Received 8 January 2022; revised 8 March 2022; accepted 10 March 2022.

Date of publication 22 March 2022; date of current version 30 March 2022.

Digital Object Identifier 10.1109/OJAP.2022.3161129

High-Gain Dual-Band Dual-Sense Circularly

Polarized Spiral Series-Fed Patch Antenna
NGHIA NGUYEN-TRONG 1 (Member, IEEE), SHENGJIAN JAMMY CHEN 1,2 (Member, IEEE),
AND CHRISTOPHE FUMEAUX 1 (Fellow, IEEE)
1 School of Electrical and Electronic Engnieering, University of Adelaide, Adelaide, SA 5005, Australia

2 College of Science and Engineering, Flinders University, Adelaide, SA 5042, Australia

CORRESPONDING AUTHOR: N. NGUYEN-TRONG (e-mail: [email protected])

This article has supplementary downloadable material available at https://doi.org/10.1109/OJAP.2022.3161129, provided by the authors.

ABSTRACT This paper demonstrates a dual-band circularly polarized (CP) two-dimensional (2D) series-
fed patch array antenna. The CP rotating sense in each band is opposite to each other, which makes them
well-suited for the downlink and uplink in satellite applications. The antenna can be designed to achieve
a very high gain with a simple feeding structure using spiral microsip line, which can be integrated on
a single-layered substrate. A theoretical framework for this type of antennas has been developed, which
facilitates an efficient and fast design procedure requiring minimal computational resource. A prototype
has been fabricated and measured, showing operating bandwidths (overlapped 10-dB return loss, 3-dB
gain, 3-dB axial ratio bandwidths) of about 7% and 5%, realized broadside gains of 21.5 and 22 dBic
at 12 and 14 GHz bands, respectively. This is an outstanding performance considering that the antenna
profile is only 0.03 free-space wavelength (at 12 GHz) and the ratio between two operating frequencies
is very tight accounting to only 1.17.

INDEX TERMS Antenna array, Archimedean spiral, circular polarization, high gain, leaky-wave antenna,
microstrip patch antenna, series-fed patch antenna, satellite communications.

I. INTRODUCTION substrate-integrated waveguide (SIW) [11] or ridge-gap

ATELLITE communications require significantly high waveguide feed [12] have been proposed as alternatives,
S antenna gain with circular polarization (CP). Typically,
the antenna needs to operate in dual band, one for uplink
however, the complexity in feeding still required multi-
layered configurations. Single-layer structures were achieved
and one for downlink. The CP rotating sense for each band in [13], [14], but still with a large number of power dividers
is opposite to each other to minimize cross-talk. Due to and matching circuits [13] or with a very narrow band-
all of these requirements, designing antennas for satellite width [14]. Another method is to utilize metasurface-inspired
communications is a challenging task, especially when other Fabry-PÃl’rot structure [15] in which each unit cell is specif-
constraints such as low profile and low cost are imposed. ically designed to achieve ±90◦ phase shift in each band;
Furthermore, since the CP rotating sense is opposite, it can be nevertheless, the structure has high profile and the axial ratio
even a more difficult task if the ratio between two operation (AR) bandwidth is small.
frequencies is small. To achieve high-gain antenna with a simple feeding and
Most of previous planar designs proposed in the litera- a low-profile structure, the concept of leaky-wave antenna
ture have utilized conventional or modified microstrip patch (LWA) and series-fed patch array antenna is often utilized.
arrays. These structures have the disadvantage of using Conventional LWAs and series-fed patch array antennas are
complicated feeding networks. This issue becomes more one-dimensional (1D) structure fed by a straight transmis-
problematic when dual-band operation with opposite rotat- sion line. Two-dimensional (2D) realization of such antennas
ing sense is required, which often results in multi-layered have recently started to gain more attention from antenna
structures [1]–[7]. Antennas utilizing metasurfaces [8]–[10], researchers and designers [16], [17] with the potential
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/

VOLUME 3, 2022 343

Nguyen-Trong et al.: HIGH-GAIN DUAL-BAND DUAL-SENSE CP SPIRAL SERIES-FED PATCH ANTENNA

of increasing the gain while retaining the advantage of

a simple feeding. Several attempts have been made to
design dual-band dual-sense CP antenna, however, these
designs still required complicated multi-layered feeding
structures [18], [19].
This paper demonstrates a passive 2D spiral series-fed
patch array antenna that can achieve all aforementioned
requirements on a single-layered substrate, using a simple
feeding method based on proximity coupling. In particu-
lar, the proposed design works as a dual-band dual-sense
CP antenna at 12 GHz and 14 GHz, i.e., the frequency
bands allocated for satellite communications in Ku-band in
Australia. It is also noted that the method presented here
can be straightforwardly adapted to other frequencies. The
FIGURE 1. (a) A non-optimized series-fed patch array with proximity coupling.
paper is structured as follows. Section II will provide a (b) Simulated instantaneous E-field distribution at 12 GHz. (c) Simulated S-parameters,
brief review on linear series-fed patch array antenna for the (d) Simulated AR at 12 GHz.

