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

net/publication/362467585

A double-leaf-shaped four-port MIMO antenna for ultra-wideband

applications

Article in International Journal of RF and Microwave Computer-Aided Engineering · August 2022

DOI: 10.1002/mmce.23349

CITATIONS READS

3 158

4 authors, including:

Rahul Kumar Garg Sarthak Singhal

The LNM Institute of Information Technology Malaviya National Institute of Technology Jaipur
15 PUBLICATIONS 70 CITATIONS 93 PUBLICATIONS 1,205 CITATIONS

SEE PROFILE SEE PROFILE

Raghuvir Singh Tomar

The LNM Institute of Information Technology
31 PUBLICATIONS 210 CITATIONS

SEE PROFILE

All content following this page was uploaded by Rahul Kumar Garg on 07 July 2023.

The user has requested enhancement of the downloaded file.

Received: 2 March 2022 Revised: 15 July 2022 Accepted: 17 July 2022
DOI: 10.1002/mmce.23349

ORIGINAL ARTICLE

A double-leaf-shaped four-port MIMO antenna for

ultra-wideband applications

Rahul Kumar Garg1 | Sarthak Singhal2 | Maroor Vikraman Deepak Nair1 |

Raghuvir Tomar1

1
ECE Department, The LNMIIT, Jaipur,
India Abstract
2
ECE Department, MNIT, Jaipur, India A CPW-fed double-leaf-shaped ultra-wideband planar antenna, along with its
four-port MIMO configuration, is presented. The MIMO element is designed
Correspondence
through combining the techniques of a modified radiator, coplanar waveguide
Rahul Kumar Garg, ECE Department,
The LNMIIT, Jaipur, India. feeding arrangement, and defected ground structure. The size and operating
Email: [email protected] bandwidth of the MIMO element are 14
19 mm2 and 3.73 GHz to
22.01 GHz, respectively. The MIMO element has a peak realized gain of
4.27 dBi and has stable radiation patterns at lower frequencies. In the MIMO
configuration, four replicas of this MIMO element are orthogonally arranged
to ensure polarization diversity. A rotated plus-shaped decoupling structure is
integrated with the MIMO design for enhancing intra-element isolation. The
proposed MIMO design offers an ultra-wide bandwidth of 18.28 GHz
(3.73 GHz–22.01 GHz), an isolation value of more than 22 dB and an overall
footprint of 40
40 mm2. ECC < 0.03, CCL < 0.35, MEG < 4 dB, and
TARC < 20 dB have been achieved. The MIMO design is fabricated and
experimentally tested. The measured data is found to agree well with the simu-
lated data. This antenna will be highly suitable for many ultra-wideband com-
munication applications in today's world.

KEYWORDS
double-leaf-shaped antenna, MIMO parameters, plus-shaped decoupling structure,
polarization diversity, time-domain analysis, ultra-wide bandwidth

1 | INTRODUCTION The performance of UWB antenna systems gets

degraded due to multipath fading. These multipath fad-
The wireless communication technology has witnessed ing effects can be mitigated using multiple-input
significant developments during its transition from first- multiple-output (MIMO) technology.1–3 MIMO technol-
(limited to voice calls and limited data transfer) to fifth- ogy employs multiple antennas at both transmitter and
generation (HD video calling and high-speed gaming) receiver ends. However, MIMO systems impose an addi-
systems. These systems require high data rates, tional challenge involving mutual coupling (low isola-
improved reliability, low-power consumption, and tion) between adjacent antenna elements. Since this
higher channel capacity. Ultra-wideband (UWB) mutual coupling can damage the performance of the
antenna systems (with an operational bandwidth of entire system, MIMO antenna designers face a significant
3.1 GHz to 10.6 GHz allocated by FCC) can successfully challenge.4–6 Microstrip patch antennas are preferred
meet these requirements. over other antenna geometries for UWB applications

Int J RF Microw Comput Aided Eng. 2022;e23349. wileyonlinelibrary.com/journal/mmce © 2022 Wiley Periodicals LLC. 1 of 16
https://doi.org/10.1002/mmce.23349
2 of 16 GARG ET AL.

