Wiley

International Journal of Antennas and Propagation

Volume 2024, Article ID 7480655, 18 pages
https://doi.org/10.1155/2024/7480655

Research Article
A Novel and Compact Metamaterial-Based Four-Element MIMO
Antenna System for Millimeter-Wave Wireless Applications with
Enhanced Isolation

Iftikhar Ud Din ,1 Nisar Ahmad Abbasi ,2 Waheed Ullah ,3 Sadiq Ullah ,1

Messaoud Ahmad Ouameur ,4 and Dushantha Nalin K. Jayakody 5,6
1
Telecommunication Engineering Department, University of Engineering and Technology, Mardan 23200, Pakistan
2
Department of Electrical Engineering, University of Hafr Al Batin, Hafr Al Batin 39524, Saudi Arabia
3
School of Electrical and Information Engineering, University of the Witwatersrand, Johannesburg, South Africa
4
Department of Electrical Engineering and Computer Engineering, University du Quebec a Trois-Rivieres, Quebec, Canada
5
Centro de Investigação em Tecnologias – Autonoma TechLab, Universidade Autónoma de Lisboa, 1150-293 Lisbon, Portugal
6
CIET, Faculty of Engineering, Sri Lanka Institute of Information Technology, Malabe 10115, Sri Lanka

Correspondence should be addressed to Dushantha Nalin K. Jayakody; [email protected]

Received 8 March 2024; Revised 18 April 2024; Accepted 23 April 2024

Academic Editor: Sandeep Kumar Palaniswamy

Copyright © 2024 Iftikhar Ud Din et al. Tis is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.
A compact fork-shaped MIMO antenna system with a 2 × 2 arrangement with four elements is presented. Te MIMO elements are
arranged orthogonally to achieve a small overall size of 36 × 28 mm2 and a wide bandwidth for 5G mm-wave applications. MIMO
elements are positioned 4 mm from each corner of the substrate to achieve compact size and minimize coupling. To improve the
isolation of the proposed MIMO system, a metamaterial slab is inserted in the middle of the substrate and between radiating
elements of the MIMO antenna system, which improves isolation by 10 dB within the whole operating band and achieves
maximum isolation of 65 dB at 34.5 GHz. Te proposed MIMO system operates in the Ka-band frequency range of 22–50 GHz
with isolation greater than 30 dB and efciency above 80% across the entire frequency spectrum for 5G communication.
Additionally, the performance parameters of MIMO are examined, including diversity gain (DG) and envelope correlation
coefcient (ECC), and it is found that they meet the required standards of DG approximately equal to 10 and ECC < 0.05. Te
proposed MIMO system has been fabricated and tested. Te measured results are consistent with the design of the simulated
structure using the CST Microwave Studio (CSTMWS) simulator.

1. Introduction such as high resolution, fast data transfer, compact size

(which allows for smaller antenna dimensions), low in-
5G is expected to increase both the data rate and the number terference (which provides systems with higher immunity to
of connected devices, while also reducing latency, energy cramming), low cost, and high security [4, 5]. As a result,
consumption, cost, and poor spectrum utilization. Te 5G mm-wave antenna arrays are being investigated to provide
communication spectrum has been split into two parts by the high throughput demanded by users while also allowing
the International Telecommunications Union (ITU). One is for antenna size reduction to put the antenna elements closer
known as the sub-6 GHz spectrum [1], and the other is the together by adding the EBG structure in the ground plane to
millimeter-wave (mm-wave) spectrum [2, 3]. We investigate make the antenna miniaturized, but with a complex design
the mm-wave frequency band, which has the potential to [6]. However, due to the inevitable and undesirable mutual
support a larger number of users due to its many benefts, coupling between antenna parts, antenna engineers are
2 International Journal of Antennas and Propagation

