Design of Dual-Band MIMO Antenna
with High Isolation for WLAN Mobile Terminal
Jung-Nam Lee, Kwang-Chun Lee, Nam-Hoon Park, and Jong-Kweon Park
In this paper, we propose a dual-band multiple-input
multiple-output (MIMO) antenna with high isolation for
WLAN applications (2.45 GHz and 5.2 GHz). The
proposed antenna is composed of a mobile communication
terminal board, eight radiators, a coaxial feed line, and
slots for isolation. The measured –10 dB impedance
bandwidths are 10.1% (2.35 GHz to 2.6 GHz) and 3.85%
(5.1 GHz to 5.3 GHz) at each frequency band. The
proposed four-element MIMO antenna has an isolation of
better than 35 dB at 2.45 GHz and 45 dB at 5.2 GHz
between each element. The antenna gain is 3.2 dBi at 2.45
GHz and 4.2 dBi at 5.2 GHz.
Keywords: MIMO, dual-band antenna, isolation,
WLAN, PIFA.
Manuscript received Apr. 23, 2012; revised Sept. 4, 2012; accepted Sept. 13, 2012.
This work was supported by the IT R&D program of MKE/KEIT, Korea (10038765,
Development of B4G Mobile Communication Technologies for Smart Mobile Services).
Jung-Nam Lee (phone: +82 42 860 1884, jnlee77@etri.re.kr), Kwang-Chun Lee
(kclee@etri.re.kr), and Nam-Hoon Park (nhpark@etri.re.kr) are with the Communications
Internet Research Laboratory, ETRI, Daejeon, Rep. of Korea.
Jong-Kweon Park (ingpark@hanbat.ac.kr) is with the Department of Radio Wave
Engineering, Hanbat National University, Daejeon, Rep. of Korea.
http://dx.doi.org/10.4218/etrij.13.0112.0250
ETRI Journal, Volume 35, Number 2, April 2013
© 2013
I. Introduction
A technology for innovatively increasing the transfer rate is
required to provide an improved quality of service for wireless
transmissions. Various kinds of multimedia services are
required in a wireless environment. The amount of data to be
transmitted has increased, as has the speed of data
transmissions. Therefore, we conduct a study on a method for
efficiently using limited frequencies. As part of the study,
research on a multiple-input multiple-output (MIMO) system
that uses a channel in the spatial domain is actively conducted.
MIMO technology uses multiple antennas at both the
transmitter and receiver to simultaneously transmit multiple
signals using the same wireless channel. MIMO technology
increases the channel capacity within limited frequency
resources and provides a high data transmission rate. Further,
MIMO technology can increase the capacity of wireless data
tenfold without using additional frequencies and can transmit a
wide range of data with high reliability. The capacity of a
MIMO system is reduced owing to connections between
receiver signals. The connections between signals received
from different antenna devices are very important parameters
in a MIMO system [1]-[8], such as compact dual-printed
inverted-F antenna diversity systems for portable wireless
devices [1], a modified planar inverted-F antenna (PIFA) and
its array for MIMO terminals [2], an integrated MIMO antenna
with a high-isolation characteristic [3], a compact four-element
diversity-antenna array for personal digital assistant (PDA)
terminals in a MIMO system [4], a three-antenna MIMO
system for WLAN operation in a PDA phone [5], a novel
miniaturized tri-band PIFA for MIMO applications [6], design
considerations for a low antenna correlation and mutual
coupling reduction in multi-antenna terminals [7], and a
Jung-Nam Lee et al.
177
Antenna 2
50-Ω coaxial
(2.45-GHz radiator)
feed cable
5.2-GHz radiator
Bending line
Antenna 1
(2.45-GHz radiator)
[mm]
5.2-GHz radiator
6.2
Ground plane
(50 mm × 100 mm)
of a mobile phone
6
4.5 2
9.5
4.5
5.5
PW2
(4.2 mm)
Antenna 4
(2.45-GHz radiator)
5.2-GHz radiator
5.2-GHz radiator
Antenna 3
(2.45-GHz radiator)
UH:8 mm
compact antenna array for MIMO applications at 1,800 MHz
and 2,450 MHz [8]. Since a large number of antenna devices
are used in a MIMO system, when the antennas are mounted to
a mobile terminal, the distance between the antennas becomes
very short, which may create stronger connections between
them. Since the antennas are connected to each other, a
relatively low gain is obtained.
In this paper, we propose a dual-band MIMO antenna with
high isolation for WLAN applications (2.45 GHz and 5.2
GHz). The proposed antenna is composed of a first radiator
with one end connected to a feed line and receiving a signal
within the first frequency band (2.45 GHz) and a second
radiator having one end connected to a ground plane and
receiving a signal within the second frequency band (5.2 GHz).