(because of advantages like wide bandwidth, improved

gain and efficiency, less complexity, ease of fabrication,
and low profile). The coplanar waveguide (CPW) feed
arrangement is one of the most suitable methods for
achieving UWB performance.7,8 In the large variety of
UWB MIMO antenna designs available in the
literature,9–45 the intra-element isolation is improved
using many different techniques, including stubs between
antenna elements, EBG (electromagnetic bandgap) struc-
tures, FSS (frequency selective surfaces), neutralization
lines, DGS (defected ground structure), slotted edge sub-
strate, and different decoupling structures. In the above-
FIGURE 1 Proposed antenna geometry (single element)
discussed geometries, either the design dimensions are
large compared to those of wireless devices, or their
bandwidth is unable to cover the future wireless commu- where λL = wavelength at 3.73 GHz). Table 1 shows the
nication services beyond 14 GHz, or the MIMO parame- optimum design dimensions.
ter values are barely acceptable. Figure 2 depicts the proposed antenna's stages of
In this work, a double-leaf-shaped microstrip growth. The evolution starts from a CPW-fed elliptical
antenna and its polarization diversity configuration monopole antenna named Antenna-I. Since the oval
for UWB applications are presented. The antenna shape near the feed line and the ground plane provides
geometry comprises a modified double-leaf shaped broader impedance matching, Antenna-I has two operat-
radiator fed using a microstrip feed line and defected ing bands (as shown in Figure 3) from 3.87 GHz to
CPW ground planes. In polarization diversity configu- 11.84 GHz (7.97 GHz bandwidth) and from 16.31 GHz to
ration, four antenna elements are arranged orthogo- 23.12 GHz (6.81 GHz bandwidth). In the second step,
nally, and a plus-shaped decoupling structure for that is, Antenna-II, the replacement of the elliptical radi-
isolation improvement is used. Both configurations ator with a modified double-leaf radiator leads to the
(single element and MIMO configuration) operate bandwidth enhancement for the first operating band
over a frequency range of 3.73 GHz to 22.01 GHz with from 7.97 GHz (3.87–11.84 GHz) to 9.55 GHz (3.94–
isolation >22 dB in MIMO configuration. The pro- 13.49 GHz) and bandwidth reduction for the second band
posed configurations are designed and analyzed using from 6.81 GHz (16.31–23.12 GHz) to 3.84 GHz (18.18–
CST-Microwave Studio. These configurations have 22.02 GHz). In the case of Antenna-III, loading CPW
advantages of broader bandwidth, miniaturized ground planes with a pair of rectangular notches further
dimensions, ECC < 0.03, CCL < 0.35, MEG < 4 dB, enhance the impedance bandwidth to 18.28 GHz (3.73–
and TARC < 20 dB over previously reported compet- 22.01 GHz) by merging both the operating bands of
itive geometries. A good agreement between simulated Antenna-II. The reason for adding the notches in the
and experimental results is seen. ground plane is that it is equivalent to adding impedance
to the antenna. This additional impedance (added by the
ground plane notches) helps in achieving the antenna
2 | S I NG LE- ELEM ENT DESIG N impedance closer to the characteristic impedance. As a
result, a broadband impedance matching is achieved that
Figure 1 shows the geometry of the proposed single enhances the overall impedance bandwidth of the
antenna element. The design is etched on an FR-4 sub- antenna.
strate (h = 1.6 mm, εr = 4.4 and tan δ = 0.025). In this The simulated normalized radiation patterns of the
context, it is worth noting that some researchers22,23 have single antenna element at a few sample frequencies
used alternate substrates due to the lossy behavior of (4 GHz, 8 GHz, 12 GHz, 16 GHz, and 20 GHz) in both
FR-4 at very high or millimeter-wave frequencies and Φ = 00 and Φ = 90 planes are shown in Figure 4. It is
some have used FR-4 due to its low-cost, good mechani- observed that the patterns are omnidirectional and bidi-
cal values and electrical insulating capabilities.24–29 The rectional in both planes for frequencies <10 GHz. At fre-
radiator's design is generated from the intersection and quencies >10 GHz, the patterns become distorted-
union of three elliptical radiators. The radiator is fed via omnidirectional in both planes. Figure 5 depicts that the
a 2.1 mm broad microstrip feed line, and a 50 Ω defected maximum value of peak realized gain in the operating
CPW feed structure. The proposed single antenna ele- band is 4.27 dBi, whereas the maximum efficiency of the
ment's total size is 14
19 mm2 (0.174 λL
0.236 λL, design is 85% at 12 GHz.
GARG ET AL. 3 of 16

TABLE 1 Optimized design dimensions of the proposed antenna (all values are in millimeters)

Parameter Value Parameter Value Parameter Value Parameter Value

Ls 19 Lg 8.05 LT 40 C1 5.6
Ws 14 Wg 5.6 WT 40 C2 7
L1 1.4 Wm 2.1 LD 22 S 1.05
W1 2.8 g 0.35 WD 0.5 θ 75
a1 5.6 a2 8.26 b1 6.3 b2 6.37

F I G U R E 2 Evolution stages of single-

element antenna

Antenna-I Antenna-II Antenna-III

0
current is concentrated inside the feedline, edges of the
ground notch, and ground plane edges closer to the feed-
Reflection Coefficient (dB)

line. At 20 GHz frequency, the surface current is mainly

-10 concentrated along the feed line and the ground plane in
the lower region of the radiator. An antenna performance
parameter named BDR (Bandwidth Dimension Ratio) is
defined as the percentage bandwidth of the antenna per
-20 unit area. The length and width of the antenna are calcu-
lated in terms of its highest wavelength.39 The obtained
Antenna-I value of BDR for the designed antenna is 3458.016, which
Antenna-II
is calculated using Equation (1). A comparison of the
Antenna-III
-30 antenna element with other geometries is presented in
3 6 9 12 15 18 21 24
Table 2. The proposed design has a percentage bandwidth
Frequency (GHz)
of 142%, impedance bandwidth of 18.28 GHz, a size reduc-
F I G U R E 3 Reflection coefficients versus frequency of tion up to 77.15%, and a BDR of 3458.016 compared with
proposed antenna for evolution stages previously reported designs.

For a detailed insight into the proposed single antenna Percentage bandwidth
BDR ¼ :
element's functionality, the surface current density distri- ðLength in terms of λL Þ
ðWidth in terms of λL Þ
butions at the frequencies of 4 GHz, 8 GHz, 12 GHz, ð1Þ
16 GHz, and 20 GHz are analyzed and are shown in
Figure 6. At 4 GHz, the surface current is distributed
inside the feedline, along the outer periphery of the 2.1 | Parametric variations
ground plane and lower half of the radiator. For 8 GHz,
the surface current is distributed uniformly over the entire The effects of various design parameters on antenna
antenna geometry. At 12 GHz, the current is spread along performance are analyzed by carrying out a parametric
the radiator's outer perimeter, the inner feedline, and the study. The chosen parameters are the dimensions of
ground plane's outer periphery. The current density is neg- the ground plane notch (L1 and W1) and the dimen-
ligible inside the ground plane and along the inner edges sions of radiator ellipses (a1, b1, a2, b2) for the analy-
of the radiator. In the case of 16 GHz, the maximum sis. For the parametric analysis, one parameter is
4 of 16 GARG ET AL.