becoming increasingly concerned about the design of an- 37.5 GHz, and a maximum isolation improvement of 35 dB
tenna arrays in mm-wave bands without compromising gain is achieved with the use of closely spaced MIMO elements in
and efciency performance [7]. Te isolation between the the form of patch antennas [15]. No additional or external
antenna elements in mm-wave bands can be increased by barriers have been placed for isolation improvement. In the
increasing the distance between them or incorporating the past few years, the metasurface loading technique has gained
isolation structures and other various techniques. In [8], a lot of attention for its potential to increase antenna gain by
a miniaturized MIMO antenna based on metamaterials for manipulating the phase and amplitude of the incident
use in 5.8-GHz WLANs is presented. To achieve a high electromagnetic (EM) wave [16, 17]. Here, a microwave-
isolation of 50 dB between the two antenna elements, a series operating dual-band metasurface is employed that can
of metamaterial absorber unit cells are integrated between manipulate refected and transmitted waves independently
them. As decoupling structures, several electromagnetic in each half-space. Te developed C-shaped metasurface has
band gap (EBG) surfaces were used in [9] within the im- two C-shaped patterns and a C-shaped slot on three layers,
pedance bandwidth of 5.8–6.4 GHz, increasing the isolation and it is able to adjust the opening and orientation angles at 7
between the patch array by around 50 dB. Te isolation and 17 GHz to change the amplitude and phase responses of
bandwidth was increased from 4.0 to 6.5 GHz (or 48%) by the refected and transmitted waves, respectively. Gain
incorporating three double F-shaped etchings into the array fuctuation in the band of interest is 8.2–10.2 dB, and the
ground plane, with a corresponding increase in isolation of improvement in endfre gain with metasurface loading is
10 to 40 dB across the total isolation bandwidth. Te mul- observed to be 73% greater at 28 GHz. With the reduction
tiresonance EBG structure and the double F-shaped ground paradigm toward nonuniform phase errors at the Vivaldi
plane help minimize surface waves as they move through the aperture through metasurface loading, the work in [18]
substrate. Reference [10] employed a slot-array patch an- intends to produce a high-gain Vivaldi MIMO antenna with
tenna and a defective ground surface (DGS) to achieve a 50- pattern diversity for the application of the 5G mm-wave
dB decoupling between 1.2–1.35 GHz with edge-to-edge band. Te proposed MIMO antenna has a maximum gain of
spacing of 0.057λ between CP antenna elements, improving 12 dB. Te high gain of the suggested antenna can also be
isolation by around 50 dB. To further mitigate the issue of useful in overcoming the attenuations in signal strength
backward radiation from slot-array DGS, a wheel-shaped caused by the atmosphere at higher frequencies in mm-wave
absorber based on the electromagnetic absorption material communication systems. Reference [19] describes a tree-
has been developed and manufactured. Without degrading shaped patch-based MIMO antenna. Te bandwidth of the
other antenna characteristics, the absorber has an absorp- impedance provided by the radiating structure is between 23
tivity of greater than 95% in the frequency range of and 40 GHz. Te developed patch antenna was also found to
1.2–1.35 GHz and reduces backward radiation by more than have a high gain of 11.45 dB in the target frequency range.
12.5 dB in the plane phi � 0° and 16.1 dB in the plane Te MIMO antenna was designed in [20] to operate in the
phi � 90°. Te use of metallic vias as a simple decoupling 24.55–26.5 GHz region. An array of metasurfaces was placed
technique for better isolating millimeter-wave MIMO di- on the MIMO antenna to achieve good isolation (>25 dB).
electric resonator antenna (DRA) elements is studied in [11]. Te MIMO antenna has a measured impedance bandwidth
At 26 GHz, isolation can be improved from 15.1 to 34.2 dB of 1.95 GHz, which makes it suitable for use in the 26 GHz
for an H-plane-connected MIMO DRA array and from 13.1 range. Furthermore, the gain is increased by 2.1 dB, with
to 43 dB for an E-plane array. Te DB decoupling method is a reported maximum gain of 10.27 dB, when a metasurface is
a novel method for decoupling MIMO patch antennas, and placed above the MIMO antennas at a distance of around
the dielectric block was proposed by the authors of [12] to 0.5 l λ. Te MIMO antenna presented in [21] operates at
reduce the coupling between close patch antenna elements a dual frequency, where the frst resonance had a bandwidth
in the E and H planes by more than 20 dB over a bandwidth of 10 GHz, covering the range from 23 GHz to 33 GHz, and
of 4.32–5.48 GHz. Te 28 GHz band (27.5–28.35 GHz) is the second resonance had a bandwidth of 3.25 GHz, cov-
allocated by the Federal Communications Commission ering the range from 37.75 GHz to 41 GHz. Te isolation
(FCC) for 5G applications, and the proposed antenna ac- between radiating elements was observed to be >20 dB at
quires a simulated impedance bandwidth (S11 below both resonance bands. Te authors in [22] built and tested
− 10 dB) from 27.25 to 28.59 GHz to cover this range. To a 4-port MIMO antenna system operating in the 25–50 GHz
increase isolation by a maximum of 12 dB at 27.5–28.35 GHz band. Te antenna’s ground plane has been truncated in the
[13], metal strips were printed on a MIMO DRA. Changing middle, and the elements’ orthogonal orientation is the
the layout of a patch antenna in a slot operating at 56.5 to foundation of the design. A high DG of 9.999 dB, low TARC
64 GHz increased isolation by a maximum of 38.6 dB [14]. A of less than − 10 dB, low CCL of 0.25 bits/s/Hz, low ECC of
U-shaped coupling slot with a microstrip feed is placed less than 0.005, and low mutual coupling of less than − 10 dB
symmetrically along the x and y axes to achieve balanced are just some of the exceptional performance parameters
radiation performance for polarization diversity. Te orig- attained by the proposed antenna system. Reference [23]
inal patch’s wideband performance was improved by sizing introduces compact MIMO antennas for submillimeter-
the U-shaped slots for resonance. To maintain isolation, the wave 5G NR bands, which are fed by CPW. Tese anten-
current must be routed through the isolation port’s feed line. nas have a fat design and ofer high gain, wide bandwidth,
Te 5G mm-wave self-isolated MIMO antenna technology is and improved radiation efciency. Te upper surface of the
simple and efcient in the mm-wave frequency range of 28- antennas includes a modifed CPW feed into a slotted
International Journal of Antennas and Propagation 3