To improve the isolation of the proposed antenna, narrow
rectangular slots are inserted on the ground plane [1]. The
measured –10 dB impedance bandwidths are 10.1% (2.35
GHz to 2.6 GHz) and 3.85% (5.1 GHz to 5.3 GHz) at each
frequency band. The proposed four-element MIMO antenna
has an isolation of better than 35 dB at 2.45 GHz and 45 dB at
5.2 GHz between each element. The antenna gain is 3.2 dBi at
2.45 GHz and 4.2 dBi at 5.2 GHz.
A
B
A: feeding point (1.5 mm)
Uw:4.2mm B: shorting point (1.0 mm)
1.5 mm
(a)
HS1
LS1 HS1
HS2
LS2
HS2
5.2-GHz slot
2.45-GHz slot
2.45-GHz slot
II. MIMO Antenna Design
An important aspect of the proposed antenna is a
configuration in which the frequency signals in different bands
do not affect each other when a plurality of small antennas are
formed in a mobile communication device. In this
configuration, the mobile communication device can achieve
increased isolation among the plurality of antennas.
The proposed antenna is composed of four radiators, an
isolation improving slot, a feed line, and a mobile
communication terminal board. Figure 1 shows the geometry
of the proposed MIMO antenna. The ground plane has a
dimension of 50 mm × 100 mm. A dielectric substrate with a
height of 0.8 mm and a relative dielectric constant of 4.5 is
used. The antenna is excited using a 50-Ω coaxial feed line.
The optimal parameters can be chosen as Hs1 = 10 mm, Hs2 = 9
mm, Ls1 = 18 mm, and Ls2 = 21.5 mm based on an extensive
simulation using Ansys HFSS [9].
The first radiator (2.45-GHz radiator) has one end connected
to a power feeding line and receives a signal within the first
frequency band (2.45 GHz). The second radiator (5.2-GHz
radiator) has one end connected to a ground plane and receives
a signal within the second frequency band (5.2 GHz). The first
stub (PW2) is extended from the other end of the first radiator
and finely adjusts the signal received by the first radiator. The
second stub (UW) is extended from the other end of the second
radiator and finely adjusts the signal received by the second
178
Jung-Nam Lee et al.
5.2-GHz slot
(c)
(b)
Fig. 1. Detailed view, (b) slot on ground plane, and (c)
photograph of proposed MIMO antenna.
radiator and a shorting plate electrically connecting the first
radiator to the ground plane.
In the antenna described in this paper, the first radiator is only
connected to the coaxial feed line, and the second radiator may
not be connected to the coaxial feed line but is connected to the
ground plane. The signal flows along the ground plane.
Therefore, even though the gain is slightly reduced, appropriate
isolation between channels is provided, and it is possible to
freely tune the received frequency signal.
As shown in Fig. 1(b), to improve the isolation between the
antennas mounted at the edges of the board, a plurality of slots
may be formed in the ground plane [1]. Since it is possible to
alter the path of the current flowing directly toward the
neighboring antennas through the ground plane using individual
slots, the isolation between antennas can be improved.
III. Measured and Simulated Results
Several cases for the available area are considered and
ETRI Journal, Volume 35, Number 2, April 2013
2
1
Case 1
4
2
1
Case 2
3
4
Case 3
3
4
0
2
3
Return loss (dB)
1
–10
–20
S11 (measured)
(a)
S22 (measured)
0
S33 (measured)
–30
S44 (measured)
–10
0
–40
–10
S11 (case 1)
–50
S21 (case 1)
S31 (case 1)
–60
S41 (case 1)
1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0
Frequency (GHz)
Magnitude (dB)
(b)
2.8
3.2
3.6 4.0 4.4 4.8
Frequency (GHz)
5.2
5.6
6.0
–30
–40
–50
–60
–10
–70
2.0
–20
S21 (measured)
S31 (measured)
S41 (measured)
–20
0
2.4
2.8
3.2
3.6 4.0
4.4
Frequency (GHz)
4.8
5.2
(b)
–30
Fig. 3. Measured (a) return loss and (b) isolation (with slot).
–40
S11 (case 2)
S22 (case 2)
S21 (case 2)
S31 (case 2)
S41 (case 2)
–50
–60
1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0
Frequency (GHz)
(c)
0
–10
Magnitude (dB)
2.4
(a)
–30
Isolation (dB)
Magnitude (dB)
2.0
–20
–20
–30
–40
S11 (case 3)
–50
S21 (case 3)
–60
S41 (case 3)
S31 (case 3)
1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0
Frequency (GHz)
(d)
Fig. 2. Comparisons of simulated coupling and reflection
coefficients of proposed antenna for three different
configurations: (a) three cases, (b) case 1, (c) case 2,
and (d) case 3.
ETRI Journal, Volume 35, Number 2, April 2013
simulated using Ansys HFSS, and the coupling and reflection
coefficients are compared to select the best position for the
elements.
Figure 2 shows the coupling and reflection coefficients for
the three considered cases. The antenna elements are better
matched at 2.45 GHz and 5.2 GHz for case 1, which is
therefore chosen.
