accounts for both the transformer-like
behavior of the flared metal plates and the parasitic near-field
coupling between the interior and exterior resonators. They are
determined by equaling the antenna admittance level simulated
using HFSS to the calculated one using the equivalent circuit
model with (9). As a result, the pair of inverters is served as
a final tuning parameter to adjust the admittance level of the
equivalent circuit model. In the current design, the inverters
and
are given by 0.0077 mhos (i.e., 130 ohms).
Fig. 14(a) and (b) compares the simulated input conductance
and susceptance, respectively, of the proposed antenna using
HFSS simulator to the calculated ones using (9). Reasonable
agreement between the results can be again observed, and
hence manifests the correctness of the proposed equivalent
circuit model. The discrepancy can be most attributed to the
inaccuracy modeling of the resonator itself and the ignored
cross coupling between the resonators 1 and 3, and etc., For
easy reference, the extracted element values of the equivalent
circuit model in Fig. 3 are summarized in Table II.

IV. CONCLUSION
In this paper, a novel band-notched ultrawideband planar
monopole antenna with multiple resonators loaded on the antenna structure has been investigated and discussed in detail. By
applying two pairs of folded strips, the antenna demonstrates
bandstop-filter-like response at the desired rejection band
with bandwidth controllability. An equivalent circuit model
involving lumped RLC parallel resonators and J-inverters has
been extracted to explain the rather complicated behavior of
the proposed design. The calculated antenna input admittance
using the equivalent circuit model agrees reasonably well with
the simulated one using the full-wave simulator. In addition,
five tuning parameters have been investigated in terms of their
tunability of the bandwidth as well as the center frequency
of the notched band, and the effects of the truncated ground
plane have been discussed as well. The antenna performance is
evaluated in terms of the return losses, radiation patterns and
gain responses, and the applicability of the proposed design
to the ultrawideband radios has been well proved. The future
works will be in further improving the accuracy of the antenna
equivalent circuit model, and in improving the rejection performance of the proposed design at the 5-GHz notched band.

MA et al.: DESIGN OF A MULTIRESONATOR LOADED BAND-REJECTED UWB PLANAR MONOPOLE ANTENNA

2883

TABLE II
ELEMENT VALUES OF THE EQUIVALENT CIRCUIT MODEL IN FIG. 3

ACKNOWLEDGMENT
The authors would like to express their sincere gratitude to an
anonymous professor who gave the initial idea of this work at
the 2006 IEEE International Symposium on Antennas and Propagation, and to Prof. S.-Y. Chen at National Taiwan University
for his help in setting up the measuring arrangement.
REFERENCES
[1] [Online]. Available: http://www.ieee802.org/15/pub/TG3a.html Task
group 3a homepage
[2] [Online]. Available: http://www.uwbforum.org/ UWB, Forum homepage
[3] [Online]. Available: http://www.wimedia.com/ WiMedia, homepage
[4] Z. N. Chen, X. H. Wu, N. Yang, and M. Y. W. Chia, Considerations
for source pulses and antennas in UWB radio systems, IEEE Trans.
Antennas Propag., vol. 52, pp. 17391748, Jul. 2004.
[5] C. E. Baum, E. G. Farr, and C. A. Frost, Transient gain of antennas
related to the traditional continuous-wave (CW) definition of gain, in
Proc. Ultra-Wideband Short-Pulse Electromagnetics 4, Jun. 1998, pp.
109118.
[6] M. Ciattaglia and G. Marrocco, Approximate calculation of time-domain effective height for aperture antennas, IEEE Trans. Antennas
Propag., vol. 53, pp. 10541061, Mar. 2005.
[7] S. Licul and W. A. Davis, Unified frequency and time-domain antenna
modeling and characterization, IEEE Trans. Antennas Propag., vol.
53, pp. 28822888, Sep. 2005.
[8] J. McLean, H. Foltz, and R. Sutton, Pattern descriptors for UWB antennas, IEEE Trans. Antennas Propag., vol. 53, pp. 553559, Jan.
2005.
[9] J. McLean, H. Foltz, and R. Sutton, Conditions for direction-independent distortion in UWB antennas, IEEE Trans. Antennas Propag., vol.
54, pp. 31783183, Nov. 2006.
[10] Y.-J. Ren and K. Chang, An annual ring antenna for UWB communications, IEEE Antennas Wireless Propag. Lett., vol. 5, pp. 274276,
2006.
[11] M. Klemm, I. Z. Kovcs, G. F. Pedersen, and G. Trster, Novel smallsize directional antenna for UWB WBAN/WPAN applications, IEEE
Trans. Antennas Propag., vol. 53, pp. 38843896, Dec. 2005.
[12] J. Liang, C. C. Chiau, X. Chen, and C. G. Parini, Study of a printed
circular disc monopole antenna for UWB systems, IEEE Trans. Antennas Propag., vol. 53, pp. 35003504, Nov. 2005.
[13] Z. N. Low, J. H. Cheong, and C. L. Law, Low cost PCB antenna
for UWB application, Antennas Wireless Propag. Lett., vol. 4, pp.
237239, 2005.
[14] Z. N. Chen, T. S. P. See, and X. Qing, Small printed ultrawideband
antenna with reduced ground plane effect, IEEE Trans. Antennas
Propag., vol. 55, pp. 383388, Feb. 2007.
[15] W. S. Lee, W. G. Lim, and J. W. Yu, Multiple band-notched planar
monopole antenna for multiband wireless systems, IEEE Microw.
Wireless Compon. Lett., vol. 15, pp. 576578, Sept. 2005.
[16] Y.-C. Lin and K.-J. Hung, Compact ultrawideband rectangular
aperture antenna and band-notched designs, IEEE Trans. Antennas
Propag., vol. 54, pp. 30753081, Nov. 2006.
[17] K.-H. Kim and S.-O. Park, Analysis of the small band-rejected antenna
with the parasitic strip for UWB, IEEE Trans. Antennas Propag., vol.
54, pp. 16881692, Jun. 2006.