circular patch, arranged orthogonally to create a 4-port elements, and the elements are arranged in an orthogonal
structure. Trough careful optimization, the antennas manner to achieve the compact size of the proposed
achieve isolation greater than 20 dB. With dimensions of MIMO system. Finally, a metamaterial slab is inserted in
24 × 24 mm2, the antennas demonstrate a bandwidth the center of the substrate and adjacent to radiating el-
ranging from 24.8 to 44.45 GHz, a maximum gain of 8.6 dBi, ements, which improves the isolation by 10 dB within the
and a minimum efciency of 85%. In [24], a small whole operating bandwidth.
24 × 20 mm2 MIMO antenna consisting of four elements
with a full ground plane is presented. Tis antenna boasts an 2. MIMO Antenna Design
impressive 85% efciency due to the utilization of AgHT-8
and Plexiglas materials. It operates efectively in two fre- 2.1. Single-Element Optimization. Te proposed antenna is
quency bands, 24.10–27.18 GHz (12% bandwidth) and designed using optimization techniques and involves two
33–44.13 GHz (28.86% bandwidth), specifcally designed for steps. In the frst step, a U-shaped patch is designed with
mm-wave 5G applications. It provides isolation exceeding a full ground plane. Ten, in the second step, the U-shaped
16 dB, maximum gain surpassing 3 dBi and efciency ex- patch remains unchanged while the ground plane is mod-
ceeding 75. Te research presented in [25] is a novel dual- ifed and made defective to achieve the fnal proposed de-
band, dual-polarization MIMO antenna designed for 5G sign, as shown in Figure 1.
applications operating at 28 and 38 GHz frequencies. Tese are key mathematical formulas assisting in the
Constructed on the Rogers TM RO4003 substrate, the an- design of the proposed antenna.
tenna has dimensions of 3.77 λ0 × 1.86 λ0. By utilizing an
aperture-coupled patch confguration and CPW trans- C
wp � 􏽰����������
mission line, it achieves dual-band operation. A parasitic 2fc εrelative + 1/2
director is used to enhance directivity at 28 GHz. Te 2-port
MIMO antenna ofers an impedance bandwidth of εrelative + 1 εrelative + 1 1
S11<− 10 dB from 27.6 to 30.3 GHz and 33.4 to 40.4 GHz, εeff � + + 􏽱���������
with a gain of over 8 dBi, isolation above 20 dB, and cross- 2 2 1 + 12h/wp
polarization level exceeding 20 dB [26]. Te compact fork-
shaped MIMO antenna technology that is being shown , (1)
ofers innovative advancements for 5G millimeter-wave 􏼐εeff + 03.􏼑􏼐wp /h + 0.264􏼑
∆L � 0.1421h
applications. First of, it achieves a small dimension of 􏼐εeff − 0258.􏼑􏼐wp /h + 0.8􏼑
36 × 28 mm2 with a large bandwidth due to its orthogonally
structured 2 × 2 MIMO confguration. Second, by in-
tegrating a metamaterial slab, isolation is enhanced by 10 dB c
Lp� 􏽰���� − 2∆L
over the band, with a peak isolation of 65 dB at 34.5 GHz. 2fo εeff
Te 5G mm-waves technology, which operates in the Ka-
band (22–50 GHz), promises efciency >80% and isolation where Ls and Ws represent the substrate’s length and width,
>30 dB across the spectrum, which is essential for 5G while h denotes the substrate’s height with efective per-
communication. Certainly, connecting the MIMO ground mittivity and relative permittivity “ εeff and ” εrelative . C stands
planes can make isolation problems worse. A connected for the speed of light, fc denotes the central frequency, and
ground MIMO arrangement with an integrated isolation ∆L represents the efective length. Figure 2 shows that the
mechanism would be more efective against interference. refection coefcient of step 1 is higher than − 10 dB, in-
Improved isolation gains may be obtained by arranging the dicating a complete mismatch. However, in step 2, the re-
metamaterial slab in a strategic way between radiating el- fection coefcient is lower than − 10 dB and provides
ements inside the interconnected ground arrangement. Tis a wideband impedance bandwidth ranging from 22 to
approach ensures lower interference and better overall 46 GHz. As a result, the proposed antenna is highly suitable
performance, ofering a more comprehensive solution for for 5G mm-wave communication.
5G millimeter-wave applications where isolation is a major In Figures 3(a) and 3(b), the design and refection co-
problem for the MIMO antenna system. efcient of a new antenna array are presented. Te antenna
In this article, a four-element MIMO antenna is consists of a 1 × 2 U-shaped radiator with a partial ground
presented for 5G mm-wave applications. Te proposed plane and is intended for use in mm-wave applications. It
antenna has a large bandwidth with decent MIMO fea- was built on a low-loss substrate made from Rogers RT-
tures for 5G millimeter-wave applications. Te large Duroid 5880, which is 0.8 mm thick with a 0.0009 loss
bandwidth of the antenna is more useful in achieving tangent and a dielectric constant of 2.2. Te substrate itself
a high data rate for communication. Te antennas of the measures 18 × 8 × 0.8 mm3. To ensure that the radiator re-
reported MIMO antenna structure are evolved in two ceives maximum power, it was connected to a simple
steps as shown in Figure 1, with a defected ground microstrip feed and designed using CST Microwave Studio
structure which mostly provides a wideband character- (CSTMWS). Refer to Table 1 for more information on the
istic. Furthermore, initially, the MIMO structure consists fnal design values. Figures 4, 5, 6, and 7 show a single
of two radiating elements which are perpendicular to each antenna’s 3D radiation patterns, 2D radiation patterns, gain,
other, and then the MIMO structure is modifed into four and efciency, respectively.
4 International Journal of Antennas and Propagation

(a) (b)

Figure 1: Optimization steps of the proposed antenna. (a) with full Ground Plane (b) with half ground plane.