sake of completeness and to demonstrate the general feed-

ing principle. The main contribution of the paper is shown
in Section III and Section IV where single-arm and multi- Figure 1 shows the S-parameters and AR results of a
arm spiral series-fed patch array antennas are analyzed and non-optimized design. The antenna is designed with Rogers
demonstrated. Measured antenna performance is then shown TMM-4 substrate (r = 4.5, tan δ = 0.002 and thickness
in Section V, followed by a conclusion. hs = 0.762 mm) with patch elements resonating at f1 ≈
12 GHz (Fig. 1(c)). As shown in Fig. 1(d), the AR is about
6.5 dB at broadside. To obtain a better AR, the authors in [20]
II. THEORY AND REVIEW OF LINEAR SERIES-FED added a rectangular slot in the patch center to control the
PATCH ARRAY two orthogonal modes in the circular patches. Later, in [21],
We first review the recent 1D series-fed patch array design the slot was rotated to provide more degrees of freedom in
concepts introduced in [20], [21], which will be exploited optimizing the AR. Nevertheless, it has been found that the
to develop the proposed 2D spiral series-fed patch array. slot is not effective when the coupling gap g is large. It
This section mainly serves as review of the general physical is noted that a relatively large value of g is needed when a
principle noting that the main contributions of this paper is small coupling is required to achieve higher gain. Therefore,
the AR optimization with an off-center via. This feature has in this paper, we modify the design by replacing the slot by
not been shown before in [20], [21]. a shorting via in the middle of the patch. Theoretically, a
via precisely at the center does not affect the fundamental
A. LINEAR SERIES-FED PATCH ANTENNA radiating modes of a circular patch, i.e., two degenerate
Series-fed patch array antennas are conventional struc- orthogonal TM11 modes. By slightly varying the position of
tures, which have been demonstrated starting four decades this via to be off-center (in both x- and y-direction), the two
ago [22]–[24]. The series-fed patches are progressively radi- orthogonal modes are disturbed and their amplitude ratio and
ating when the electromagnetic power is transmitted along phase can be optimized to improve the AR.
the host transmission line (typically a microstrip line). In Figure 2 shows the results for an optimized design. It can
this paper, we will focus on the CP realization mechanism be noticed that the reflection coefficient has improved as the
where circular patches are fed by proximity coupling [20], AR is optimized towards 0 dB by controlling the asymmetry
in a generic configuration illustrated in Fig. 1(a). This of the structure. This agrees with the analysis in [25] as in
method has two main advantages, namely that the propa- this case both shunt and series power radiate with the same
gation constant β remains nearly the same as that of the magnitude and as a consequence, the antenna efficiency can
feeding microstrip line, and that the leakage rate αrad can approach 100% when a sufficient number of series elements
be conveniently controlled by varying the gap g between is used.
the host microstrip line and the patches. This antenna
can be classified as a longitudinally asymmetrical LWA B. DUAL-BAND CONCEPT
as analyzed in [25]. For this type of antenna, both shunt To open a second band at a frequency f2 > f1 , another
and series power are radiating with 90◦ phase difference, array of smaller circular patches resonating at f2 is added
resulting in CP radiation. For broadside radiation, the dis- on the other side of the transmitting microstrip line [21].
tance between two elements is selected as λg where λg is Due to the asymmetry, i.e., depending whether the patches
the guided-wavelength of the feeding microstrip line. This are on the left or right of the microstrip line, the rotating
will ensure that all patches radiate with the same phase sense of these patches are opposite. Since each patch array is
(Fig. 1(b)). only excited in a relatively small bandwidth at the resonance

344 VOLUME 3, 2022

FIGURE 2. (a) A CP series-fed patch array with proximity coupling optimized with
off-center vias. (b) Simulated instantaneous E-field distribution at 12 GHz.
(c) Simulated S-parameters, (d) Simulated AR at 12 GHz.
FIGURE 4. Design of spiral microstrip line with constant width w (number of turns
N = 1.5 in this diagram).

dividers. As an alternative with simple feeding, we propose a

spiral series-fed patch array. The concept is to use a uniform
microstrip line in a spiral configuration to feed the whole
array. In this section, the feeding is demonstrated first, fol-
lowed by the explanation on how the patches are arranged
to obtain broadside radiation. Based on this, a single-arm
spiral series-fed patch antenna is designed, followed by a
discussion on the radiation efficiency.

A. SPIRAL FEEDING
A conventional Archimedean spiral is utilized for the
host transmission line. The parametric equation P(ϕ) =
(x(ϕ), y(ϕ)) for a spiral starting at the origin in cartersian
coordinates is derived as

a x(ϕ) = 2π a
ϕ cos ϕ
FIGURE 3. (a) A CP dual-band dual-sense series-fed patch array with proximity r(ϕ) = ϕ⇒ (1)
coupling optimized with off-center vias. (b) Simulated instantaneous E-field 2π y(ϕ) = 2π sin ϕ.
a
ϕ
distribution at 12 GHz adn 14 GHz. (c) Simulated S-parameters, (d) Simulated AR at 12
and 14 GHz. In this equation, ϕ is the azimuth angle that runs from 0 to
N × 2π where N is the number of turns (N does not need
to be an integer); a is the constant distance between two
frequency f1 or f2 , they do not significantly affect each other. successive turns (as shown in Fig. 4).
Thus, this design concept is still working when f1 and f2 are The angle α(ϕ) of the outward normal vector with respect
quite close to each other. It is noted that when f1 = f2 , the to the x-axis (see Fig. 4) can be derived as
RHCP and LHCP modes are added up, resulting in a linearly    
dx − cos ϕ + ϕ sin ϕ
polarized antenna. Fig. 3 shows the optimized design and its α(ϕ) = arctan − = arctan . (2)
performance for f1 = 12 GHz and f2 = 14 GHz. The AR is dy sin ϕ + ϕ cos ϕ
less than 3 dB across the 3 dB-beamwdith at both operating It is noted that values of α(ϕ) cover the whole 2π range
frequencies. while the function arctan only yields values in range of
[−π/2, −π/2], so care needs to be taken when perform-
III. SINGLE-ARM SPIRAL SERIES-FED PATCH ANTENNA ing (3) to ensure α(ϕ) is a continuous function. A more
In order to increase the gain, more radiating elements need to convenient equation to calculate α(ϕ) is
be added alongside with a reduction in the coupling between ⎧ 
the patches and the host transmission line. In a linear arrange- ⎨ arctan − cos ϕ+ϕ sin ϕ if dy/dϕ ≥ 0
 sin ϕ+ϕ cos ϕ
α(ϕ) = (3)
ment, this results in a very long 1D structure with a fan-beam ⎩ arctan − cos ϕ+ϕ sin ϕ + π if dy/dϕ < 0.
sin ϕ+ϕ cos ϕ
pattern where frequency-dependent beam squinting can be
a serious issue. An array of parallel 1D structures can be The value of α(ϕ) is critical for the patch placement which
designed but it still requires a feeding network with power is shown in the next section.

