See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/258649323

Dual-Frequency Decoupling for Two Distinct Antennas

Article  in  IEEE Antennas and Wireless Propagation Letters · December 2012

DOI: 10.1109/LAWP.2012.2226635

CITATIONS READS

12 74

4 authors, including:

Jacob Coetzee Xinyi Tang

Queensland University of Technology National University of Singapore
77 PUBLICATIONS   678 CITATIONS    24 PUBLICATIONS   357 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Waveguide slot arrays View project

Wideband phase shifters View project

All content following this page was uploaded by Jacob Coetzee on 04 April 2018.

The user has requested enhancement of the downloaded file.

This is the author’s version of a work that was submitted/accepted for pub-
lication in the following source:

Albannay, Mohammed Masoud, Coetzee, Jacob, Tang, Xinyi, & Mouthaan,

Koenraad
(2012)
Dual-frequency decoupling for two distinct antennas.
IEEE Antennas and Wireless Propagation Letters, 11, pp. 1315-1318.

This file was downloaded from: https://eprints.qut.edu.au/54261/

c Copyright 2012 IEEE

This work has been submitted to the IEEE for possible publication. Copy-
right may be transferred without notice, after which this version may no
longer be accessible

Notice: Changes introduced as a result of publishing processes such as

copy-editing and formatting may not be reflected in this document. For a
definitive version of this work, please refer to the published source:

https://doi.org/10.1109/LAWP.2012.2226635
 1

Dual-Frequency Decoupling for

Two Distinct Antennas
Mohammed M. Albannay, Student Member, IEEE, Jacob C. Coetzee, Member, IEEE,
Xinyi Tang, Member, IEEE and Koen Mouthaan, Member, IEEE

Abstract—Constant development of new wireless standards
increases the demand for more radiating elements in compact
end-user platforms. A decrease in antenna separation gives rise
to increased antenna coupling, resulting in a reduction of the
signal-to-interference-plus-noise-ratio (SINR) between transmit-
ter and receiver. This paper proposes a decoupling network which
provides dual band port isolation for a pair of distinct antennas.
A prototype has been fabricated to verify the theory.
Index Terms—Dual-band decoupling, decoupling network, port
isolation, mutual coupling.
I. I NTRODUCTION
isolation. In its simplest form, a decoupling network consists
of reactive components connected between neighboring anten-
nas. Typically, a decoupling network for an N port antenna
array is a 2N port network, incorporating reactive components
and/or transmission line sections and/or stubs. Port isolation
between two identical antennas has been achieved for single
band [7]–[9] and dual band cases [10], [11].
In this paper, a decoupling network for two closely spaced,
distinct antennas operating at two different frequencies is intro-
duced. It provides a matched port for each of the frequencies
while maintaining isolation at both frequencies. The theory is
validated by measurements on a fabricated prototype.

D EMAND for high performance, reliability and data rate

communication has catalyzed the evolution of wireless
standards in the past three decades [1] . Consequently, modern II. T HEORY AND D ESIGN OF D ECOUPLING N ETWORK
end-user platforms are equipped with miniaturized antennas to
The distinct antenna pair in Fig. 1 is characterized by
operate with the latest standards whilst allowing for backward
scattering matrix Sa , given by
compatibility. It is common practice in the wireless industry
" a a
#
to allocate a distinct radiator for each operating band cluster S11 S12
a
in the end-user platform. User equipment manufacturers are S = a a (1)
S12 S22
constantly decreasing the physical dimensions of the platform ,
to enhance aesthetics and utility. This inevitably leads to a a
where 6=
S11 S22 .
The corresponding admittance matrix can
aggravated mutual coupling between closely spaced radiators.
be calculated from
Mutual coupling causes feedback looping between radiators " a a
#
[2], [3]. This in turn decreases SINR, which causes significant a a a −1
Y11 Y12
transmission degradation [3]. Y = Y0 (I − S )(I + S ) = a a (2)
Y12 Y22
This intrinsic problem can be overcome with the use of ,