0 0 0
0 0 0
330 30 330 30 330 30
-10 -10 -10

300 -20 60 -20 300 -20 60

300 60
-30 -30 -30

-40 -40 -40

0 Φ =00 Φ =00
270 Φ
-50=0 90 -50=900 -50=900
270 Φ 90 270 Φ 90
Φ =900
8 GHz 12 GHz
4 GHz

240 120 240 120 240 120

210 150 210 150 210 150

180 180 180

(A) (B) (C)

0 0
0 0
330 30 330 30
-10 -10

300 -20 60 -20

300 60
-30 -30

-40 -40
Φ =00 0
270 -50 90 270 Φ
-50=0 90
Φ =900
Φ =900
16 GHz
20 GHz

240 120 240 120

210 150 210 150

180 180

(D) (E)

FIGURE 4 Radiation pattern of single element antenna (simulated): (A) at 4 GHz, (B) at 8 GHz, (C) at 12 GHz, (D) at 16 GHz, and
(E) at 20 GHz

5 90
Figure 7A illustrates that as the width of the ground
80
4 plane notch (W1) is increased, the lower cutoff shifts to a
70 higher frequency and the upper cutoff shifts to a lower
Radiation Efficiency (%)

3
Realized Gain (dB)

60 frequency. The ground plane notch length (L1) variation

2 50 affects the impedance matching of the antenna, as shown
in Figure 7B. The lower and upper cutoff frequencies
1 40
shift towards the higher side due to variation in the
30 length of defects. The chosen optimized dimensions of
0
20 the defects are W1 = 2.8 mm and L1 = 1.4 mm for the
-1
10 best possible scenario.
Gain
Efficiency In both cases, when the dimensions a1 and b1
-2 0
3 6 9 12 15 18 21 24 (Ellipse-1) increase, the lower and upper cutoff fre-
Frequency (GHz) quencies shift to the higher values. The effect on the
impedance matching is opposite of each other. Match-
F I G U R E 5 Simulated peak realized gain and efficiency of
ing improved in the case of dimension a1 variations,
single element antenna
but it deteriorates in the case of b1. The chosen opti-
mized values are a1 = 5.6 mm, and b1 = 6.3 mm. In
varied at a time, and others are kept fixed. The results the dimensional variation of Ellipse-2 and Ellipse-3,
obtained after these parametric variations are shown (both are identical); the increment in dimension a2
in Figure 7. lowered the cutoff frequencies and deteriorated the
GARG ET AL. 5 of 16

F I G U R E 6 Surface current
4 GHz 8 GHz 12 GHz
distribution: (A) at 4 GHz, (B) at
8 GHz, (C) at 12 GHz, (D) at
16 GHz, and (E) at 20 GHz

(A) (B) (C)

16 GHz 20 GHz

(D) (E)

TABLE 2 Single antenna element comparison

Dimension Bandwidth

S. no. Ref. mm
mm λL
λL GHz % Ratio Size reduction (% λL2) BDR

1. 17 20
25 0.20
0.25 3–11 114 3.67 17.87 2280
2. 18 32
32 0.33
0.33 3.1–10.6 109 3.42 62.29 1000.918
3. 19 25
35 0.25
0.35 3–11 114 3.67 53.06 1302.857
4. 20 39
15 0.39
0.15 3–11 114 3.67 29.8 1948.717
5. 22 30
32 0.31
0.33 3.1–14 154 4.51 59.85 1505.376
6. 23 40
40 0.424
0.424 3.18–11.5 113 3.62 77.15 628.559
7. Here 14
19 0.174
0.236 3.73–22.01 142 5.9 -- 3458.016

impedance matching. The opposite phenomenon 3.1 | Simulated results

is observed when we increase dimension b2. The cho-
sen optimum values are a2 = 8.26 mm, and b2 Figure 9 depicts the scattering parameters of the pro-
= 6.37 mm. posed MIMO antenna, both with and without the
decoupler. The MIMO configuration's impedance
bandwidth is found to be equal to that of the antenna
3 | F OUR-POR T MIMO A NTENNA element. In the absence of the decoupling structure,
DESIGN isolation greater than 20 dB (|S21| and |S41|) between
adjacent antenna elements is observed. The isolation
The proposed four-port MIMO antenna configuration is between the diagonal components, on the other hand,
shown in Figure 8. Four identical antenna elements are is (jS31 j ffi 15 dB) relatively poor compared to other isola-
positioned orthogonally to achieve polarization diversity. tion values. Due to the small footprint of the MIMO
The overall physical size of the MIMO configuration is antenna design, the performance degradation between
40
40 mm2. The arm length of the plus-shaped decou- the diagonal antenna elements in terms of strong mutual
pling structure and the angle (θ) value are optimized coupling is seen. It becomes necessary to nullify the
using iterative simulations. The optimized values are mutual coupling effects to maintain the antenna
listed in Table 1. performance in MIMO configuration. A plus-shaped
6 of 16 GARG ET AL.

0 0
W1=2.1 mm W1=2.45 mm W1=2.8 mm L1=0.7 mm L1=1.05 mm L1=1.4 mm
W1=3.15 mm W1=3.5 mm L1=1.75 mm L1=2.1 mm
-5
-10
Reflection Coefficient (dB)

Reflection Coefficient (dB)

-10

-20
-15

-20
-30

-25

-40
-30

-50 -35
3 6 9 12 15 18 21 24 3 6 9 12 15 18 21 24
Fre quenc y (GHz) Fre quenc y (GHz )
(A) (B)
0 0
a1=4.55 mm a1=4.9 mm a1=5.25 mm b1=6.02 mm b1=6.37 mm b1=6.72 mm
a1=5.6 mm a1=5.95 mm b1=7.07 mm b1=7.42 mm

-10
-10
Reflection Coefficient (dB)

Reflection Coefficient (dB) -20

-20

-30

-30
-40

-40 -50

3 6 9 12 15 18 21 24 3 6 9 12 15 18 21 24
Frequency (GHz) Frequency (GHz)
(C) (D)
0 0
a2=7.56 mm a2=7.91 mm a2=8.26 mm b2=5.67 mm b2=6.02 mm b2=6.37 mm
a2=8.61 mm a2=8.96 mm b2=6.72 mm b2=7.07 mm
Reflection Coefficient (dB)

Reflection Coefficient (dB)

-10 -10

-20 -20

-30 -30

3 6 9 12 15 18 21 24 3 6 9 12 15 18 21 24
Frequency (GHz) Frequency (GHz)
(E) (F)