[18] T. G. Ma and S. J. Wu, Ultrawideband band-notched folded strip

monopole antenna, IEEE Trans. Antennas Propag., vol. 55, pp.
24732479, Sept. 2007.
[19] R. C. Hua, C. F. Chou, and T. G. Ma, On the design of multipole
band-notched UWB planar monopole antenna with good frequency
selectivity, in Proc. Progress in Electromagnetic Research Progress
Symp. (PIERS), Beijing, China, Mar. 2630, 2007, p. 1378.
[20] T. G. Ma and S. J. Wu, Ultrawideband band-notched U-shape folded
monopole antenna and its radiation characteristics, in Ultra-Wideband
Short Pulse Electromagnetics. New York: Springer, 2007, pp. 4956.
[21] T. G. Ma and C. H. Tseng, An ultrawideband coplanar waveguide-fed
tapered ring slot antenna, IEEE Trans. Antennas Propag., vol. 54, pp.
11051110, Apr. 2006.
[22] J. S. Hong and M. J. Lancaster, Microstrip Filters for RF/Microwave
Application. New York: Wiley, 2001.
Tzyh-Ghuang Ma (S00M06) was born in Taipei,
Taiwan, R.O.C., in 1973. He received the B.S. and
M.S. degrees in electrical engineering and the Ph.D.
degree in communication engineering from National
Taiwan University, Taipei, Taiwan, R.O.C., in 1995,
1997, and 2005, respectively.
In 2005, he joined the faculty of the Department
of Electrical Engineering, National Taiwan University of Science and Technology (NTUST), Taipei,
Taiwan, R.O.C., where he is currently an Associate
Professor. His research interests include ultrawideband antenna and RF front-end circuit designs, mobile antenna designs, radio
frequency identification (RFID), and microwave passive circuit designs.

Ren-Ching Hua was born in Taipei, Taiwan, R.O.C.,

in 1981. He received the B.S. degree in communications and guidance engineering from National
Taiwan Ocean University, Keelung, Taiwan, R.O.C.,
in 2005 and the M.S. degree in electrical engineering
from National Taiwan University of Science and
Technology (NTUST), Taipei, Taiwan, R.O.C., in
2007.
In 2007, he joined the R.O.C. Army for mandatory military service. His research interests include
mobile antenna designs and RFID reader antenna designs.

Chin-Feng Chou was born in Hsinchu City, Taiwan,

R.O.C., in 1982. He received the B.S. degree in electrical engineering from Southern Taiwan University
of Science and Technology (STUT), Tainan, Taiwan,
R.O.C. in 2004 and the M.S. degree in electrical engineering from National Taiwan University of Science
and Technology (NTUST), Taipei, Taiwan, R.O.C.,
in 2007.
In 2007, he joined Tecom Co. Ltd., Taiwan,
R.O.C., where he works on RF circuits in Bluetooth.
His research interests include mobile antenna designs, RF front-end circuit module for ultrawideband radios, and filter designs.