-5
Refection coefcient (dB)

-10

-15

-20

-25

-30

-35
20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Freqency (GHz)

S1, 1 FG
S1, 1 HG
Figure 2: Refection coefcient responses of the optimization steps of the proposed antenna.

Te radiation patterns in both principal planes E and H As can be seen in Figure 10, these MIMO setups are
at 29 GHz are shown in Figure 5. It is observed that the highly helpful in achieving a large bandwidth that ranges
bidirectional radiation pattern is achieved for the E-plane, from 22 to 50 GHz and a strong isolation of greater than
while the omnidirectional radiation pattern is attained for 20 dB between the radiation elements.
the H-plane.
Te refection coefcient of an antenna is infuenced by
2.3. Four-Element MIMO Confguration. For reliable
the length of its ground plane. In this investigation, varia-
5G mm-wave communication, the proposed antenna has
tions in the ground plane’s length during parametric ana-
been transformed into a four-port MIMO system with high
lyses have afected the antenna’s performance. Specifcally,
channel capacity and high data trafc. In Figure 11(a), the
when the ground plane changes in length, the values above
layout details of the proposed MIMO antenna system are
− 10 dB are observed for refection coefcients for lengths of
shown. Good isolation of more than 20 dB is achieved
1, 2, 3, and 4 mm. However, in Figure 8, optimal perfor-
without a decoupling structure by placing the MIMO ele-
mance is achieved with a ground plane length (lg) of 5 mm, ments in a perpendicular confguration on the top of the
resulting in the desired wide impedance bandwidth ranging
substrate. Figure 11(b) shows that from the back, each ra-
from 22 to 46 GHz. Tis indicates that a properly di-
diating element has its own partial ground plane that is
mensioned ground plane is crucial for achieving the desired
orthogonal to the rest of the element. Te MIMO ar-
antenna performance, ensuring refection coefcients below
rangement has a size of 36 × 28 × 0.8 mm3 and consists of
− 10 dB across the desired frequency range.
four radiating elements. Te alignment of antennas in
MIMO setups plays a crucial role in enhancing signal di-
2.2. Two-Element System. As shown in Figure 9, the single versity and minimizing interference. By placing antennas
antenna array provided here has been transformed into orthogonal or at right angles to each other, it optimizes the
a two-element MIMO structure with overall dimensions of spatial spacing, resulting in reduced correlation among
28 × 18 × 0.8 mm3. Te MIMO elements are arranged on the transmitted signals and improved channel capacity. Addi-
top of the substrate at right angles to one another. tionally, this confguration simplifes signal processing
International Journal of Antennas and Propagation 5

Wp1

Lp1
Lp

Wp

Lf Lg

Wf Wg
(a)
0
-5
-10
Refection coefcient (dB)

-15
-20
-25
-30
-35
-40
-45
-50
20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency (GHz)
Single element
1×2 Array
(b)

Figure 3: (a) Layout of the proposed antenna. (b) Refection coefcient.

techniques such as Maximum Ratio Combining and Zero 2.4. Unit Cell Optimization. Te unit cell described in this
Forcing, thereby enhancing the overall performance of paper is optimized for use with millimeter waves at fre-
wireless communication systems. Simulated S-parameters of quencies between 24 and 44 GHz. To design the unit cell
the proposed antenna are shown in Figure 12(a). It can be suggested, a Rogers RT duroid 5880 substrate of thickness
seen that the − 10 dB bandwidth extends from 22 to 50 GHz 0.8 mm was used. In the frequency domain of a unit cell,
with an overall isolation better than − 22 dB. In addition, the CSTMWS is used to run the simulation. Te proposed
refection coefcient of proposed MIMO with the connected optimized unit cell, represented in Figure 13(b), is made up
ground plane above − 10 dB at lower frequencies ranges from of two rectangular split-ring resonators with a combined H-
24 to 33 GHZ, while at higher frequencies, it attains an shaped arrangement in the middle. Te gradual evolution of
impedance bandwidth of 17 GHz from 33 to 50 GHz as the unit cell is depicted in a similar fashion in Figure 13(a).
shown in Figure 12(b), while the transmission coefcient Te unit cell operates in the 24–44 GHz region of the mm-
below − 20 dB shows good isolation among radiating waves spectrum, as shown by the refection coefcient plot in
elements. Figure 13(c). Te transmission coefcient curve establishes
6 International Journal of Antennas and Propagation