VOLUME 3, 2022 345

Nguyen-Trong et al.: HIGH-GAIN DUAL-BAND DUAL-SENSE CP SPIRAL SERIES-FED PATCH ANTENNA

of the microstrip line at position ϕ1 = 0 and ϕ2 such that

L(ϕ2 ) = λg . We choose the radiating phase reference at
the first patch ψr1 = ψr (ϕ = 0) = 0◦ . Since L(ϕ2 ) = λg ,
the phase of the microstrip line at ϕ1 and ϕ2 is the same.
However, due to the curvature of the microstrip line, the
field is rotated by an angle made of the two normal vectors
n1 and n2 . From the definition of α(ϕ), this angle can be
calculated as α(ϕ2 ) − α(ϕ1 ). Therefore, the excitation phase
of the second patch is ψr2 = α(ϕ2 ) − α(ϕ1 ). By this argu-
ment, the excitation phase of a patch placed at an arbitrary
position P(ϕ) at a distance L(ϕ) from the spiral feed can be
found as
FIGURE 5. Placement of two patches along a spiral section. ψr (ϕ) = α(ϕ) − α0 − ϕDL
L(ϕ)
Let the width of the feeding microstrip line be w, the = α(ϕ) − α0 − 2π , (10)
λg
equations for the two edges of the microstrip line can be
derived as where α0 is the angle of the normal vector corresponding to
the first patch which is used as a phase reference, and ϕDL
x(ϕ) = aϕ cos ϕ ± x(ϕ) (4) is calculated as in (10).
y(ϕ) = aϕ sin ϕ ± y(ϕ), (5) Physically, (10) implies that to make sure that all of the
patches have the same excitation phase, we need to select
where
the location of the patches such that L(ϕ) compensates for
w
x(ϕ) = cos α(ϕ) (6) the field rotation due to the curvature of the microstrip line.
2 For the case that the patches are placed on the other side
w
y(ϕ) = sin α(ϕ). (7) of the microstrip line (for the 2nd band), the normal vector
2
towards the patch is now in the opposite direction. Hence,
Equations (5)-(8) are obtained by “shifting” the original spi-
the excitation phase is
ral curve along the normal direction at each location by
+w/2 and −w/2. This ensures that the microstrip line main- ψr (ϕ) = −α(ϕ) − α0 − ϕDL
tains its constant width w at any point along the spiral curve L(ϕ)
= −α(ϕ) − α0 − 2π . (11)
(Fig. 4). λg
The length of the spiral from its center is
 Using (10) and (11), it is straightforward to select two sets

ϕ  2  2 of angles {ϕi }, {ϕi } such that the excitation phase ψr (ϕi ) (and
dx dy
L(ϕ) = + dϕ ψr (ϕi )) are the same for all indices i. Furthermore, since α0
dϕ dϕ
0
  is a constant, it can be ignored when selecting {ϕi } and{ϕi }.
1 1
= ϕ 1 + ϕ + ln ϕ + 1 + ϕ .
2 2 (8) These two sets correspond to a spiral array of patches on
2 2 either side of the feeding microstrip line. The analysis here
It is noted that this length L(ϕ) is important to calculate the also provides a tool to design the antenna with an arbitrary
phase delay from the center of the spiral to a position P(ϕ) beam direction rather than broadside, but this is beyond the
along the spiral. This phase delay is scope of this paper.
L(ϕ)
ϕDL = L(ϕ)β = 2π , (9) C. PRELIMINARY DESIGN
λg
Using the framework developed in this section, a dual-band
where β is the propagation constant of the host microstrip dual-sense CP single-arm spiral series-fed patch array has
line. been designed for operation at 12 GHz and 14 GHz as shown
in Fig. 6. The parameters for the spiral are a = 15.3 mm
B. PATCH PLACEMENT and N = 2.125. The value of a should be chosen large
The patch placement is critical to ensure that all the patches enough to have sufficient space for the patches of both bands,
are radiating with uniform phase to obtain broadside radi- and small enough to avoid grating lobe in the array. Here
ation. For a linear 1D series-fed patch array, the distance a is about 0.7λ0 at f2 = 14 GHz (where λ0 is the free-
between two patches can be simply selected as λg . For a space wavelength). Optimized parameters (unit: mm) for this
spiral feeding, the radiating phase of the patch is not only antenna are the microstrip line width w = 0.8, radii of the
affected by the phase delay provided by the microstrip line, patches R1 = 3.27, R2 = 2.75, location of off-centered vias
but also by curvature of the feeding line defined by α(ϕ). xoff1 = 0.27, yoff1 = 0.56, xoff2 = 0.29, yoff2 = 0.53, and gap
To understand this, one can consider the simple case sizes g1 = 0.52, g2 = 0.59. These parameters are defined in
shown in Fig. 5. Two patches are placed on the outside Fig. 3(a).