F I G U R E 7 Parametric variation of the antenna parameters (A) variation in W1, (B) variation in L1, (C) variation in a1, (D) variation in
b1, (E) variation in a2, and (F) variation in b2

decoupling structure has been placed between antenna parasitic element (plus shape decoupler) between
elements to improve the isolation (that is, to minimize antenna elements ensured the diversion in the surface
the mutual coupling). Iterative simulations were used to current distribution (in this case, absorbing the portion of
optimize the decoupler's location, size, and tilt angle. The the coupled current between antenna elements). This
GARG ET AL. 7 of 16

resulted in the enhancement of the isolation of MIMO Figure 10C–F. It is observed that the impedance
configuration. As a result of the parasitic element bandwidth of the elements reduces from 18.28 GHz
approach of isolation improvement (the presence of a (3.73 GHz to 22.01 GHz) to 12.1 GHz (7.9 GHz to
decoupling structure), it is observed that the intra-port 20 GHz) for connected ground planes. The mutual
isolation value between the diagonal elements is coupling between the elements is below the accept-
improved to more than 22 dB without any significant able limit of 15 dB. This degradation in the imped-
effects on the impedance bandwidth and isolation ance bandwidth of antenna elements may be
between the adjacent elements. attributed to the mutual coupling effects between the
radiator and ground connecting line. It is also
observed that there is a significant change in scatter-
3.2 | Effect of connected ground on the ing parameters with variations in the ground connect-
MIMO design ing line width.

The coplanar ground planes of the adjacent elements

of the proposed MIMO antenna are connected for 3.3 | Effect of decoupler rotation
both with and without a decoupler (as shown in
Figure 10A–B) via metallic lines to investigate the This section shows the effect of rotating the decoupler
effect of connected ground planes. The width of con- (at various possible angles) on the performance of the
necting line (CGW) is varied from 0.2 mm to 0.3 mm. MIMO design. The decoupler is symmetric to its shape
The comparison of the proposed MIMO antenna with and has equal dimensions, and the angle of the decoupler
discrete ground planes (NCG) and connected ground significantly affects the antenna performance. The angle
planes in terms of scattering parameters is shown in is measured in an anti-clockwise direction from the
antenna axis (presented in Figure 8B). The simulation
results of the MIMO design (in terms of obtained S-
parameters) at various angles are presented in Figure 11.
The decoupler is rotated for the 50 ≤ θ ≤ 90 , and the
results in terms of the S-parameters are observed. The
angle <50 is omitted to protect the overlapping of the
decoupler and other antenna elements. An unwanted
band-notch region between 3.5 GHz to 5 GHz and a shift
in the upper cutoff frequency of the design are observed
as a result of decoupler rotation. It is also observed that
the decoupler's angle severely impacts the isolation
values between the antenna elements. In the worst sce-
F I G U R E 8 MIMO configuration of the proposed antenna: nario, the isolation between the antenna elements is
(A) without decoupler and (B) with plus-shaped decoupler dropped to 15 dB. By carefully examining all the values,

0 -20
S11/22/33/44- Without Decoupler S13/31- Without Decoupler
S11/22/33/44- With Decoupler S13/31- With Decoupler

-10
-30
S-Parameters (dB)

S-Parameters (dB)

-20
-40
-30

-50
-40
S12/21 -Without Decoupler S14/41- Without Decoupler
S12/21 -With Decoupler S14/41- With Decoupler
-50 -60
3 6 9 12 15 18 21 24 3 6 9 12 15 18 21 24
Frequency (GHz) Frequency (GHz)
(A) (B)

FIGURE 9 S-parameters (simulated) of the design with and without the plus-shaped decoupler
8 of 16 GARG ET AL.

(A) (B)
0 0

-10

-10
-20
S11 (dB)

S21 (dB)
-30
-20

-40

-30
-50
NCG - with decoupler NCG - with decoupler
NCG - without decoupler NCG - without decoupler
CGW = 0.2 mm -60 CGW = 0.2 mm
CGW = 0.3 mm CGW = 0.3 mm
-40
3 6 9 12 15 18 21 24 3 6 9 12 15 18 21 24
Frequency (GHz) Frequency (GHz)
(C) (D)
0 0

-10
-10

-20

-20
S31 (dB)

S41 (dB)

-30

-40
-30

-50
-40 NCG - with decoupler NCG - with decoupler
NCG - without decoupler -60 NCG - without decoupler
CGW = 0.2 mm CGW = 0.2 mm
CGW = 0.3 mm CGW = 0.3 mm
-50 -70
3 6 9 12 15 18 21 24 3 6 9 12 15 18 21 24
Frequency (GHz) Frequency (GHz)
(E) (F)

F I G U R E 1 0 (A) Connected ground planes of MIMO antenna without decoupler, (B) Connected ground planes of MIMO antenna with
decoupler, and (C) to (F) Effect of connecting the ground planes on S-parameters (where CGW = connected ground width; NCG = not
connected ground)

the optimum value of the rotation angle is chosen to be at a few sample frequencies of 4 GHz, 8 GHz, 12 GHz,
θ = 75 , for the best possible S-parameters. 16 GHz, and 20 GHz. It is observed that in the absence of
Surface current distributions are analyzed to observe decoupling structure, the current is being coupled to
the effects of the decoupler by exciting one of the ports of other antenna elements signifying the high mutual cou-
the MIMO antenna (as all the antenna elements are iden- pling at all frequencies. In the presence of a decoupler,
tical), and the remaining ports terminated with 50 Ω the maximum current is concentrated at the decoupler,
load. Figure 12 displays the surface current distributions thereby isolating the other terminated antenna elements.
GARG ET AL. 9 of 16

-20

-10 -30

S21 (dB)
S11 (dB)

-40

-20

-50
θ=90 0 θ=85 0 θ=80 0 θ=90 0 θ=85 0 θ=80 0
θ=75 0 θ=70 0 θ=65 0 θ=75 0 θ=70 0 θ=65 0
θ=60 0 θ=55 0 θ=50 0 θ=60 0 θ=55 0 θ=50 0
-30 -60
3 6 9 12 15 18 21 24 3 6 9 12 15 18 21 24

Frequency (GHz) Frequency (GHz)

(A) (B)

-20 -20

-30 -30
S31 (dB)

S41 (dB)

-40 -40

-50 -50
θ=90 0 θ=85 0 θ=80 0 θ=90 0 θ=85 0 θ=80 0
θ=75 0 θ=70 0 θ=65 0 θ=75 0 θ=70 0 θ=65 0
θ=60 0 θ=55 0 θ=50 0 θ=60 0 θ=55 0 θ=50 0
-60 -60
3 6 9 12 15 18 21 24 3 6 9 12 15 18 21 24