Table 1: Design parameters of the fork-shaped antenna and MIMO the entire transmission in the operating frequency band,
system. which also shows that the loss of a unit cell is close to zero.
Symbol Dimension size (mm) As shown in Figure 14(a), a metamaterial slab was
Ls 18
subsequently constructed by periodic repetition of 5 × 4 unit
Lf 5 cells placed in the center of the proposed MIMO antenna
Lp 1 system, and a 3 × 3 unit cell of metamaterials was inserted
Lp1 2 between two radiating elements in situations when the two
Lg 5 antennas are perpendicular. Te MIMO system’s perfor-
Ls1 18 mance was improved after using a metamaterial because of
Ls2 36 its bandpass behavior, which is inserted within MIMO
Ws 8 setups. Figure 14(c) shows an improvement in isolation over
wf 1.7 the entire wide bandwidth, with the refection coefcient
Wp 5 curve of the MIMO system before and after insertion of the
Wp1 1
metamaterial having the same impedance bandwidth of
Wg 8
Ws1 28
22–50 GHz.
Ws1 28
3. Experimental Results
y For experimental verifcation, the presented MIMO system
with a metamaterial slab is developed as shown in Figure 15.
dB To investigate the impacts of the metamaterial on MIMO
5.94 performance, it must be inserted into the MIMO system at
4.45 a specifed orientation. In the following subsections, the
2.97
Phi 1.48 results of the simulations and the measurements compared
x
0 y to one another will be discussed. Discrepancies between
Theta –8.52 simulated and experimental outcomes in MIMO antennas,
z
–17
–25.5 irrespective of their size, require careful examination. Several
z x factors such as material properties, boundary conditions,
–34.1
environmental factors, manufacturing accuracy, experi-
Figure 4: 3D far-feld radiation pattern of the proposed single mental setups, electromagnetic interactions, frequency
antenna. variations, and model simplifcations impact these difer-
ences. It is essential to use thoughtful analysis to understand
0 the root causes and improve simulation techniques. Tis
5 enables adjustments in parameters, assumptions, and ex-
-30 30 perimental setups. By continuously refning this process, the
0
-5 accuracy of both simulations and experiments improves,
-10
resulting in better MIMO antenna behavior and more
-60 60 precise performance evaluations.
-15
Figures 16(a) and 16(b) show a comparison between the
-20
simulated and measured values for the S-parameters. Ant 1
-25 and Ant 2’s impedance bandwidth was 28 GHz, and their
-90 -30 90 simulated and measured refection coefcients were in the
22–50 GHz range. Te transmission coefcients of Ant1 and
Ant2 with Ant3 and Ant4 antennas are also investigated. As
can be seen in Figure 16(b), isolation levels within the
-120 120 operating band range from a minimum of 27 dB to a max-
imum of 65 dB. Fabrication defects, calibration errors, and
cable losses contribute to a small but noticeable discrepancy
-150 150 between the simulated and observed fndings.
180 Te radiation patterns of the proposed MIMO system are
measured in an anechoic chamber. Figures 17(a) and 17(b)
E-plane illustrate the measured and simulated pattern for H and E
H-plane planes at 28 and 38 GHz, respectively. Tere are small dif-
Figure 5: E- and H-plane patterns of the proposed single antenna ferences between the measured and simulated results due to
at 29 GHz. calibration errors and cable losses during measurement.
International Journal of Antennas and Propagation 7

10
9
8
7
6
Gain (dB)
5
4
3
2
1
0
-1
20 25 30 35 40 45 50
Frequency (GHz)
Figure 6: Gain of the proposed single antenna.

1
0.9
0.8
0.7
Total Efciency

0.6
0.5
0.4
0.3
0.2
0.1
0
20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency (GHz)
Figure 7: Total efciency of the proposed single antenna.

-5
Refection coefcient (dB)

-10

-15

-20

-25

-30

-35
20 22 24 26 28 30 32 34 36 38 40 42 44 46 48
Frequency (GHz)

S1, 1 (lg=1) S1, 1 (lg=4)

S1, 1 (lg=2) S1, 1 (lg=5)
S1, 1 (lg=3)
Figure 8: Efects of the length (lg) of the ground plane on the proposed antenna.
8 International Journal of Antennas and Propagation

Ls1

Ws1
Figure 9: Two-element MIMO system’s front and back views.

-10

-20
S-parameters (dB)

-30

-40

-50

-60

-70
22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency (GHz)

S1, 1 S2, 1
S1, 2 S2, 2
Figure 10: Two-element MIMO antenna system’s S-parameters.

Ls2

Ws2
(a) (b)
Figure 11: Continued.
International Journal of Antennas and Propagation 9

(c) (d)

Figure 11: (a, c) Front view and (b, d) back view of the proposed four-element MIMO system.

-10
S-parameters (dB)

-20

-30

-40

-50

-60
22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency (GHz)

S1, 1 S1, 2 S1, 3 S1, 4 S2, 1 S2, 2

S2, 3 S2, 4 S3, 1 S3, 2 S3, 3 S3, 4
S4, 1 S4, 2 S4, 3 S4, 4
(a)
Figure 12: Continued.
10 International Journal of Antennas and Propagation

-10

-20

S-parameters
-30

-40

-50

-60

-70
20 25 30 35 40 45 50
Frequency (GHz)

S1, 1 S1, 2 S1, 3 S1, 4 S2, 1 S2, 2

S2, 3 S2, 4 S3, 1 S3, 2 S3, 3 S3, 4
S4, 1 S4, 2 S4, 3 S4, 4
(b)

Figure 12: S-parameters of the proposed MIMO system in each port (a) with an individual ground plane and (b) with a connected
ground plane.