346 VOLUME 3, 2022

 14.3 dBi at 14 GHz. It is noted that the realized gain should
also take the absorption at Port 2 into account (as shown in
the next subsection). The slightly higher gain at 12 GHz is
due to the fact that the effective aperture for the 12 GHz
array is larger than that for 14 GHz since the patches for
14 GHz are placed inside of the spiral (see Fig. 6).

D. ANTENNA EFFICIENCY DISCUSSION

In terms of efficiency, a series-fed patch array antenna can
be considered as a LWA, where the radiation efficiency is
calculated as
αrad αrad αc + αd
ηe = = =1− . (12)
FIGURE 6. Design of a dual-band dual-sense CP antenna with a single-spiral αc + αd + αrad αtotal αtotal
microstrip line. Ground plane size RGND = 42.5 mm.
In (12), αrad , αc and αd are the leakage rate, the attenuation
due to the conductor loss and dielectric loss, respectively.
The total antenna efficiency is

ηe-total = ηe 1 − |S11 |2 − |S21 |2 . (13)

Let us assume that the LWA is designed such that Pterm of

power is absorbed at the termination (Pterm = |S21 |2 ≈ 1%
for the single-arm spiral series-fed patch array in Fig. 6).
For a spiral with a total length of Ltotal , the total attenuation
constant is calculated as
ln(Pterm )
αtotal = (14)
−2Ltotal
Based on this, the leakage rate should be chosen as
αrad = αtotal − αc − αd . (15)
In order to increase the gain, the antenna size, and there-
fore the spiral length Ltotal need to increase. Thus αtotal need
to be decreased according to (14). Meanwhile, the loss rate
of the conductor and substrate (αc and αd ) stays constant.
Therefore, αrad decreases as in (15), which then reduces
FIGURE 7. Simulated results of the single-arm spiral series-fed patch array in Fig. 6:
(a) S-parameters; (b) Axial ratio at broadside versus frequency; (c) Realized gain at
the antenna efficiency according to (12). As an example, let
12 GHz; (d) Realized gain at 14 GHz. us attempt to re-design the antenna in Fig. 6 for a higher
gain by enlarging the spiral feed with N = 4 while keeping
Pterm = 0.01. From the simulation results of the single-arm
The antenna is excited perpendicularly from the bottom spiral antenna in Fig. 6, i.e., ηe = 0.821 with N = 2.125 and
by a 50- coaxial connector at the onset of the spiral (Port 1 a = 15.3 mm, the sum of conductor and substrate loss is
in Fig. 6). Given a substrate with a fixed thickness (TMM-4, calculated as αc + αd = 1.86 Np/m. When N = 4, the total
r = 4.5 and thickness hs = 0.762 mm), the 50- microstrip attenuation rate is calculated from (14) as αtotal = 2.97 Np/m.
line width can be calculated as w50- = 1.5 mm. However, Thus, the efficiency is only 37% (about −4.3 dB).
it is empirically found that a smaller value of w makes the Figure 8 shows the calculated theoretical radiation effi-
control of the coupling with the patch more convenient. Thus, ciency for different numbers of turns N for the antenna in
we choose here w = 0.8 mm (which yields a slightly higher Fig. 6. All other parameters remain the same and antenna
characteristic impendance of 70 ). In this case, a single is designed for fixed Pterm = 0.01. The result demonstrates
quarter-wavelength impedance matching section added at the that at some point, we cannot increase the gain by making
beginning of the feed is sufficient to achieve good matching the antenna larger because the radiation efficiency degra-
to a 50- connector. dation will offset the gain increase through increase of the
The antenna performance is summarized in Fig. 7. The aperture. Our various simulations indicate that for the con-
simulated radiation efficiency is 82.1% and 81.5% at 12 GHz sidered materials (copper and TMM4 substrate with loss
and 14 GHz, respectively (both conductor and dielectric tangent tan δ = 0.002, thickness 0.762 mm), practically the
losses are taken into account in simulation). The simulated gain cannot exceed 18 dBic at 12 GHz for a single-spiral
realized gain at broadside reaches 15.3 dBi at 12 GHz and feed configuration.

VOLUME 3, 2022 347

Nguyen-Trong et al.: HIGH-GAIN DUAL-BAND DUAL-SENSE CP SPIRAL SERIES-FED PATCH ANTENNA

FIGURE 8. Calculated radiation efficiency of the antenna in Fig. 6 designed with FIGURE 9. Final design of the dual-band dual-sense CP antenna with 4 spiral
different sizes. All other parameters remain the same. Pterm is fixed at 0.01. microstrip line arms.