Figure 3(a) shows the measured return losses of the proposed
antenna for case 1. The antenna is measured using an Anritsu
vector network analyzer in an anechoic chamber. The return
losses are greater than 10 dB from 2.35 GHz to 2.6 GHz and
from 5.1 GHz to 5.3 GHz. Figure 3(b) shows the measured
isolation of the four-element antenna of Fig. 1. As can be seen
from the figure, the antenna isolations have about 35 dB (S21),
29 dB (S31), and 30 dB (S41) at 2.45 GHz, and 45 dB (S21), 38
dB (S31), and 35 dB (S41) at 5.2 GHz, respectively.
Many techniques for improving the isolation have been
introduced [10]-[18]. It is difficult to integrate other elements
inside the mobile terminal because the electronic band gap
bores several holes into the ground plane [10]-[12]. The
isolation is much improved when a short-circuit strip with
Jung-Nam Lee et al.
179
0
S21(without slot)
S31(without slot)
S41(without slot)
Return loss (dB)
Isolation (dB)
–10
0
S21(with slot)
S31(with slot)
S41(with slot)
–20
–30
–10
–20
–40
Pw2=4 mm
–30
–50
2.40
2.45
Frequency (GHz)
(a)
0
–10
Pw2=5 mm
Pw2=6 mm
2.50
1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0
Frequency (GHz)
(a)
S21(without slot)
S21(with slot)
S31(without slot)
S41(without slot)
S31(with slot)
S41(with slot)
0
–20
–10
Return loss (dB)
Isolation (dB)
Pw2=2 mm
Pw2=3 mm
–30
–40
–20
Uw=3 mm
Uw=4 mm
–50
–60
5.10
–30
5.15
5.20
Frequency (GHz)
(b)
5.25
Uw=2 mm
Uw=5 mm
Uw=6 mm
5.30
1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0
Frequency (GHz)
(b)
Fig. 4. Comparison of measured isolations of proposed antenna:
(a) 2.45-GHz isolations (with and without slot) and (b)
5.2-GHz isolations (with and without slot).
0
Without slot
With slot
S21
S31
S41
S21
S31
S41
2.45 GHz
18 dB
15 dB
17 dB
35 dB
29 dB
31 dB
5.2 GHz
15 dB
19 dB
15 dB
45 dB
38 dB
35 dB
Return loss (dB)
–10
Table 1. Isolations with and without slot.
UH=8 mm
UH=7 mm
–20
–30
–40
many forms is added between the antennas [13]-[16]. However,
the mobile terminal is affected manually. When using the
parasitic element [17], [18], the space to arrange another
element in the mobile terminal is decreased. In this paper, we
form a slot on the ground plane to improve the isolation
method [1].
Figure 4 and Table 1 show a comparison of the measured
isolations with and without the use of a slot. When a slot is not
used, the isolation at operating frequencies of 2.45 GHz and 5.2
GHz is measured at less than –20 dB. On the other hand, when
a slot is used, the isolation of the operating frequency is
180
Jung-Nam Lee et al.
UH=6 mm
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
Frequency (GHz)
(c)
6.5 7.0
Fig. 5. Effects of antenna geometry (PW2, UW, and UH) on return
loss: (a) PW2, (b) UW, and (c) UH.
measured at over –30 dB.
We study the effects of the antenna geometry (PW2, UW, and
UH) on the return loss. The effects of a 2.45-GHz tuning stub
length (PW2) on the return loss are shown in Fig. 5(a). From the
figure, the length of the tuning stub has a profound effect on the
lower frequency band (2.45 GHz) but only a small effect on the
ETRI Journal, Volume 35, Number 2, April 2013
upper frequency band (5.2 GHz). As the length of the tuning
stub (PW2) increases, the operating frequency band (2.45 GHz)
is moved to a lower frequency. The effects of the 5.2-GHz
radiator length (UW) and height (UH) on the return loss are
shown in Figs. 5(b) and 5(c), respectively. As shown in the
figures, changes of the 5.2-GHz radiator length (UW) and
height (UH) have a negligible effect on the lower frequency
band (2.45 GHz). Only characteristics of the upper frequency
band (5.2 GHz) are changed. As the length (UW) and height
(UH) of the 5.2-GHz radiator increase, the operating frequency
band (5.2 GHz) is shifted to the lower frequency. According to
the above results, the lower (2.45 GHz) and upper frequency
(5.2 GHz) bands can be controlled by changing PW2, UW, and
UH.