(a)
Ws

Wc
Lc

Ls Lp

Wp

(b)
0
-2
-4
S-parameters (dB)

-6
-8
-10
-12
-14
-16
-18
-20
20 22 24 26 28 30 32 34 36 38 40 42 44
Frequency (GHz)

S1, 1 S2, 1
S1, 2 S2, 2
(c)

Figure 13: (a) Unit cell optimization steps. (b) Proposed unit cell geometry. (c) S-parameter of the unit cell.
International Journal of Antennas and Propagation 11

(a) (b)
0

-10

-20
S-parameters (dB)

-30

-40

-50

-60

-70
22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency (GHz)

S1, 1 S2, 2 S3, 3 S4, 4

S1, 2 S1, 3 S1, 4 S2, 1
S2, 3 S2, 4 S3, 1 S3, 2
S3, 4 S4, 1 S4, 2 S4, 3
(c)

Figure 14: (a) Proposed MIMO antenna system based on metamaterials. (b) Back view. (c) S-parameters.

(a) (b)

Figure 15: Fabricated MIMO antenna. (a) Front view and (b) back view of the metamaterial-based MIMO system.
12 International Journal of Antennas and Propagation

-5

-10

Refection coefcient (dB)

-15

-20

-25

-30

-35

-40

-45
22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency (GHz)

S1, 1 S1, 1 Mea

S2, 2 S2, 2 Mea
(a)
0

-10
Transmission coefcient (dB)

-20

-30

-40

-50

-60

-70
22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency (GHz)

S1, 2 S1, 3 S1, 4 S2, 1

S2, 3 S2, 4 S1, 2 Mea S1, 3 Mea
S1, 4 Mea S2, 1 Mea S2, 3 Mea S2, 4 Mea
(b)

Figure 16: Simulated and measured results. (a) Refection coefcient and (b) transmission coefcient of the metamaterial-based MIMO system.

Figure 18 shows the simulated and measured co- and It has been investigated that if we increase the frequency,
cross-polarization patterns in the E and H planes. Te the radiation becomes more directive and the gain of the
MIMO antenna presented here exhibits favorable charac- radiating patch improves [27, 28]. Moreover, as the gain
teristics across frequency bands at 28 and 38 GHz fre- increases, the efciency of the radiating element also in-
quencies. Te cross-polarizations of the E-plane and creases, and a maximum total efciency of 94% is achieved at
copolarizations of the H-plane have the same behavior at 26 GHz and more than 80% efciency over the overall
both frequencies, that is, 28 GHz and 38 GHz. Tis signifes operating bandwidth, as depicted in Figure 19. From the
excellent isolation between polarizations, which is crucial for plots, it shows that measured results are closely in agreement
reducing interference and ensuring reliable communication with simulated results.
in MIMO systems. Te simulated and measured results Te antenna confgurations, channel conditions, signal
closely align, confrming the antenna’s consistent perfor- processing techniques, hardware specifcations, data pro-
mance and validating the design methodology. Tese cessing approaches, environmental infuences, and temporal
characteristics hold great promise for enhancing the capacity fuctuations can lead to variations in MIMO measurement
and reliability of MIMO systems, particularly in high- processes. By elucidating these diferences, the credibility
frequency applications. and scientifc precision can be enhanced, thereby advancing
International Journal of Antennas and Propagation 13

0 0
5 5
-30 30 -30 30
0
0
-5
-60 -5 60 -60 60
-10
-10
-15

-90 -15 90 -90 -20 90

-120 120 -120 120

-150 150 -150 150

180 180

E-plane S H-plane M E-plane S H-plane M

H-plane S E-plane M H-plane S E-plane M

(a) (b)

Figure 17: Simulated and measured results of the E and H planes. (a) 28 GHz and (b) 38 GHz.