where ψr (ϕ) is calculated as in (10) or (11). The above

IV. MULTI-ARM SPIRAL SERIES-FED PATCH ANTENNA
equation is important to ensure that all patches in all arms
Instead of one spiral arm, Na number of arms can be used.
radiate with the same phase. The ± sign is selected depend-
For the same aperture size, the length of each arm Ltotal is
ing on the side of the patches relatively to the feeding lines.
reduced roughly by Na times. Equations (14) and (15) then
For the case Na = 4, αarm = 0◦ , 90◦ , 180◦ and 270◦ .
infer that the total attenuation constant and leakage rate can
To avoid reflection at the terminations which can degrade
be chosen larger. This will result in better radiation efficiency
the AR performance, the antenna is terminated with 4
according to (12). Furthermore, using additional spiral arms
matched ports as shown in Fig. 9. In this case, the relative
will result in a more even power distribution in the antenna
total power absorbed at all termination is
aperture which will further improves the aperture efficiency
and beam radial symmetry. Pterm = |S21 |2 + |S31 |2 + |S41 |2 + |S51 |2 ≈ 4|S21 |2 . (17)
In general, if the number of spiral arms is larger, a higher
antenna efficiency can be achieved. The number of spiral One issue with a uniform LWA or series-fed patch array
arms is limited by practical considerations on the feeding is that the elements closer to the feed radiate with a higher
network at the spiral center. If Na arms are used and no intensity than the elements near the termination. This sig-
feeding network is implemented, the impedance of each nificantly reduces the aperture efficiency, especially with
microstrip line needs to be 50  × Na . High value of long structures. This problem is mitigated with the proposed
microstrip line characteristic impedance can result in higher Na -arm spiral structure since the radiating power can radi-
conductor loss and the required thin strips may become dif- ally spread further, and is therefore distributed more evenly
ficult to realize. Thus, an impedance matching section is in the aperture. To further improve the aperture efficiency,
generally needed, which nevertheless is still much simpler the gap profile {g1i }, {g2i } can be selected such that the gaps
than a typical feeding network in a 2D microstrip patch are larger for small i and smaller for large i (i is the ele-
array. ment index from the feed to termination). In this design, we
For demonstration, a design with Na = 4 spiral arms is choose a simple linear gap profile as
shown in Fig. 9. All parameters for the patches are the i−1
same as for the design of single-arm spiral series-fed patch g1i = g1start + (g1end − g1start ) (18)
N1 − 1
array. The parameters for the spiral are a = 61.2 mm
i−1
and N = 1. It is noted that a now is 4 times the g2i = g2start + (g2end − g2start ) (19)
value that would be used in a single-arm spiral antenna, N2 − 1
which allows to accommodate the 4 intertwined arms. By where i = 1 to N1 in (18) and i = 1 to N2 in (19); N1 and
choosing the value of a in this way, the spacing for the N2 are the number of patch elements for the f1 and f2 -band,
patch placement remains the same. A quarter-wavelength respectively. The values for the final design in Fig. 9 are
impedance transformer (which is designed at 13 GHz) is g1start = 0.56 mm, g1end = 0.36 mm, g2start = 0.69 mm,
used at the center to excite 4 arms of the spiral microstrip g2end = 0.39 mm. These values are optimized such that
line. Pterm ≈ 0.01 (or equivalently |S21 | ≈ −26 dB) while maxi-
If an arm is rotated by an angle αarm , the radiating phase mizing the realized gain at broadside for an optimal aperture
of an element corresponding to that arm is adjusted as efficiency.
For demonstration, Fig. 10 shows the simulated electric
ψr-rotate (ϕ) = ψr (ϕ) ± αarm , (16) field distribution at 12 GHz and 14 GHz of the optimized

348 VOLUME 3, 2022

FIGURE 10. Simulated instantaneous electric field at 12 GHz and 14 GHz of the final
4-arm spiral series-fed patch array design.

FIGURE 12. Simulated and measured S-parameters of the 4-arm spiral series-fed
patch array.

FIGURE 11. Photographs of the fabricated 4-arm spiral series-fed patch array
prototype.

FIGURE 13. Simulated and measured AR at broadside across the frequency.

design. It can be observed that all patches are radiating

with almost the same phase, due to the procedure to select
the patch positions demonstrated in Section III-B. At each The |S21 | results show a slight shift to the left (of about
operating frequency, only the corresponding patches are 0.3 − 0.4%) for the resonance frequency of the patches.
excited. The power is distributed more evenly compared to This is mainly due to the tolerance of the substrate material
the previous single-arm design. This results in a more sym- (TMM-4) which may have a slight variation in its rela-
metrical radiation pattern and a higher aperture efficiency. tive permittivity. The discrepancy in measured and simulated
It is also noted that there is a small amount of coupling |S11 | can be attributed to the fabrication tolerance at the feed
among the patches, however, this coupling does not signif- where the quarter-wavelength microstrip line transformer’s
icantly affect the antenna performance as will be shown in width is only wt = 0.22 mm (see in the inset in Fig. 9).
the next section with measurement results. Figure 13 shows the simulated and measured AR at
broadside across the frequency range. The measured 3-dB
AR-bandwidths extend from 11.25 GHz to 12.4 GHz and
V. MEASUREMENT from 13.25 GHz to 14.4 GHz. Compared to simulation, the
The 4-arm spiral series-fed patch array optimized in measured AR bandwidth is also slightly shifted to the left,
Section VI has been fabricated and measured in an in- which confirms the behavior observed with the measured
house anechoic chamber. A photograph of the antenna in its |S21 | parameters. Wide 3-dB AR bandwidths are achieved
measurement setup is shown in Fig. 11. The simulated and as about 8% at both bands in both simulation and measure-
measured S-parameters are shown in Fig. 12. Since |S21 |, ment. At broadside, the measured cross-polarization is lower
|S31 |, |S41 | and |S51 | are similar to each other, only |S11 | than −18 dB and −25 dB at 12 and 14 GHz, respectively.
and |S21 | are shown. Both simulated and measured reflection The antenna radiation patterns are shown in Figs. 14
coefficients show good performance in a very wide band- and 15 for 12 GHz and 14 GHz, respectively. The results
width. This is one of the advantage of this type of antennas demonstrate that the antenna exhibits a directive beam at
as impedance matching is relatively simple and robust to broadside, which confirms that the radiating phase of all
achieve compared to conventional patch arrays. patches are close to each other. Dual-sense CP has also