To analyze in detail the effect of the isolation slot on the
ground plane, the excited E-field and H-field and the surface
current in the ground plane at 2.45 GHz and 5.2 GHz for the
proposed MIMO system are studied. The E and H fields and
the surface current are simulated using CST Microwave Studio
[19]. Figures 6(a) through 6(c) show the simulated E and H
fields and surface current for the case of antenna 1 excited at
2.45 GHz. As shown in the figures, it can be seen that the E and
H fields and surface current are mostly focused in the 2.45GHz slot and less so in the 5.2-GHz slot. The isolations of S21,
S31, and S41 are greatly improved, which agrees with the results
shown in Fig. 4(a). The simulated results for the case of
antenna 1 excited at 5.2 GHz are shown in Figs. 6(d) through
6(f). As shown in the figures, the E and H fields and surface
current are mostly focused in the 5.2-GHz slot and less so in
the 2.45-GHz slot. The simulated results in Fig. 4(b) show the
improvement in the measured isolations of S21, S31, and S41.
Figs. 6(g) and 6(h) respectively show the surface current of
2.45 GHz mostly focused in the 2.45-GHz radiator and the
surface current of 5.2 GHz mostly focused in the 5.2-GHz
radiator. The radiators (2.45 GHz and 5.2 GHz) do not affect
one another.
The measured radiation patterns of the proposed MIMO
antenna at resonant frequencies of 2.45 GHz and 5.2 GHz are
plotted in Fig. 7. The radiation pattern results for three planes
are shown, and the Eθ and Eφ components appear to be
comparable in all three planes. The radiation patterns of the
proposed MIMO antenna are typical of those of conventional
PIFAs. For frequencies of over 2.45 GHz (2.35 GHz to 2.6
GHz) and 5.2 GHz (5.1 GHz to 5.3 GHz), the radiation
patterns are similar to those shown in Figs. 7(a) through 7(f).
Stable measured peak MIMO antenna gains of about 3.2 dBi at
2.45 GHz and 4.2 dBi at 5.2 GHz are obtained.
Calculation of the antenna correlation can be done using two
methods. The first method uses a far-field radiation pattern [20],
and the second method uses the scattering parameters [21]. In
ETRI Journal, Volume 35, Number 2, April 2013
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(f)
Fig. 6. Simulated E-field, H-field, and surface current at 2.45 and
5.2 GHz: (a) 2.45-GHz E-field, (b) 2.45-GHz H-field, (c)
2.45-GHz surface current, (d) 5.2-GHz E-field, (e) 5.2GHz H-field, (f) 5.2-GHz surface current, (g) 2.45-GHz
radiator surface current, and (h) 5.2-GHz radiator surface
current.
this paper, the envelope correlation is calculated using the
scattering parameters, based on a formula proposed by Blanch
[22]:
ρe =
*
| S11* S12 + S 21
S 22 |2
.
2
(1 − (| S11 | + | S21 | ))(1 − (| S22 |2 + | S12 |2 ))
2
In the design of small antennas on a mobile phone or small
electric device, the radiation efficiencies are low, owing to the
compressed electrical size of the mobile terminals. Therefore,
the energy conservation in the envelope correlation is expected
to be inadequate. The expression using the scattering
parameters has a disadvantage in the case of small mobile
terminal antennas. However, the envelope correlation
calculated using the scattering parameters provides sufficiently
Jung-Nam Lee et al.
181
0
0
330
0
30
300
30
60
–20
0
330
0
300
60
–20
330
0
–10
30
300
–10
60
90
–20
Eθ
Eφ
240
0
210
180
0
120
90
–40 270
–20
240
0
150
210
300
60
90
–20
Eθ
Eφ
240
0
210
330
0
–10
210
(a)
0
30
300
–30 270
90
210
330
0
–10
180
2.45 GHz
xy-plane
120
150
z
30
90
300
–10
60
120
150
120
210
180
–10
120
150
60
0
300
60
–10
90
–30 270
–10
–10
90
Eθ
Eφ
Eθ
Eφ
10
330
10
30
–10
60
120
150
–50 270
–10
330
10
5.2 GHz
xz-plane
120
210
330
180
150
z
30
180
150
y
0
z
30
300
300
60
10
60
–20
90
Eθ
Eφ
120
240
210
10
330
180
60
120
210
180
–10
150
90
Eθ
Eφ
–20
120
240
10
210
10
330
180
150
0
30
300
60
90
–30 270
Eθ
Eφ
–20
–10
120
240
120
240
0
210
10
330
180
150
210
10
30
90
–20
–20
120
90
210
180
150
(e)
5.2 GHz
xy-plane
180
150
210
330
180
z
30
300
–20
210
180
330
–20
0
150
180
150
0
30
300
60
–40 270
120
240
120
210
x
–20
90
Eθ
Eφ
90
240
0
60
60
Eθ
Eφ
y
0
30
300
0
150
0
–40 270
–20
120