miniaturized surface acoustic wave (SAW) duplexers [4], where I is an 2 × 2 identity matrix and Y0 is the characteristic
utilizing a frequency division duplex (FDD) scheme. This admittance of the system.
approach occupies a very small footprint, but its complex This pair of antennas can be decoupled using the circuit
architecture makes it difficult to design and fabricate for shown in Fig.1. It consists of two transmission lines with
arbitrary frequency bands. A miniaturized microstrip diplexer characteristic impedance Z01 and Z02 and electrical lengths θ1
is discussed in [5]. It provides dual-band isolation, but tailoring and θ2 . A shunt element with an admittance of jB completes
the design for an arbitrary pair of frequencies would also the network. Using transmission line theory, the admittance
require significant effort. Suspended neutralizing lines are matrix at ports 10 and 20 can be calculated from
used in [6] to reduce coupling between two planar inverted-F " 0 0
#
antennas. While the solution is simple in principle, it lacks Y11 Y12
0 −1
a robust design procedure and is difficult to implement in Y = BA = 0 0 (3)
Y12 Y22
practice. ,
An alternative approach to overcoming mutual coupling is with
through the use of a passive decoupling network to obtain port " #
a a
cos θ1 + jZ01 Y11 sin θ1 jZ01 Y12 sin θ1
M. M. Albannay and J. C. Coetzee are with the School of Electrical A= a a
Engineering and Computer Science, Queensland University of Technology,
jZ02 Y12 sin θ2 cos θ2 + jZ02 Y22 sin θ2
Brisbane, QLD 4001, Australia (e-mail: albannay@student.qut.edu.au; and " #
a a
jacob.coetzee@qut.edu.au). Y11 cos θ1 + j sin(θ1 )/Z01 Y12 cos θ1
X. Tang and K.Mouthaan are with the Department of Electrical and B= a a
Computer Engineering, National University of Singapore, Singapore 119077, Y12 cos θ2 Y22 cos θ2 + j sin(θ2 )/Z02
.
Singapore (e-mail: eletangx@nus.edu.sg; k.mouthaan@nus.edu.sg).
 2

1 2

HPF ϕHPF ( fi )
Z02
Z01 1 2 Z02 2
θ1 θ2 ϕ ( fi )= _
= θ2 ( f i )
BPF ϕBPF ( fi )

1’ 2’

1’’ jB 2’’
Fig. 2. Implementation of line 2 with a high-pass filter cascaded to a band-
pass filter to simultaneously realize the required phase shift φ(f1 ) and φ(f2 ).
Fig. 1. Distinct antenna pair with decoupling network
C C C1 L1

With the addition of the parallel admittance jB, the C2 L2 C2 L2

L
admittance matrix at ports 100 and 200 becomes
" #
00 G011 + j(B11
0
+ B) G012 + j(B12 0
− B) (a) (b)
Y = (4)
G012 + j(B12
0
− B) G022 + j(B22 0
+ B)
.
Fig. 3. Topology of (a) high-pass filter and (b) band-pass filter.
For an arbitrary choice of Z01 , Z02 and θ1 , the electrical length
θ2 can be adjusted so that G012 = 0. By selecting B = B12 0
,
equation (4) reduces to and electrical length θ1 (f1 ) for line 1 in Fig. 1. For this trans-
" # mission line, the electrical length θ1 (f2 ) has a predetermined
00 G011 + j(B110 0
+ B12 ) 0 value of
Y = (5) f2
0 G022 + j(B22 0
+ B120
) θ1 (f2 ) = θ1 (f1 ). (8)
.
f1
The zero off-diagonal elements in (5) indicate that the ports
The parameter θ1 (f2 ) is thus no longer be available for the
are decoupled.
minimization of the cost function, leaving only Z01 , Z02 and
Port isolation for the antenna pair at two frequencies
θ1 (f1 ) as variables for the optimization procedure. For line
fi (i = 1, 2) is accomplished by calculating the characteristic
2, a high-pass/band-pass filter cascade can be employed, as
impedances Z01 and Z02 as well as the required electrical
shown in Fig. 2. The topologies of the two filters are shown
lengths θ1 (fi ) and θ2 (fi ). The voltage reflection coefficient
in Fig. 3. The required phase relation of the high-pass and
seen at ports 100 and 200 is given by
band-pass filter is given by
 Y0 − Yii00 
 