VOLUME 3, 2022 349

Nguyen-Trong et al.: HIGH-GAIN DUAL-BAND DUAL-SENSE CP SPIRAL SERIES-FED PATCH ANTENNA

FIGURE 14. Normalized radiation pattern in the xz-plane (top) and yz-plane FIGURE 15. Normalized radiation pattern in the xz-plane (top) and yz-plane
(bottom) at 14 GHz. Legends are the same in both figures.
(bottom) at 12 GHz. Legends are the same in both figures.

been verified by measurement. The 3-dB beamwidth in both

xz- and yz-planes is about 13◦ . The ARs as a function of
zenith angle θ at 12 GHz and 14 GHz are shown in Fig. 16
with good performance. The measured AR is below 3 dB
within the 3-dB beam width. It is noted that due to the
slight frequency shift, the performance is indeed better at
about 11.8 GHz compared to the target operating frequency
12 GHz (see Fig. 13).
Finally, Fig. 17 shows the simulated and measured real-
ized broadside gains, which agree quite well to each other.
The antenna achieves a measured maximum realized gain
of about 21.6 dBic and 22.9 dBic at 12 GHz and 14 GHz,
respectively (see Fig. 17). The corresponding simulated val-
ues are 21.5 dBic and 22.0 dBic. It is noted the AR (Fig. 13)
and gain measurement data (Fig. 17) are smoothen using FIGURE 16. Simulated and measured AR versus θ at 12 GHz and 14 GHz.
moving average due to standing wave effect between trans-
mitter and receiver. Slightly higher values in peak measured
gain can be attributed to this standing wave effect and Finally, Table 1 shows a comparison with other high-gain
other errors in gain calibration. Nevertheless, discrepancies dual-CP antennas in the literature (λ0 is the free-space wave-
of about 1 dB are expected for typical gain measurement length at the lower band). Only designs with realized gain
tolerances. The 3-dB gain bandwidth is approximately 7% higher than 13.0 dBi are selected for comparison. It can be
and 5% at the lower and upper bands, respectively. The sim- seen that the proposed design has a unique structure with
ulated total efficiency is 81% for both bands. The simulated unique concept, which makes it low-cost and easy for inte-
total efficiency is taking into account both dielectric and con- gration. Only the proposed design and [13] can achieve the
ductor loss, as well as the reflection and termination power. functionality with a single-layer substrate. However, [13]
Reasonable agreement in the simulated and measured gain stills used complicated feeding network with relatively
indicates a similar value for the practical efficiency of the large element spacing which resulted in higher sidelobe
antenna. level.

350 VOLUME 3, 2022

TABLE 1. Performance comparison with literature on High-Gain Dual-CP antennas.