–40 270
0
–20
Eθ
Eφ
240
330
0
60
Eθ
Eφ
240
30
–40 270
0
60
0
300
0
150
300
–30
240
60
–30 270
–10
90
0
–50 270
90
Eθ
Eφ
–30
x
Eθ
Eφ
0
–30
–50 270
300
–20
10
30
300
10
x
–10
–30
–10
330
–50 270
–10
y
0
30
0
0
–30
240
10
–10
90
Eθ
Eφ
–30
0
0
210
–30 270
–10
330
–20
–10
0
–30
–30
150
180
(d)
0
300
5.2GHz
yz-plane
120
240
–20
240
180
210
10
30
–20
210
120
240
0
(c)
–10
–10
0
300
0
150
180
90
Eθ
Eφ
0
10
30
–20
240
0
330
10
0
–30 270
90
Eθ
Eφ
–20
60
–20
0
–20
–30 270
210
–20
240
330
30
–30 270
90
Eθ
Eφ
240
0
–10
Eθ
Eφ
0
–20
–10
–30 270
180
150
0
–10 300
60
–20
x
330
0
180
–20
10
30
60
210
210
(b)
300
y
0
z
120
240
x
0
0
90
Eθ
Eφ
y
30
150
–20
–10
150
300
–10
0
–20
Eθ
Eφ
240
0
–10
60
–20
–10
330
0
180
60
–20
0
0
–20
120
30
300
–30 270
–10
–20
0
150
–20
Eθ
Eφ
240
2.45 GHz
xz-plane
120
–30 270
90
–20
180
180
330
–10
–40 270
120
210
0
60
Eθ
Eφ
240
–10
0
30
300
–20
–40 270
120
150
0
330
0
–20
180
90
–30 270
–20
Eθ
Eφ
30
330
0
2.45 GHz
yz-plane
0
–20
–20
–40 270
330
0
90
Eθ
Eφ
120
240
210
180
150
(f)
Fig. 7. Measured radiation patterns of proposed MIMO antenna: (a) yz plane (2.45 GHz), (b) xz plane (2.45 GHz), (c) xy plane
(2.45 GHz), (d) yz plane (5.2 GHz), (e) xz plane (5.2 GHz), and (f) xy plane (5.2 GHz).
accurate results.
Figure 8 shows the simulated envelope correlation
182
Jung-Nam Lee et al.
coefficients obtained. The scattering parameters are found
using CST Microwave Studio. The calculated envelope
ETRI Journal, Volume 35, Number 2, April 2013
1.0
ρe 21
Mobile
case
ρe 31
Correlation coefficient
0.8
ρe 41
2.45 GHz
5.2 GHz
ρe 32
ρe 42
0.6
LCD
ρe 43
Battery
0.4
Head
phantom
0.2
(a)
0.0
2
3
4
Frequency (GHz)
5
6
Fig. 8. Simulated envelope correlation coefficient.
Table 2. Envelope correlation coefficients.
Operating frequency 2.45 GHz
ρe21
ρe31
ρe41
ρe32
ρe42
ρe43
0.0020
0.0027
0.0110
0.0110
0.0027
0.0028
Operating frequency 5.2 GHz
ρe21
ρe31
ρe41
ρe32
ρe42
ρe43
0.0030
0.0005
0.0058
0.0058
0.0005
0.0030
correlation coefficients are depicted in Table 2. As shown in
Table 2, the envelope correlation coefficients of the proposed
MIMO antenna are substantially less than 0.5 at both 2.45 GHz
and 5.2 GHz. Therefore, the four antennas are highly
decorrelated.
IV. SAR Simulation of Hand and Head Phantoms
To analyze the characteristic of the MIMO antenna when the
antenna is positioned in the hand and head, the antenna is
simulated by applying the hand and head phantoms.
The proposed antenna includes a mobile case, LCD, and
battery. The mobile case is made of rubber (εr = 3), and the
LCD and battery are PEC. The hand and head phantoms are
applied to the earlier case. To model the hand and head
phantoms, we use the shell (εr = 5, loss tangent = 0.05, and
material density = 1,000 kg/m3) and liquid (εr = 42, loss tangent
= 0.99, and material density = 1,000 kg/m3) dielectric material.
Figure 9 shows the specific absorption rate (SAR) simulation
model and the simulated 1-g SAR values for the proposed
antenna. The SAR simulated using the IEEE C95.3 averaging
method and the input power for the SAR testing is 0.126 W (24
dBm) for the WLAN frequency bands (2.45 and 5.2 GHz) [23].
The SAR values for 1-g tissue in the cases of the hand and
ETRI Journal, Volume 35, Number 2, April 2013
(b)
Fig. 9. SAR simulation modeling: (a) head phantom and (b) hand
phantom.
head are studied using the SAR simulation model provided by
CST Microwave Studio. The return loss and isolation for the
two cases at each operating frequency is given in Fig. 10.
Figure 10 shows the simulated return loss, isolation, and
correlation coefficients resulting from using the hand and head
phantoms. It is clear that there is no distinguishable difference
in return loss between using the hand phantom and using only
the antenna (without the hand phantom). The isolation level
increases with the hand phantom. However, the envelope
correlation coefficients are substantially less than 0.5. When
applying the head phantom, the lower (2.45 GHz) and upper
(5.2 GHz) frequency bands move to the low frequency, and the
level of isolation increases. Nevertheless, the envelope
correlation coefficients are substantially less than 0.5.