|Γi | =  . (6) φHP F (fi ) + φBP F (fi ) = −θ2 (fi ), i = 1, 2. (9)
Y0 + Y 00 ii
The band-pass
√ filter is designed at a center frequency of
An infinite number of solutions are available. The most appro-
f0 = f1 f2 and produces an antisymmetric phase response
priate solution is obtained by minimizing the mismatch at ports
at frequencies f1 and f2 , so that
100 and 200 . A cost function g = max[|Γ1 (f1 )|, |Γ2 (f2 )|] can
be used to optimize parameters Z01 , Z02 , θ1 (f1 ) and θ1 (f2 ), φBP F (f1 ) = −φBP F (f2 ). (10)
while θ2 (f1 ) and θ2 (f2 ) are computed during each iterative
From (9) and (10), it follows that
step of the optimization procedure by respectively solving
φHP F (f1 ) + φHP F (f2 ) = −θ2 (f1 ) − θ2 (f2 ), (11)
G012 (fi ) = 0, i = 1, 2. (7)
with φHP F (fi ) defined as [13]
Subsequently, the susceptance of the shunt elements is ob-  
0
tained as B(fi ) = B12 (fi ). 4πCfi Z02
φHP F (fi ) = tan−1 . (12)
(2πCfi Z02 )2 − 1
III. I MPLEMENTATION OF D ECOUPLING N ETWORK The capacitance C can be found by solving the non-linear
Transmission lines with specified values of electrical length equation (11). Subsequently, L can be calculated from [13]
at two different frequencies can be realized by composite Z02
left/right handed transmission lines [12]. However, this ap- L= . (13)
2πf1 sin (φHP F (f1 ))
proach would require a large number of lumped elements.
From (9), the required phase shift of the band-pass filter is
A simplified approach can be adopted by using a simple
transmission line section with characteristic impedance Z01 φBP F (f1 ) = −φHP F (f1 ) − θ2 (f1 ). (14)
 3