[3] C.-X. Mao, S. Gao, Y. Wang, Q. Luo, and Q.-X. Chu, “A

shared-aperture dual-band dual-polarized filtering-antenna-array with
improved frequency response,” IEEE Trans. Antennas Propag., vol. 65,
no. 4, pp. 1836–1844, Apr. 2017.
[4] C.-X. Mao, S. Gao, Y. Wang, Q.-X. Chu, and X.-X. Yang, “Dual-band
circularly polarized shared-aperture array for c -/ x -band satellite
communications,” IEEE Trans. Antennas Propag., vol. 65, no. 10,
pp. 5171–5178, Oct. 2017.
[5] S. Mohammadi-Asl, J. Nourinia, C. Ghobadi, M. Majidzadeh,
E. Mostafapour, and A. Haghshenas, “High gain modified dual-band
and dual-circularly polarized array antenna,” Wireless Pers. Commun.,
vol. 97, no. 2, pp. 2131–2144, 2017.
[6] M. Wang, D.-Y. Wang, W. Wu, and D.-G. Fang, “Single-layer, dual-
port, dual-band, and orthogonal-circularly polarized microstrip antenna
array with low frequency ratio,” Wireless Commun. Mobile Comput.,
vol. 2018, pp. 1–10, Apr. 2018.
[7] C.-X. Mao, Z. H. Jiang, D. H. Werner, S. S. Gao, and W. Hong,
“Compact self-diplexing dual-band dual-sense circularly polarized
FIGURE 17. Simulated and measured broadside realized gain.
array antenna with closely spaced operating frequencies,” IEEE Trans.
Antennas Propag., vol. 67, no. 7, pp. 4617–4625, Jul. 2019.
[8] K. Li, L. Li, Y.-M. Cai, C. Zhu, and C.-H. Liang, “A novel design of
low-profile dual-band circularly polarized antenna with meta-surface,”
VI. CONCLUSION
IEEE Antennas Wireless Propag. Lett., vol. 14, pp. 1650–1653, 2015.
In this paper, a theoretical framework for a spiral 2D [9] J. Zhu, Y. Yang, S. Li, S. Liao, and Q. Xue, “Dual-band dual circu-
dual-band dual-sense series-fed patch array with proximity larly polarized antenna array using FSS-integrated polarization rotation
coupling has been developed. Critical aspects of this type AMC ground for vehicle satellite communications,” IEEE Trans. Veh.
Technol., vol. 68, no. 11, pp. 10742–10751, Nov. 2019.
of antennas, including feeding, element placement, radia- [10] T. Yue, Z. H. Jiang, and D. H. Werner, “A compact metasurface-
tion efficiency and aperture efficiency have been thoroughly enabled dual-band dual-circularly polarized antenna loaded with
discussed. Single-arm and multi-arm spiral series-fed patch complementary split ring resonators,” IEEE Trans. Antennas Propag.,
vol. 67, no. 2, pp. 794–803, Feb. 2019.
arrays were demonstrated with the multi-arm design show- [11] Z.-J. Guo, Z.-C. Hao, H.-Y. Yin, D.-M. Sun, and G. Q. Luo, “Planar
ing a better performance. A 4-arm spiral series-fed patch shared-aperture array antenna with a high isolation for millimeter-
array prototype has been validated experimentally, show- wave low earth orbit satellite communication system,” IEEE Trans.
Antennas Propag., vol. 69, no. 11, pp. 7582–7592, Nov. 2021.
ing an operating bandwidth (overlapped −10-dB impedance [12] Y. Liu, Z. Yue, Y. Jia, Y. Xu, and Q. Xue, “Dual-band dual-circularly
bandwidth, 3-dB broadside gain bandwidth and 3-dB AR polarized antenna array with printed ridge gap waveguide,” IEEE
bandwidth) of 7% at the 12 GHz and 5% at the 14 GHz Trans. Antennas Propag., vol. 69, no. 8, pp. 5118–5123, Aug. 2021.
[13] J.-D. Zhang, W. Wu, and D.-G. Fang, “Dual-band and dual-circularly
band. The antenna shows outstanding performance for a polarized shared-aperture array antennas with single-layer substrate,”
single-layered substrate antenna without any external feeding IEEE Trans. Antennas Propag., vol. 64, no. 1, pp. 109–116, Jan. 2016.
network and with a very low profile of only 0.03λ0 . [14] J.-D. Zhang, L. Zhu, N.-W. Liu, and W. Wu, “Dual-band and
dual-circularly polarized single-layer microstrip array based on mul-
tiresonant modes,” IEEE Trans. Antennas Propag., vol. 65, no. 3,
pp. 1428–1433, Mar. 2017.
REFERENCES [15] C. Chen, Z.-G. Liu, H. Wang, and Y. Guo, “Metamaterial-inspired
[1] S. Ye, J. Geng, X. Liang, Y. J. Guo, and R. Jin, “A compact dual-band self-polarizing dual-band dual-orthogonal circularly polarized Fabry–
orthogonal circularly polarized antenna array with disparate ele- Pérot resonator antennas,” IEEE Trans. Antennas Propag., vol. 67,
ments,” IEEE Trans. Antennas Propag., vol. 63, no. 4, pp. 1359–1364, no. 2, pp. 1329–1334, Feb. 2019.
Apr. 2015. [16] M. ElSherbiny, A. E. Fathy, A. Rosen, G. Ayers, and S. M. Perlow,
[2] A. Garcia-Aguilar, J.-M. Inclan-Alonso, L. Vigil-Herrero, “Holographic antenna concept, analysis, and parameters,” IEEE Trans.
J.-M. Fernandez-Gonzalez, and M. Sierra-Perez, “Low-profile Antennas Propag., vol. 52, no. 3, pp. 830–839, Mar. 2004.
dual circularly polarized antenna array for satellite communications [17] T. Zhao, D. R. Jackson, J. T. Williams, and A. A. Oliner, “General for-
in the x band,” IEEE Trans. Antennas Propag., vol. 60, no. 5, mulas for 2-D leaky-wave antennas,” IEEE Trans. Antennas Propag.,
pp. 2276–2284, May 2012. vol. 53, no. 11, pp. 3525–3533, Nov. 2005.