The comparison of the simulated results for the three cases
(that is, using antenna only, hand phantom, or head phantom) at
each operating frequency is given in Tables 3 and 4.
Figure 11 shows the simulated SAR field distribution of the
hand and head phantoms excited at 2.45 GHz and 5.2 GHz.
We simulate the effect of the SAR by the hand and head
phantoms. The input power is 24 dBm at 2.45 GHz and 5.2
GHz, and the average SAR of about 1 g is obtained for the
position of the hand and head phantoms. The average SARs for
when the MIMO antennas are located in the hand and head
phantoms are shown in Table 5.
V. Conclusion
A dual-band MIMO antenna with high isolation for WLAN
applications (2.45 GHz and 5.2 GHz) was designed and
presented in this paper. The proposed antenna is composed of
two radiators, the first with one end connected to a feed line
and receiving a signal within the first frequency band (2.45
GHz) and the second with one end connected to the ground
plane and receiving a signal within the second frequency band
(5.2 GHz). To improve the isolation of the proposed antenna,
Jung-Nam Lee et al.
183
0
0
S11 (with hand phantom)
S22 (with hand phantom)
S33 (with hand phantom)
S44 (with hand phantom)
–20
–30
2.0
2.4
2.8
3.2
S11 (with head phantom)
S22 (with head phantom)
S33 (with head phantom)
S44 (with head phantom)
–30
–40
3.6 4.0 4.4 4.8
Frequency (GHz)
(a)
5.2
5.6
6.0
2.0
2.4
2.8
3.2
3.6 4.0 4.4 4.8
Frequency (GHz)
(b)
5.2
5.6
6.0
1.0
0
1.0
0.9
–10
0.9
–20
0.8
–20
S21(with hand phantom)
S31(with hand phantom)
S41(with hand phantom)
ρ21(with hand phantom)
ρ31(with hand phantom)
ρ41(with hand phantom)
–40
–50
2.45 GHz
–60
0.7
0.6
0.5
0.4
–70
0.3
–80
0.2
–90
0.1
–100
2.0
2.4
2.8
3.2
3.6 4.0 4.4 4.8
Frequency (GHz)
5.2
5.6
0.8
5.2 GHz
–30
Isolation (dB)
5.2 GHz
Correlation coefficient
0
–10
–30
Isolation (dB)
–20
–40
2.45 GHz
–50
–60
0.7
0.6
0.5
0.4
–70
0.3
–80
0.2
–90
0.1
–100
2.0
0.0
6.0
S21(with head phantom)
S31(with head phantom)
S41(with head phantom)
ρ21(with head phantom)
ρ31(with head phantom)
ρ41(with head phantom)
2.4
2.8
3.2
3.6 4.0 4.4 4.8
Frequency (GHz)
(c)
5.2
Correlation coefficient
–10
Return loss (dB)
Return loss (dB)
–10
5.6
0.0
6.0
(d)
Fig. 10. Simulated isolation and correlation coefficients: (a) return loss with hand phantom, (b) return loss with head phantom, (c)
isolation and correlation coefficient with hand phantom, and (d) isolation and correlation coefficient with head phantom.
Table 3. Simulated isolation and correlation coefficient comparison.
Frequency
2.45 GHz
Isolation
(dB)
5.2 GHz
Phantom
S21
S31
S41
Only
antenna
35
29
31
Hand
15
15
20
Head
13
14
17
Only
antenna
45
Hand
20
23
14
Head
16
15
12
38
35
Frequency
2.45 GHz
Correlation
coefficient
5.2 GHz
Phantom
S21
S31
S41
Only
antenna
0.0020
0.0027
0.011
Hand
0.027
0.059
0.044
Head
0.093
0.127
0.063
Only
antenna
0.003
0.0005
0.0058
Hand
0.029
0.025
0.022
Head
0.042
0.068
0.080
Table 4. Simulated return loss comparison.
Frequency Phantom
Return
loss 2.45 GHz
(dB)
184
Port 1
Port 2
Port 3
Port 4
Only
antenna
30
15
30
30
Hand
11
11
10
13
Head
12
10
10
11
Jung-Nam Lee et al.