Frequency (GHz) Frequency (GHz)

(C) (D)

FIGURE 11 Effect of decoupler rotations on the S-parameters of the design

3.4 | Measured results patterns in both ɸ = 00 and ɸ = 90 planes at various fre-
quencies are shown in Figure 15. At frequencies
The proposed four-port MIMO antenna prototype with <10 GHz, radiation patterns are omnidirectional and
optimized design dimensions is fabricated and tested. bidirectional in ɸ = 0 and ɸ = 90 planes, respectively.
Figure 13 shows images of the fabricated prototype and At frequencies >10 GHz, distorted-omnidirectional pat-
the measurement setup in an anechoic chamber. All terns in both planes are observed.
ports of the proposed antenna design are symmetric and
identical. The comparison of simulated and measured
scattering parameters is shown in Figure 14. It is 4 | MIMO PERFORMANCE
observed that the measured impedance bandwidth is PARAMETERS
18.19 GHz (3.64–21.79 GHz), and port isolation values
are better than 20 dB, in comparison with the simulated The performance evaluation parameters for a MIMO
bandwidth of 18.28 GHz and isolation of 22 dB, respec- antenna are ECC (envelope correlation coefficient), DG
tively. Minor discrepancies between the simulated and (diversity gain), CCL (channel capacity loss), MEG (mean
measured results can be attributed mainly to substrate effective gain), and TARC (total active reflection coeffi-
losses, soldering, fabrication, and human errors. cient).1 ECC represents the correlation of radiation pat-
In far-field radiation pattern measurement, one port terns of one antenna with those of adjacent antenna
was connected to the excitation signal, and the other elements. It is also a measure of communication channel
ports were matched with 50 Ω load. The radiation isolation when they transmit simultaneously. The lower
10 of 16 GARG ET AL.

4 GHz

(A)

8 GHz

F I G U R E 1 3 Photograph: (A) fabricated antenna,

(B) measurement setup in an anechoic chamber

(B) 0

12 GHz -10
S-Parameters (dB)
-20

-30

-40

(C) -50
S11/22/33/44_Sim S11/22/33/44_Meas
S21_Sim S21_Meas
S31_Sim S31_Meas
S41_Sim S41_Meas
-60
16 GHz 3 6 9 12 15 18 21 24
Frequency (GHz)

F I G U R E 1 4 S-parameters of the proposed MIMO antenna

(simulated and measured)

radiation pattern (using Equation (2)) or S-parameters

(D) (using Equation (3)). In both cases, the peak ECC value
is 0.03. The simulated and experimental values for the
20 GHz proposed design are demonstrated in Figure 16A.
Ð Ð !  2
  !  
 4π  F i ðθ, ϕÞ  F j ðθ, ϕÞdΩ
ECCij ¼ Ð Ð ! 2 Ð Ð ! 2 , ð2Þ
   
4π  F i ð θ, ϕ Þ  dΩ 4π  F j ð θ, ϕÞ  dΩ

 2
(E)   
Sii Sij þ Sji Sjj 
ECCij ¼   2    2  2  , ð3Þ
 1  jSii j  Sji  : 1  Sjj   Sij  
2
F I G U R E 1 2 Surface current distribution with and without
decoupling structure: (A) at 4 GHz, (B) at 8 GHz, (C) at 12 GHz,
(D) at 16 GHz, and (E) at 20 GHz
where Ω = solid angle, i and j are antennas for
correlation.
the ECC value, the lesser the correlation between anten- Another important MIMO parameter is the DG,
nas. The upper threshold value of ECC for any MIMO which represents the effectiveness of diversity. It is
system is 0.4. ECC can be calculated by using either achieved in a multichannel wireless environment and
GARG ET AL. 11 of 16

0 0 0
0 0 0
330 30 330 30 330 30

-10 -10 -10

300 60 300 60 300 60

-20 -20 -20

-30 -30 -30

4 GHz 8 GHz 12 GHz
270 -40 90 270 -40 90 270 -40 90
Φ =00_Sim Φ = 00_Meas Φ =00_Sim Φ =00_Meas Φ =00_Sim Φ =00_Meas
Φ = 900_Sim Φ = 900_Meas Φ =900_Sim Φ =900_Meas Φ =900_Sim Φ =900_Meas

240 120 240 120 240 120

210 150 210 150 210 150

180 180 180

(A) (B) (C)

0 0
0 0
330 30 330 30

-10 -10

300 60 300 60
-20 -20

-30 -30
16 GHz 18 GHz
270 -40 90 270 -40 90
Φ =00_Sim Φ =00_Meas Φ =00_Sim Φ =00_Meas
Φ =900_Sim Φ =900_Meas Φ =900_Sim Φ =900_Meas

240 120 240 120

210 150 210 150

180 180

(D) (E)

F I G U R E 1 5 Simulated and measured radiation patterns of designed MIMO antenna: (A) at 4 GHz, (B) at 8 GHz, (C) at 12 GHz, (D) at
16 GHz, and (E) at 18 GHz

quantifies the diversity effects on the communication sys- MEG is the measure of antenna performance when
tem. The DG values calculated using Equation (4), shown the effect of the environment is also taken into consider-
in Figure 16B, are approximately constant at 10 for both ation and can be calculated by using Equations (6) and
simulated and experimental data. (7) based on scattering parameters. The observed values
of MEG for this antenna are within the limits of 6 dB
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2 to 4 dB.
DGij ¼ 10 1  ECCij  , ð4Þ  X K  
MEGi ¼ 0:5 1  Sij  , ð6Þj¼1

CCL is the estimated maximum transmission rate for MEGij ¼ MEGi  MEGj , ð7Þ
reliable transmission in a rich multipath channel. The
allowable CCL for a MIMO antenna is 0.4 bits/s/Hz. The where, K = number of antenna elements in MIMO
maximum value of CCL for the proposed antenna com- configuration.
puted, using Equation (5), illustrated in Figure 16C, The square root of total reflected power divided by
is 0.35. total incident power is TARC and can be calculated using
Equation (8). It depicts the degree of interference
 
σ 11 σ 12  between the antenna elements. The TARC of the four-
CCL ¼  log 2    , ð5Þ
σ 21 σ 22  port MIMO antenna can also be calculated from the S-
  2    parameters. The TARC value is <20 dB for this antenna
where, σ ii ¼ 1  jSii j2  Sij  , σ ij ¼  Sii Sij þ Sji Sjj . for the entire operating band.
12 of 16 GARG ET AL.