the understanding of the proposed MIMO systems as uniform across the patch. Tis uniform distribution in-
compared to the literature. dicates that the energy is radiating away from the antenna
A metamaterial-based four-element MIMO (multiple structure as electromagnetic waves.
input multiple output) system is being analyzed at 28 GHz
frequency as shown in Figure 20. When port one (Antenna 4. MIMO Performance Parameters
1) is excited, the surface current distribution on its radiating
element is primarily concentrated near the feed region. Tis Some of the MIMO antenna performance parameters which
concentration of current is high, as the electromagnetic can explore MIMO performance are discussed below.
energy is being injected into the antenna structure. Now,
because of the specifc orientation of the metamaterial slab
among the radiating elements, the efects on the other ra- 4.1. Envelope Correlation Coefcient (ECC). Te envelope
diating elements are negligible as shown in Figure 20(a). Tis correlation coefcient (ECC), which is a widely used MIMO
means that when Antenna 1 is excited, it does not signif- performance parameter, gives information on the correla-
cantly infuence the behavior of the other radiating elements. tion between the radiating elements in the MIMO antenna
Tis is a desirable characteristic for MIMO systems because setup as given in equation (2). A low level of correlation is
it helps to isolate the signals transmitted or received by each observed for independent antenna elements. Te proposed
antenna element, reducing interference and improving MIMO structure has a low value of ECC< 0.005, which is less
overall system performance. Furthermore, the metamaterial than the standard threshold of 0.5. Te wide band antenna
slab is strategically placed among the MIMO elements. As ECC is estimated from the radiation pattern [29].
a result, the current from the radiating elements also dis- 􏼌􏼌 􏼌
tributes among the unit cells of the metamaterial slab. Tis 􏼌􏼌J4π M 􏼁(θ, ϕ) × 􏼐 M (θ, ϕ)􏼑dΩ􏼌􏼌􏼌2
􏼌 i j 􏼌
redistribution of current helps to reduce coupling among the ECC � 􏼌􏼌 􏼌􏼌2 􏼌􏼌 􏼌􏼌2 , (2)
radiating elements. Coupling can lead to interference and J4π􏼌􏼌 Mi (θ, ϕ)􏼁􏼌􏼌 dΩJ4π􏼌􏼌􏼌􏼐Mj (θ, ϕ)􏼑􏼌􏼌􏼌 dΩ
degradation of signal quality in MIMO systems, so mini-
mizing it is crucial for achieving high-performance com- where Mi (θ, ϕ) and Mj (θ, ϕ) note the radiation patterns
munication. Te surface current density distribution on the when ports i and j are excited, respectively; and the term Ω
patch, near the feed region of each radiating element, shows denotes the solid angle. Te proposed four-element
strong current densities corresponding to the energy being metamaterial-based MIMO structure has a very low value
input into the antenna. However, as the distance from the of ECC, as shown in Figure 21. Good isolation is shown
feed region increases, the current density becomes more among radiating elements.
14 International Journal of Antennas and Propagation

0 0
5 5
-30 30 -30 30
0 0

-60 -5 60 -60 -5 60

-10 -10

-90 -15 90 -90 -15 90

-120 120 -120 120

-150 150 -150 150

180 180

Cross-Pol (Simu) Co-Pol (Simu) Co-Pol (Simu) Cross-Pol (Simu)

Cross-Pol (Meas) Co-Pol (Meas) Co-Pol (Meas) Cross-Pol (Meas)
(a) (b)
0 0
5 5
30 30 30 30
0 0

-5 -5
60 60 60 60
-10 -10

-15 -15

90 -20 90 90 -20 90

120 120 120 120

150 150 150 150

180 180

Co-Pol (Simu) Cross-Pol (Simu) Cross-Pol (Simu) Co-Pol (Simu)

Co-pol (Meas) Cross-Pol (Meas) Cross-Pol (Meas) Co-Pol (Meas)
(c) (d)

Figure 18: Simulated and measured radiation patterns of the proposed antenna in the E-plane and the H-plane. (a) 28 GHz (E-plane).
(b) 28 GHz (H-plane). (c) 38 GHz (E-plane). (d) 38 GHz (H-plane).

4.2. Diversity Gain (DG). Te DG of a MIMO antenna is the Table 2 shows a comparison of the proposed antenna
loss in transmission power for the MIMO confguration. It with the relevant work reported in the literature. Te pro-
can be calculated using equation (3) [29]. Figure 21 shows posed MIMO system has a wider bandwidth, higher ef-
the DG curves of the MIMO antenna. Te fgure shows that ciency, and lower ECC than most of the reported work. In
the DG is about 10 dB over the entire band of the MIMO contrast to current approaches, the proposed antenna
antenna, which ensures good diversity performance. structure introduces a compact fork-shaped MIMO system
􏽱���������� with an orthogonally oriented 2 × 2 confguration aug-
DG � 10 1 − |ECC|2 . (3) mented by a metamaterial slab placed strategically for
International Journal of Antennas and Propagation 15

1
0.9
0.8
0.7

Total efciencies
0.6
0.5
0.4
0.3
0.2
0.1
0
22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency (GHz)

Ant 1 Ant 2 Ant 3 Ant 4

Ant 1 Mea Ant 2 Mea Ant 3 Mea Ant 4 Mea
Figure 19: Simulated and measured total efciencies of the metamaterial-based MIMO system.

A (m)
20
18.2
16.4
14.5
12.7
10.9
9.09
7.27
5.45
3.64
1.82
0

(a) (b) (c) (d)

Figure 20: Surface current distribution at 28 GHz across each port of the metamaterial-based MIMO system. (a) Port 1. (b) Port 2. (c) Port 3.
(d) Port 4.