Frequency Phantom
Return
loss
(dB)
5.2 GHz
Port 1
Port 2
Port 3
Port 4
Only
antenna
28
17
32
21
Hand
13
13
14
13
Head
16
9
10
8
ETRI Journal, Volume 35, Number 2, April 2013
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
(j)
(k)
(l)
(m)
(n)
(o)
(p)
Fig. 11. Simulated SAR field distribution of hand and head phantoms: (a) port 1 at 2.45 GHz, (b) port 2 at 2.45 GHz, (c) port 3 at 2.45
GHz, (d) port 4 at 2.45 GHz, (e) port 1 at 5.2 GHz, (f) port 2 at 5.2 GHz, (g) port 3 at 5.2 GHz, (h) port 4 at 5.2 GHz, (i) port 1
at 2.45 GHz, (j) port 2 at 2.45 GHz, (k) port 3 at 2.45 GHz, (l) port 4 at 2.45 GHz, (m) port 1 at 5.2 GHz, (n) port 2 at 5.2 GHz,
(o) port 3 at 5.2 GHz, and (p) port 4 at 5.2 GHz.
Table 5. Simulated SAR values.
Frequency
SAR
Phantom
Port 1
Port 2
2.45 GHz
SAR
(W/kg)
Hand
2.6
2.5
1.6
2.2
Head
0.7
0.7
0.49
0.52
ETRI Journal, Volume 35, Number 2, April 2013
Port 3
Port 4
Frequency
SAR
Phantom
Port 1
Port 2
Port 3
Port 4
5.2 GHz
SAR
(W/kg)
Hand
2.1
2.2
1.9
1.8
Head
0.3
0.43
0.32
0.19
Jung-Nam Lee et al.
185
narrow rectangular slots are inserted into the ground plane. The
measured –10 dB impedance bandwidths are 10.1% (2.35
GHz to 2.6 GHz) and 3.85 % (5.1 GHz to 5.3 GHz) at each
frequency band. The proposed four-element MIMO antenna
has an isolation of better than 35 dB at 2.45 GHz and 45 dB at
5.2 GHz between each element. The antenna gain is 3.2 dBi at
2.45 GHz and 4.2 dBi at 5.2 GHz. We have shown that the
envelope correlation coefficients of the proposed MIMO
antenna are substantially less than 0.5. To analyze the
characteristic of the MIMO antenna when the antenna was
positioned in the hand and on the head, the antenna was
simulated by applying the hand and head phantoms.
Propag., vol. 54, Jan. 2006, pp. 90-100.
[13] A. Diallo et al., “Study and Reduction of the Mutual Coupling
Between Two Mobile Phone PIFAs Operating in the DCS 1800
and UMTS Bands,” IEEE Trans. Antennas Propag., vol. 54, Nov.
2006, pp. 3063-3074.
[14] A. Chebihi et al., “A Novel Isolation Technique for Closely
Spaced PIFAs for UMTS Mobile Phones,” IEEE Antennas
Wireless Propag. Lett., vol. 7, 2008, pp. 665-668.
[15] K.L. Wong and W.Y. Chen, “Small-Size Printed Loop-Type
Antenna Integrated with Two Stacked Coupled-Fed Shorted Strip
Monopoles for Eight-Band LTE/GSM/UMTS Operation in the
Mobile Phone,” Microw. Opt. Technol. Lett., vol. 52, July 2010,
pp. 1471-1476.
[16]
Z.
Jin, J.H. Lim, and T.Y. Yun, “Small-Size and High-Isolation
References
MIMO Antenna for WLAN,” ETRI J., vol. 34. no. 1, Feb. 2012,
[1] M. Karaboikis et al., “Compact Dual-Printed Inverted-F Antenna
pp. 114-117.
Diversity Systems for Portable Wireless Devices,” IEEE [17] K.S. Min, D.J. Kim, and Y.M. Moon, “Improved MIMO Antenna
Antennas Wireless Propag. Lett., vol. 3, no. 1, 2004, pp. 9-14.
by Mutual Coupling Suppression between Elements,” European
[2] Y. Gao et al., “Modified PIFA and Its Array for MIMO Terminals,”
Conf. Wireless Technol., Paris, France, Oct. 2005, pp. 125-128.
IEE Proc. Microw. Antennas Propag., vol. 152, Aug. 2005, pp. [18] Z. Li, Z. Du, and K. Gong, “A Dual-Slot Diversity Antenna with
255-259.
Isolation Enhancement Using Parasitic Elements for Mobile
[3] K. Chung and J.H. Yoon, “Integrated MIMO Antenna with High
Handsets,” Asia Pacific Microw. Conf., Singapore, Dec. 2009, pp.
Isolation Characteristic,” Electron. Lett., vol. 43, Feb. 2007, pp.
1821-1824.
199-201.
[19] Computer Simulation Technology (CST), Microwave Studio
[4] C.C. Chiau, X. Chen, and C.G. Parini, “A Compact Four-Element
(MWS), 3D EM simulation software.
Diversity-Antenna Array for PDA Terminals in a MIMO System,” [20] R.G. Vaughan and J.B. Andersen, “Antenna Diversity in Mobile
Microw. Opt. Technol. Lett., vol. 44, Mar. 2005, pp. 408-412.
Communications,” IEEE Trans. Veh. Technol., vol. 36, Nov. 1987,
[5] K.L. Wong et al., “Three-Antenna MIMO System for WLAN
pp. 149-172.