The component values for the band-pass filter are given by

[13]
Z02 Z02
L1 = rb L2 =
2πf0 2πf0 b
1 b
C1 = C2 = , (15)
2πrbf0 Z02 2πf0 Z02
where
p
1 + tan |φBP F |2 + 1 + tan |φBP F |2
r= (16)
1 + tan |φBP F |2
and p Fig. 4. Dual band monopole antenna pair with 4.5 mm separation with
fr −1 + 2/r proposed decoupling network and single stub matching networks.
b= , (17)
1 − fr2
with fr = f1 /f0 being the normalized frequency. The shunt
The option of using a transmission line section and band-
susceptance B(f1 ) and B(f2 ) can be realized as a parallel
pass filter to implement line 2 was also considered. Results
or series LC circuit. Element values can be calculated using
showed that the transmission line and band-pass filter approach
relations provided in [10].
yields inferior bandwidth compared to the high-pass filter
Alternatively, a transmission line section can be used instead
solution.
of the high-pass filter. The physical length of the transmission
line with characteristic impedance of Z02 can be calculated by TABLE I
D ESIGN PARAMETERS FOR D ECOUPLING N ETWORK
θ2 (f1 ) + θ2 (f2 )
lT L = vp , (18)
2π(f1 + f2 ) Z01 45 Ω
θ1 (f1 ) 125◦
where vp is the phase velocity of the line. From (9), the Line 1
θ1 (f2 ) 170◦
(Transmission Line)
required phase shift from the band-pass filter would be Length 34.5 mm
Track width 2.11 mm
2πf1 lT L
φBP F (f1 ) = − θ2 (f1 ). (19) Line 2
Z02 46 Ω
vp θ2 (f1 ) 320◦ (−40◦ )
(Filters)
θ2 (f2 ) 344◦ (−16◦ )
IV. M EASUREMENT R ESULTS φHP F (f1 ) 32◦
To verify the theory, a pair of printed monopole antennas φHP F (f2 ) 24◦
High-pass filter
C 6.68 pF
were used. The length of the monopoles were chosen to L 7.2 nH
achieve resonance at f1 = 1.8 GHz and f2 = 2.45 GHz. |φBP F (fi )| 8◦
Using the procedure described in Sections II and III, the design L1 1.64 nH
parameters were computed and are shown in Table I. C1 3.5 pF
Band-pass filter
L2 14.2 nH
Discrete inductors and capacitors from Murata’s LQW se-
C2 0.4 pF
ries and GJM series were used on Rogers RO4003 substrate Shunt Elements L 23.4 nH
(r = 3.55, tan δ = 0.0027, thickness = 0.81 mm) to realize (Series LC circuit) C 0.25 pF
the prototype. An Agilent Technologies 8510C Network An-
alyzer was used to measure the scattering parameters. The
final prototype is displayed in Fig. 4. The antenna pair is TABLE II
characterized by the measured scattering parameters shown P ERFORMANCE OF D ECOUPLING AND M ATCHING N ETWORK
in Fig. 5. Port coupling |S21 | is high, being −4.33 dB at
Return Loss |S11 (f1 )| −12.8 dB
f1 and −6.25 dB at f2 . Note that the resonance of the two Return Loss |S22 (f2 )| −13.1 dB
antennas is shifted from the original design values due to Without decoupling & Port Coupling |S21 (f1 )| −4.33 dB
the effects of mutual coupling. Conventional stub matching matching networks Port Coupling |S21 (f2 )| −6.25 dB
networks were used to match the decoupled ports (100 and Antenna Efficiency (f1 ) 57.3%
200 ) to the system impedance of Z0 = 50 Ω, as seen in Antenna Efficiency (f2 ) 71.2%
Return Loss |S11 (f1 )| −18.6 dB
Fig. 4. However this is achieved at the penalty of reduced
Return Loss |S22 (f2 )| −22.4 dB
bandwidth, which is a common trait for decoupling circuits, With decoupling & Port Coupling |S21 (f1 )| −22.6 dB
as pointed out in [7] and [14]. This results in matched ports matching networks Port Coupling |S21 (f2 )| −28.2 dB
with increased isolation between them, as displayed in Fig. Antenna Efficiency (f1 ) 88.4%
6. The performance of the decoupling and matching network Antenna Efficiency (f2 ) 85.2%
is summarized in Table II. Note that the quoted values for
efficiency are simulated values which exclude metalization
losses. The lower dissipation of the decoupling and matching
network at the higher frequency is due to differences in current
distribution at the two frequencies.
 4

0 0

-5
-5
-10

-15 -10 S11

|S| (dB)
|S| (dB)

-20
-15
S21
S11
-25 f1 f2 S22
S21
-30 -20
S22
-35
-25 f1 f2
-40

-45 -30
1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 2.6 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 2.6
Frequency (GHz) Frequency (GHz)
Fig. 5. Measured scattering parameters for the original antenna pair. Fig. 6. Measured scattering parameters of antenna pair with decoupling and
matching network.