0.04 10.00
ECC(1,2)_Sim ECC(1,3)_Sim ECC(1,4)_Sim
ECC(1,2)_Meas ECC(1,3)_Meas ECC(1,4)_Meas
ECC(1,2)-FarField ECC(1,3)-FarField ECC(1,4)-FarField

0.03
ECC

Diversity Gain
0.02 9.95

0.01

DG(1,2)_Sim DG(1,3)_Sim DG(1,4)_Sim

DG(1,2)_Meas DG(1,3)_Meas DG(1,4)_Meas
0.00
9.90
3 6 9 12 15 18 21 24 3 6 9 12 15 18 21 24

Frequency (GHz) Frequency (GHz)

(A) (B)
0.5

0 CCL(1,2)_Sim CCL(1,3)_Sim CCL(1,4)_Sim

CCL(1,2)_Meas CCL(1,3)_Meas CCL(1,4)_Meas
0.4
Mean Effective Gain (dB)

-2
CCL (Bits/Sec/Hz) 0.3

-4
0.2

-6
MEG1/2/3/4_Sim MEG1/2/3/4_Meas 0.1
MEG12_Sim MEG12_Meas
MEG13_Sim MEG13_Meas
-8
0.0
3 6 9 12 15 18 21 24
3 6 9 12 15 18 21 24
Frequency (GHz) Frequency (GHz)

(C) (D)
0

TARC(1,2)_Sim TARC(1,3)_Sim TARC(1,4)_Sim

TARC(1,2)_Meas TARC(1,3)_Meas TARC(1,4)_Meas

-20
TARC (dB)

-40

-60

3 6 9 12 15 18 21 24
Frequency (GHz)
(E)

FIGURE 16 MIMO performance parameters: (A) ECC, (B) DG, (C) MEG, (D) CCL, and (E) TARC of proposed MIMO antenna
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
   
 Sii þ Sij ejθ 2 þ  Sji þ Sjj ejθ 2 based on the physical size of MIMO configuration,
TARCij ¼ : ð8Þ impedance bandwidth, percentage bandwidth, and
2
obtained MIMO parameters, are listed in Table 3. It is
The performance comparison of the designed MIMO observed that the proposed design has the highest imped-
antenna with the performance of the available antennas, ance bandwidth of 18.28 GHz (~142%), comparable or
GARG ET AL. 13 of 16

TABLE 3 Comparison of the proposed MIMO antenna with existing literature

Size Size reduction Connected Isolation Peak Minimum Peak

Ref. B.W. (GHz) (mm
mm) (% λL2) Ports ground (dB) ECC DG CCL
14 3–10.9 50
30 NC 2 Yes 20 0.06 9.6 a

24 2.5–14.5 50
30 NC 2 Yes 20 0.04 7.4 a

25 2.8–20 18
36 NC 2 Yes 20 0.02 9.95 a

28 1.51–13.25 50
35 NC 2 Yes 21.7 0.059 9.9 0.35

29 3–10.7 43
34.9 NC 2 Yes 25 0.05 9.8 a

30 3–11 50
25 NC 2 Yes 20 0.005 a a

33 3.1–10.6 35
40 NC 2 Yes 16 0.01 9.95 a

34 3.1–5 35
33 NC 2 Yes 22 a a a

35 3–11 21
27 NC 2 Yes 15 0.05 a a

23 3.18–11.5 80
80 65.65 4 No 15 0.015 9.94 0.5

26 3.01–12.5 81
87 65.18 4 No 20 0.2 a a

27 3.4–3.8 60
60 46.58 4 No 19 0.12 a a

31 2.8–13.3 72
72 45.30 4 Yes 18 0.06 9.95 a

32 3.1–10.6 70
41 19.23 4 No 17 0.12 9.95 0.4

Here 3.73–22.01 40
40 - 4 No 22 0.03 9.97 0.35

Abbreviation: NC, not compared.

a
Values not available.

smaller footprint, isolation of >22 dB, ECC < 0.03, Group delay is almost stable and less than 2.5 ns for the
CCL < 0.35 and DG > 9.97 as compared to other similar operating bandwidth. The correlation between the trans-
structures. mitted and received pulses is signified by the fidelity fac-
tor and can be calculated using Equation (9). The
calculated values of the fidelity factor are 83.9% for the
5 | TIM E-DOM AIN A NALY SIS F2F configuration and 78% for the S2S configuration.
2 Z ∞ 3
The time-domain analysis of the UWB antenna is neces-
6 St ðt ÞSr ðt þ τÞdτ 7
sary for system applications. Two similar antennas are
Fidelity factor ¼ max 6
4Z ∞
∞ Z ∞ 7, ð9Þ
5
placed at a distance enough to ensure the far-field sce- 2 2
jSt ðt Þj dt jSr ðt Þj dt
nario (15 cm in this case) in the time-domain analysis.44 ∞ ∞
One antenna works as a transmitter and the other as a
receiver. A short Gaussian pulse excites the transmitter where St ðt Þ ¼ transmitted pulse, Sr ðt Þ ¼ received pulse,
antenna, received at the other end, and the obtained and τ ¼ group delay.
results are analyzed. The observed results in the time-
domain analysis are the nature of transmitted and
received pulses, transmission coefficient, isolation phase, 6 | CONCLUSIONS
group delay, and fidelity factor. Here, antennas are
arranged in two different configurations named face-to- A double-leaf-shaped monopole antenna for ultra-
face (F2F) and side-by-side (S2S) for the time-domain wideband applications is reported. In addition, its
analysis. The configuration setups and the obtained four-port MIMO configuration is also investigated.
results are presented in Figure 17. The sizes of an antenna element and its MIMO con-
The value of the transmission coefficient in both con- figuration are 14
19 mm2 and 40
40 mm2, respec-
figurations is better than the 40 dB throughout the oper- tively. The techniques of a modified radiator, coplanar
ating frequency band. The linear variation of the waveguide feeding, and defected ground structure are
isolation phase describes the absence of the out-of-phase combined to achieve ultra-wide bandwidth of
components. In both cases, the isolation phase varies lin- 18.28 GHz (3.73–22.01 GHz). A plus-shaped decou-
early from 180 to +180 for the entire bandwidth. pling structure between the antenna elements is
14 of 16 GARG ET AL.