Operation in a PDA Phone,” Microw. Opt. Technol. Lett., vol. 48, [21] R.G. Vaughan and J.B. Andersen, Channels, Propagation and
July 2006, pp. 1238-1242.
Antennas for Mobile Communications, Stevenage, England: IET,
[6] M. Manteghi and Y. Rahmat-Samii, “A Novel Miniaturized
Jan. 2003. http://www.theiet.org/resources/books/electro/19103.cfm
Triband PIFA for MIMO Applications,” Microw. Opt. Technol. [22] S. Blanch, J. Romeu, and I. Corbella, “Exact Representation of
Lett., vol. 49, Mar. 2007, pp. 724-731.
Antenna System Diversity Performance from Input Parameter
[7] J. Thaysen and K.B. Jakobsen, “Design Considerations for Low
Description,” Electron. Lett., vol. 39, no. 9, 2003, pp. 705-707.
Antenna Correlation and Mutual Coupling Reduction in Multi- [23] IEC, “Human Exposure to Radio Frequency Fields from Handantenna Terminals,” European Trans. Telecommun., vol. 18, 2007,
Held and Body-Mounted Wireless Communication Devices –
pp. 319-326.
Human Models, Instrumentation, and Procedures – Part 1:
[8] B. Lindmark and S. Garcia-Garcia, “Compact Antenna Array for
Procedure to Determine the Specific Absorption Rate (SAR) for
MIMO Applications at 1800 and 2450 MHz,” Microw. Opt.
Hand-Held Devices Used in Close Proximity to the Ear
Technol. Lett., vol. 48, Oct. 2006, pp. 2034-2037.
(Frequency Range of 300 MHz to 3 GHz),” IEC 62209-1, Feb.
[9] Ansys, High Frequency Structure Simulator (HFSS), ver. 10, 2005.
2005.
[10] D. Sievenpiper et al., “High-Impedance Electromagnetic Surfaces
with a Forbidden Frequency Band,” IEEE Trans. Microw Theory
Techniques, vol. 47, Nov. 1999, pp. 2059-2074.
[11] F. Yang and Y. Rahmat-Samii, “Microstrip Antennas Integrated
with Electromagnetic Band-Gap (EBG) Structures: A Low
Mutual Coupling Design for Array Applications,” IEEE Trans.
Antennas Propag., vol. 51, Oct. 2003, pp. 2936-2946.
[12] L. Li et al., “Locally Resonant Cavity Cell Model for
Electromagnetic Band Gap Structures,” IEEE Trans. Antennas
186
Jung-Nam Lee et al.
ETRI Journal, Volume 35, Number 2, April 2013
Jung-Nam Lee was born in the Rep. of Korea
in 1977. He received his BS and MS from the
Department of Information and Communication
Engineering, Hanbat National University,
Daejeon, Rep. of Korea, in 2004 and 2006,
respectively. He received his PhD in radio wave
engineering from Hanbat National University in
2010. He thereafter joined the Mobile RF Research Team of ETRI,
where he is currently a senior member of the engineering staff. His
research interests are small antenna, RFID antenna, UWB antenna, and
small base station antenna design.
Kwang-Chun Lee received his BS/MS in
electronics engineering from Chung-Ang
University, Seoul, Rep. of Korea, in 1986 and
1988, respectively. Since 1988, he has been with
ETRI, Daejeon, Rep. of Korea. His main research
interests include high efficiency power amplifiers
and RF technologies for mobile communication
system design.
Nam-Hoon Park received his BS from
Chonnam University, Gwangju, Rep. of Korea,
in 1984, his MS from Chung-Ang University,
Seoul, Rep. of Korea, in 1987, and his PhD
from Chungnam National University, Daejeon,
Rep. of Korea, in 1999, all in computer science.
Since 1988, he has been with ETRI, Daejeon,
Rep. of Korea, where he has participated in various communication
systems development. He is now working as the leader of the Mobile
Communications Technology Research Department. His research
interests include the design and development of computer networks,
broadband wireless signaling networks, next-generation mobile
communication systems, SDR, and cognitive radio.
Jong-Kweon Park was born in the Rep. of
Korea in 1969. He received his BS in
electronics engineering from Kyungpook
National University, Daegu, Rep. of Korea, in
1994, and his MS and PhD in electrical
engineering from the Korea Advanced Institute
of Science and Technology (KAIST), Daejeon,
Rep. of Korea, in 1997 and 2001, respectively. From 2001 to 2002, he
was a research engineer at ETRI, Daejeon, Rep. of Korea. In 2002, he
joined the Department of Information and Communication
Engineering, Hanbat National University, where he is currently a
professor. His research interests are wave scattering analysis and the
design of fractal antennas, ultra-wideband (UWB) antennas, DVB-H
antennas, and T-DMB antennas.
ETRI Journal, Volume 35, Number 2, April 2013
Jung-Nam Lee et al.
187