VOLUME 3, 2022 351

Nguyen-Trong et al.: HIGH-GAIN DUAL-BAND DUAL-SENSE CP SPIRAL SERIES-FED PATCH ANTENNA

[18] R. Shavit, L. Pazin, Y. Israeli, M. Sigalov, and Y. Leviatan, “Dual SHENGJIAN JAMMY CHEN (Member, IEEE)
frequency and dual circular polarization microstrip nonresonant array received the M.E. and Ph.D. degrees in electrical
pin-fed from a radial line,” IEEE Trans. Antennas Propag., vol. 53, and electronic engineering from The University
no. 12, pp. 3897–3905, Dec. 2005. of Adelaide, Australia, in 2013 and 2017, respec-
[19] K. Wincza and S. Gruszczynski, “Series-fed dual-band dual-polarized tively.
antenna lattice fed by slot-coupled power dividers,” IEEE Antennas From 2017 to 2021, he was a Postdoctoral
Wireless Propag. Lett., vol. 15, pp. 1065–1068, 2016. Researcher and a Lecturer with the School
[20] S. J. Chen, C. Fumeaux, Y. Monnai, and W. Withayachumnankul, of Electrical and Electronic Engineering, The
“Dual circularly polarized series-fed microstrip patch array with University of Adelaide. He joined the College of
coplanar proximity coupling,” IEEE Antennas Wireless Propag. Lett., Science and Engineering, Flinders University as
vol. 16, pp. 1500–1503, 2017. a Lecturer in 2022. His current research interests
[21] C. D. Bui, N. Nguyen-Trong, and T. K. Nguyen, “A planar include antenna design and engineering, microwave absorbers, and wearable
dual-band and dual-sense circularly polarized microstrip patch leaky- and reconfigurable electromagnetic structures based on novel materials.
wave antenna,” IEEE Antennas Wireless Propag. Lett., vol. 19, Dr. Chen received scholarships, including the Australian Postgraduate
pp. 2162–2166, 2020. Award 2013 and the Simon Rockliff Scholarship 2015. He was also
[22] T. Metzler, “Microstrip series arrays,” IEEE Trans. Antennas Propag., the recipient of a number of awards, including the Young Scientist
vol. 29, no. 1, pp. 174–178, Jan. 1981. Best Paper Award at the International Conference on Electromagnetics in
[23] D. M. Pozar and D. H. Schaubert, “Comparison of three series fed Advanced Applications (ICEAA) 2015, the Young Scientist Best Paper
microstrip array geometries,” in Proc. IEEE Antennas Propag. Soc. Award and Travel Bursary Award at ICEAA 2016, an Honorable Mention
Int. Symp., vol. 2, 1993, pp. 728–731. at IEEE AP-S Symposium on Antennas and Propagation (APS/URSI)
[24] P. Hallbjorner, I. Skarin, K. From, and A. Rydberg, “Circularly 2017, a CST University Publication Award 2017, and the Best Paper
polarized traveling-wave array antenna with novel microstrip patch Award at IEEE Asia–Pacific Microwave Conference 2021. He serves as a
element,” IEEE Antennas Wireless Propag. Lett., vol. 6, pp. 572–574, Reviewer of some prestigious journals, including the IEEE TRANSACTIONS
2007. ON A NTENNAS AND P ROPAGATION , the IEEE T RANSACTIONS ON
[25] S. Otto, Z. Chen, A. Al-Bassam, A. Rennings, K. Solbach, and MICROWAVE THEORY AND TECHNIQUES, the IEEE ANTENNAS AND
C. Caloz, “Circular polarization of periodic leaky-wave antennas WIRELESS PROPAGATION LETTERS, IET Microwaves, Antennas &
with axial asymmetry: Theoretical proof and experimental demonstra- Propagation, IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND
tion,” IEEE Trans. Antennas Propag., vol. 62, no. 4, pp. 1817–1829, SYSTEMS, IEEE ACCESS, and Electronic Letters. He was the Chair of the
Apr. 2014. IEEE South Australia Joint Chapter on Microwave Theory and Techniques
& Antennas and Propagation from 2019 to 2020.

CHRISTOPHE FUMEAUX (Fellow, IEEE) received

the Diploma and Ph.D. degrees in physics from
the ETH Zurich, Switzerland, in 1992 and 1997,
respectively.
From 1998 to 2000, he was a Postdoctoral
Researcher with the School of Optics, University
of Central Florida, Orlando. In 2000, he joined
the Swiss Federal Office of Metrology, Bern,
Switzerland, as a Scientific Staff Member. From
2001 to 2008, he was a Research Associate and a
Lecturer with the Laboratory for Electromagnetic
Fields and Microwave Electronics, ETH Zurich. Since 2008, he has been
NGHIA NGUYEN-TRONG (Member, IEEE) with The University of Adelaide, Australia, where he is currently a Professor
received the Ph.D. degree (Doctoral Research with the School of Electrical and Electronic Engineering. His main research
Medal) in electrical engineering from The interests concern electromagnetics, antenna engineering, and the application
University of Adelaide, Adelaide, SA, Australia, of RF design principles across the electromagnetic spectrum. He was the
in 2017. recipient of the ETH Medal for his doctoral dissertation, the 2018 Edward
He is currently a Lecturer with The University E. Altshuler Prize, the 2014 IEEE SENSORS JOURNAL, and the 2004 ACES
of Adelaide. His main research interests include Journal best paper awards. He also received best conference paper awards
microwave circuits, advanced materials, absorbers, at the 2012 Asia–Pacific International Symposium on Electromagnetic
and various types of antennas. Compatibility and the 17th Colloque International sur la Compatibilité
Dr. Nguyen-Trong was one of the recipients of Electromagnétique 2014. More than ten of his students have received stu-
the Best Student Paper Award at the 2014 IWAT, dent awards with joint papers at IEEE conferences. He was the recipient of
the 2015 IEEE MTT-S NEMO, and the 2017 ASA Conferences, and the the University of Adelaide Stephen Cole the Elder Award for Excellence
Best Paper Award at the 2018 and 2020 AMS Conference. He has been in Higher Degree by Research Supervisory Practice in 2018. From 2011 to
continuously selected as a Top Reviewer for IEEE TRANSACTIONS ON 2015, he was a Future Fellow of the Australian Research Council. He served
ANTENNAS AND PROPAGATION, in 2018, 2019, 2020, and 2021, and IEEE as an Associate Editor for the IEEE TRANSACTIONS ON MICROWAVE
ANTENNA WIRELESS AND PROPAGATION LETTERS, in 2018 and 2021. THEORY AND TECHNIQUES from 2010 to 2013. From 2013 to 2016, he
He serves as a Technical Co-Chair for the 2020 Australian Microwave served as a Senior Associate Editor and later as the Associate Editor-in-
Symposium (AMS) and 2022 IEEE International Symposium on Antennas Chief for the IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION.
and Propagation. He is listed among Australia’s Top 40 Early Career Since March 2017, he has been serving as the Editor-in-Chief for the IEEE
Researchers by The Australian, November 2021. ANTENNAS AND WIRELESS PROPAGATION LETTERS.

352 VOLUME 3, 2022