(A) (B)
1.0 -40
Input Signal
F2F
S2S

Transmission Coefficient (dB)

0.5 -50
Normalized Amplitude

0.0 -60

-0.5 -70
S2S
F2F

-80
-1.0
0.0 0.5 1.0 1.5 2. 0 3 6 9 12 15 18 21 24
Time (ns ) Frequency (GHz)

(C) (D)
3
S2S
S2S
180 F2F
F2F
Isolation Phase (degree)

Group Delay (nS)

90 2

-90

-180 0
3 6 9 12 15 18 21 24 3 6 9 12 15 18 21 24
Frequency (GHz) Frequency (GHz)

(E) (F)

F I G U R E 1 7 Time-domain analysis: (A) F2F configuration, (B) S2S configuration, (C) normalized transmitted and received pulses,
(D) transmission coefficient, (E) isolation phase, and (F) group delay

integrated to improve the isolation to more than the proposed design with similar designs in the litera-
22 dB in the operating bandwidth. The MIMO perfor- ture shows either comparable or better bandwidth,
mance parameters are within the desirable limits. The compact physical size, and acceptable MIMO perfor-
prototype of the MIMO configuration is fabricated mance parameters. The proposed design is suitable for
and experimentally tested. The measured experimental modern ultra-wideband communication and diversity
results agree well with the expected result. Comparing applications.
GARG ET AL. 15 of 16

A C K N O WL E D G M E N T 14. Wang E, Wang W, Tan X, Wu Y, Gao J, Liu Y. A UWB MIMO

The authors acknowledge the help and measurement lab- slot antenna using defected ground structures for high isola-
oratory facility support provided by the ECE Department, tion. Int J RF Microw Comput Aided Eng. 2020;30(5):e22155.
15. Rao PK, Mishra R. Elliptical shape flexible MIMO antenna
GWEC Ajmer, India, and thank the organization.
with high isolation for breast cancer detection application.
IETE J Res. 2020;1-9.
CONFLICT OF INTEREST 16. Tang Z, Wu X, Zhan J, Hu S, Xi Z, Liu Y. Compact UWB-
None of the authors have a conflict of interest to disclose. MIMO antenna with high isolation and triple band-notched
characteristics. IEEE Access. 2019;7:19856-19865.
DATA AVAILABILITY STATEMENT 17. Medkour H, Lakrit S, Das S, Madhav BTP, VasuBabu K. A
Data sharing is not applicable to this article as no data- compact printed UWB MIMO antenna with electronically
reconfigurable WLAN band-notched characteristics. J Circuits
sets were generated or analysed during the current study.
Syst Comput. 2021;31:2250045.
18. Kumar N, Kiran KU. Meander-line electromagnetic bandgap
ORCID structure for UWB MIMO antenna mutual coupling reduction
Sarthak Singhal https://orcid.org/0000-0002-1160-6150 in E-plane. AEU-Int J Electron Commun. 2020;127:153423.
19. Wang L, Du Z, Yang H, et al. Compact UWB MIMO antenna
R EF E RE N C E S with high isolation using fence-type decoupling structure. IEEE
1. Sharawi MS. Printed MIMO Antenna Engineering. Artech Antennas Wirel Propag Lett. 2019;18(8):1641-1645.
House; 2014. 20. Luo S, Wang D, Chen Y, Li E, Jiang C. A compact dual-port
2. Chen Z, Zhou W, Hong J. A miniaturized MIMO antenna with UWB-MIMO antenna with quadruple band-notched character-
triple band-notched characteristics for UWB applications. IEEE istics. AEU-Int J Electron Commun. 2021;136:153770.
Access. 2021;9:63646-63655. 21. Khan MK, Feng Q, Zheng Z. Experimental investigation and
3. Kshetrimayum RS. An introduction to UWB communication design of UWB MIMO antenna with enhanced isolation. Prog
systems. IEEE Potentials. 2009;28(2):9-13. Electromagn Res C. 2021;107:287-297.
4. Alibakhshikenari M, Babaeian F, Virdee BS, et al. A comprehen- 22. Li Y, Li W, Yu W. A multi-band/UWB MIMO/diversity
sive survey on “Various decoupling mechanisms with focus on antenna with an enhanced isolation using radial stub-loaded
metamaterial and metasurface principles applicable to SAR and resonator. Appl Comput Electromagn Soc J. 2013;8-20.
MIMO antenna systems”. IEEE Access. 2020;8:192965-193004. 23. Hasan MN, Chu S, Bashir S. A DGS monopole antenna loaded
5. Zhang S, Lau BK, Sunesson A, He S. Closely-packed UWB with U-shape stub for UWB MIMO applications. Microw Opt
MIMO/diversity antenna with different patterns and polariza- Technol Lett. 2019;61(9):2141-2149.
tions for USB dongle applications. IEEE Trans Antennas Pro- 24. Iqbal A, Saraereh OA, Ahmad AW, Bashir S. Mutual coupling
pag. 2012;60(9):4372-4380. reduction using F-shaped stubs in UWB-MIMO antenna. IEEE
6. Khattak MI, Khan MI, Anab M, Ullah A, Al-Hasan M, Access. 2017;6:2755-2759.
Nebhen J. Miniaturized CPW-fed UWB-MIMO antennas with 25. Chandel R, Gautam AK, Rambabu K. Design and packaging of
decoupling stub and enhanced isolation. Int J Microw Wirel an eye-shaped multiple-input–multiple-output antenna with
Technol. 2021;14:1-9. high isolation for wireless UWB applications. IEEE Trans Com-
7. Kaiser T, Zheng F, Dimitrov E. An overview of ultra-wide-band pon Packag Manuf Technol. 2018;8(4):635-642.
systems with MIMO. Proc IEEE. 2009;97(2):285-312. 26. Srivastava K, Kumar A, Kanaujia BK, Dwari S, Kumar S. A
8. Deng JY, Guo LX, Liu XL. An ultrawideband MIMO antenna CPW-fed UWB MIMO antenna with integrated GSM band and
with a high isolation. IEEE Antennas Wirel Propag Lett. 2015; dual-band notches. Int J RF Microw Comput Aided Eng. 2019;
15:182-185. 29(1):e21433.
9. Sarkar C, Rao D, Saha C, Siddiqui JY. Ultra-wideband MIMO 27. Saxena S, Kanaujia BK, Dwari S, Kumar S, Choi HC, Kim KW.
monopole antenna with WLAN band rejection. IETE J Res. Planar four-port dual circularly-polarized MIMO antenna for
2021;1-8. sub-6 GHz band. IEEE Access. 2020;8:90779-90791.
10. Bahmanzadeh F, Mohajeri F. Simulation and fabrication of a 28. Dey AB, Pattanayak SS, Mitra D, Arif W. Investigation
high-isolation very compact MIMO antenna for ultra-wideband and design of enhanced decoupled UWB MIMO antenna
applications with dual band-notched characteristics. AEU-Int J for wearable applications. Microw Opt Technol Lett. 2021;
Electron Commun. 2021;128:153505. 63(3):845-861.
11. Khan AA, Naqvi SA, Khan MS, Ijaz B. Quad-port miniaturized 29. Modak S, Khan T. A slotted UWB-MIMO antenna with qua-
MIMO antenna for UWB 11 GHz and 13 GHz frequency bands. druple band-notch characteristics using mushroom EBG struc-
AEU-Int J Electron Commun. 2021;131:153618. ture. AEU-Int J Electron Commun. 2021;134:153673.
12. Sharma MK, Kumar M, Saini JP. Design and analysis of a com- 30. Dadel M, Parveen Z, Srivastava S. Compact UWB MIMO
pact UWB-MIMO antenna with improved isolation for UWB/- antennas with high isolation. IETE J Res. 2021;1-7.
WLAN applications. Wirel Pers Commun. 2021;119(4):2913- 31. Kumar S, Lee GH, Kim DH, Mohyuddin W, Choi HC,
2928. Kim KW. Multiple-input-multiple-output/diversity antenna
13. Ramanujam P, Venkatesan PR, Arumugam C, Ponnusamy M. with dual band-notched characteristics for ultra-wideband
Design of miniaturized super wideband printed monopole applications. Microw Opt Technol Lett. 2020;62(1):336-345.
antenna operating from 0.7 to 18.5 GHz. AEU-Int J Electron 32. Yang L, Xu M, Li C. Four-element MIMO antenna system for
Commun. 2020;123:153273. UWB applications. Radioengineering. 2019;28(1):60-67.
16 of 16 GARG ET AL.