V. C ONCLUSION
This paper describes a method for achieving port isola-
tion between two closely-spaced antennas operating at two
different frequencies. The theory is verified by measured
results on a fabricated prototype. The operating bandwidth
with the addition of the decoupling and matching network
can be improved by decreasing the initial mutual coupling
and impedance mismatch observed at the antenna terminals.
Note that the intention is to use a separate port for each of the
frequencies while maintaining isolation at both frequencies.
Isolated ports were therefore individually matched to only
frequencies f1 and f2 , respectively. Theoretically, the proposed
design can be applied to an arbitrary set of frequencies.
However, increased frequency separation may require higher
order filter networks.
R EFERENCES
[1] T. Rappaport, A. Annamalai, R. Buehrer and W. Tranter, “Wireless
communications: past events and a future perspective,” IEEE Commun.
Mag., vol. 40, no. 5, pp. 148–161, May 2002.
[2] H.-S. Lui, H. T. Hui and M. S. Leong, “A note on the mutual-coupling
problems in transmitting and receiving antenna arrays,” IEEE Antennas
Propagat. Mag., vol. 51, pp. 171–176, Oct. 2009.
[3] J. Shi, Q. Luo and H. Sun, “Channel reciprocity of compact antenna
array and the calibration,” in 2011 IEEE 22nd Int. Symp. Personal Indoor
and Mobile Radio Communication (PIMRC), Sept. 2011, pp. 1953–1957.
[4] H. Dong, T. Wu, K. Cheema, B. Abbott, C. Finch and H. Foo, “Design
of miniaturized RF SAW duplexer package,” IEEE Trans. Ultrason.
Ferroelectr. Freq. Control, pp. 849–858, July 2004.
[5] T. Yang, P.-L. Chi and T. Itoh, “Compact quarter-wave resonator and its
applications to miniaturized diplexer and triplexer,” IEEE Trans. Microw.
Theory Tech., vol. 59, pp. 260–269, Feb. 2011.
[6] A. Diallo, C. Luxey, P. Le Thuc, R. Staraj and G. Kossiavas, “Study
and reduction of the mutual coupling between two mobile phone PIFAs
operating in the DCS1800 and UMTS bands,” IEEE Trans. Antennas
Propag., vol. 54, pp. 3063–3074, Nov. 2006.
[7] J. Weber, C. Volmer, K. Blau, R. Stephan and M. Hein, “Miniaturized
antenna arrays using decoupling networks with realistic elements,” IEEE
Trans. Microw. Theory Tech., vol. 54, pp. 2733 –2740, June 2006.
[8] S.C. Chen, Y.S. Wang and S.J. Chung, “A decoupling technique for
increasing the port isolation between two strongly coupled antennas,”
IEEE Trans. Antennas Propag., vol. 56, pp. 3650–3658, Dec. 2008.
[9] J.C. Coetzee and Y. Yu, “Closed-form design equations for decoupling
networks of small arrays,” Electron Lett., vol. 44, pp. 1441–1442, Dec.
2008.
[10] J. C. Coetzee, “Dual-frequency decoupling networks for compact an-
tenna arrays,” Int. J. Microw. Sci. Technol., vol. 2011, pp. 3459–3466,
2011.
[11] X. Tang, K. Mouthaan and J. C. Coetzee, “Dual-band decoupling and
matching network design for very closely spaced antennas,” in Proc.
2012 European Microwave Conference, to be published.
[12] I.-H. Lin, M. DeVincentis, C. Caloz and T. Itoh, “Arbitrary dual-band
components using composite right/left-handed transmission lines,” IEEE
Trans. Microw. Theory Tech., vol. 52, pp. 1142–1149, April 2004.
[13] X. Tang and K. Mouthaan, “Design considerations for octave-band phase
shifters using discrete components,” IEEE Trans. Microw. Theory Tech.,
vol. 58, pp. 3459–3466, Dec. 2010.
[14] H.J. Chalupka, X. Wang and J.C. Coetzee, “Compact arrays for mobile
platforms: Trade-off between size and performance for SDMA and
MIMO applications,” in Proc. 48th Internationales Wissenschaftliches
Kolloquium, Techniche Universität Ilmenau, 22–25 Sept. 2003.

View publication stats