33. Zhang S, Ying Z, Xiong J, He S. Ultrawideband MIMO/diversity 41. Zhao L, Liu F, Shen X, Jing G, Cai YM, Li Y. A high-pass
antennas with a tree-like structure to enhance wideband isola- antenna interference cancellation chip for mutual coupling
tion. IEEE Antennas Wirel Propag Lett. 2009;8:1279-1282. reduction of antennas in contiguous frequency bands. IEEE
34. Zhang S, Pedersen GF. Mutual coupling reduction for UWB Access. 2018;6:38097-38105.
MIMO antennas with a wideband neutralization line. IEEE 42. Jiang T, Jiao T, Li Y. A low mutual coupling MIMO antenna
Antennas Wirel Propag Lett. 2015;15:166-169. using periodic multi-layered electromagnetic band gap struc-
35. Wu L, Cao X, Yang B. Design and analysis of a compact UWB- tures. Appl Comput Electromagn Soc J. 2018;305-311.
MIMO antenna with four notched bands. Prog Electromagn Res 43. Liu F, Guo J, Zhao L, Huang GL, Li Y, Yin Y. Ceramic
M. 2022;108:127-137. superstrate-based decoupling method for two closely packed
36. Raheja DK, Kanaujia BK, Kumar S. Compact four-port MIMO antennas with cross-polarization suppression. IEEE Trans
antenna on slotted-edge substrate with dual-band rejection Antennas Propag. 2020;69(3):1751-1756.
characteristics. Int J RF Microw Comput Aided Eng. 2019;29(7): 44. Addepalli T, Desai A, Elfergani I, et al. 8-port semi-circular arc
e21756. MIMO antenna with an inverted L-strip loaded connected
37. Chandel R, Gautam AK, Rambabu K. Tapered fed compact ground for UWB applications. Electronics. 2021;10(12):1476.
UWB MIMO-diversity antenna with dual band-notched 45. Sharawi MS. Current misuses and future prospects for printed
characteristics. IEEE Trans Antennas Propag. 2018;66(4): multiple-input, multiple-output antenna systems [wireless cor-
1677-1684. ner]. IEEE Antennas Propag Mag. 2017;59(2):162-170.
38. Bilal M, Saleem R, Abbasi HH, Shafique MF, Brown AK. An
FSS-based nonplanar quad-element UWB-MIMO antenna sys-
tem. IEEE Antennas Wirel Propag Lett. 2016;16:987-990. How to cite this article: Garg RK, Singhal S,
39. Chen KR, Row JS. A compact monopole antenna for super
Nair MVD, Tomar R. A double-leaf-shaped four-
wideband applications. IEEE Antennas Wirel Propag Lett. 2011;
10:488-491.
port MIMO antenna for ultra-wideband
40. Li J, Zhang X, Wang Z, et al. Dual-band eight-antenna array applications. Int J RF Microw Comput Aided Eng.
design for MIMO applications in 5G mobile terminals. IEEE 2022;e23349. doi:10.1002/mmce.23349
Access. 2019;7:71636-71644.

View publication stats