0.0004 10

0.00035 9.9995

0.0003 9.999
Diversity Gain

0.00025 9.9985
ECC

0.0002 9.998

0.00015 9.9975

0.0001 9.997

0.00005 9.9965

0 9.996
20 25 29 34 38 43 47
Frequency (GHz)

ECC Mea DG Mea

ECC Sim DG Sim
Figure 21: Simulated and measured ECC and DG of the metamaterial-based MIMO system.
 16

Table 2: Performance of the proposed mm-wave antenna compared to the recent literature.
Frequency
Year ref. Antenna size (mm2) Type of antenna Decoupling method Isolation (dB) Efciency (%) ECC
of antenna (GHz)
2021 [12] N.G Printed patch 5–10 Dielectric block DB 20.6 97 N.G
2019 [13] 20 × 20 DRA 26–30 Metal strips 24 N.G 0.013
2020 [19] 80 × 80 × 1.57 Printed patch 20–40 N.A 20 70 0.0014
2021 [20] 30 × 43 Printed patch 24–27 Metasurface 45 N.G 0.1e(− 6)
2023 [21] 48 × 12 Printed patch 23–41 Defected ground structure 20 82 0.00015
2023 [22] 33 × 33 Printed patch 20–50 Orientation and DGS N.G 80 0.005
Proposed 36 × 28 Printed patch 20–50 Metamaterial and DGS <30 94 0.0005
N.A stands for not available. N.G stands for not good.
International Journal of Antennas and Propagation
International Journal of Antennas and Propagation 17

improved isolation. Tis innovation achieves a maximum Microwave and Optical Technology Letters, vol. 61, no. 3,
isolation of 65 dB at 34.5 GHz, a signifcant improvement of pp. 781–786, 2019.
10 dB over the whole operating band. Te proposed antenna [6] X. Shen, Y. Liu, L. Zhao, G. Huang, X. Shi, and Q. Huang, “A
geometry maintains compactness and provides greater miniaturized microstrip antenna array at 5G millimeter-wave
isolation performance over previous methods, thus playing band,” IEEE Antennas and Wireless Propagation Letters,
a vital role in 5G millimeter-wave applications. vol. 18, no. 8, pp. 1671–1675, 2019.
[7] S. F. Jilani and A. Alomainy, “Millimetre-wave T-shaped
MIMO antenna with defected ground structures for 5G
5. Conclusions cellular networks,” IET Microwaves, Antennas & Propagation,
vol. 12, no. 5, pp. 672–677, 2018.
Tis paper presents a four-element metamaterial-based [8] A. Madni, R. M. Hasan Bilal, and W. T. Khan, “A compact
MIMO antenna system for 5G mm-wave communication metamaterial based high isolation MIMO antenna for 5.8
operating for a wide band of frequencies from 22 to 50 GHz. GHz WLAN applications,” IEEE International Symposium on,
Te MIMO elements are arranged in orthogonal orientation 2022.
on the top of the substrate with a size of 36 × 28 × 0.8 mm3 [9] Antennas and Propagation Society International Symposium,
and achieved good isolation >30 dB over a wide frequency in Proceeding of the IEEE International Symposium on An-
range. Te MIMO system is developed from the single- tennas and Propagation and USNC-URSI Radio Science
element to the four-element MIMO system to improve the Meeting (AP-S/URSI), pp. 245-246, Denver, CO, USA, July
channel capacity and increase the data rate. Moreover, the 2022.
metamaterial was inserted in the MIMO system which [10] O. Sokunbi and H. Attia, “Highly reduced mutual coupling
between wideband patch antenna array using multiresonance
improves the isolation of the proposed MIMO system and
EBG structure and defective ground surface,” Microwave and
attains a maximum isolation of 65 dB and a maximum total
Optical Technology Letters, vol. 62, no. 4, pp. 1628–1637, 2020.
efciency of 94%. In addition, the MIMO performance [11] Y. M. Pan, X. Qin, Y. X. Sun, and S. Y. Zheng, “A simple
parameters ECC and diversity gain are analyzed, which decoupling method for 5G Millimeter-wave MIMO dielectric
describe good MIMO characteristics. Finally, the measured resonator antennas,” IEEE Transactions on Antennas and
and simulated results show similarity, which makes it Propagation, vol. 67, no. 4, pp. 2224–2234, 2019.
a suitable candidate for 5G mm-wave applications. [12] D. Gao, Z.-X. Cao, S. D. Fu, X. Quan, and P. Chen, “A novel
slotarray defected ground structure for decoupling microstrip
Data Availability antenna array,” IEEE Transactions on Antennas and Propa-
gation, vol. 68, no. 10, pp. 7027–7038, 2020.
Te data used to support the fndings of this study are [13] M. Li, M. Y. Jamal, L. Jiang, and K. L. Yeung, “Isolation
available from the corresponding author upon request. enhancement for MIMO patch antennas sharing a common
thick substrate: using a dielectric block to control space-wave
coupling to cancel surface-wave coupling,” IEEE Transactions
Conflicts of Interest on Antennas and Propagation, vol. 69, no. 4, pp. 1853–1863,
2021.
Te authors declare that they have no conficts of interest. [14] Y. Zhang, J. Deng, M. Li, D. Sun, and L. Guo, “A MIMO
dielectric resonator antenna with improved isolation for 5G
Acknowledgments mm-wave applications,” IEEE Antennas and Wireless Prop-
agation Letters, vol. 18, no. 4, pp. 747–751, 2019.
Tis work was supported by the national funds through [15] T. H. Jang, H. Y. Kim, D. M. Kang, S. H. Kim, and C. S. Park,
FCT - Fundação para a Ciência e a Tecnologia as part the “60 GHz low-profle, wideband dual-polarized U-slot coupled
project AIEE-UAV (no. 2022.03897.PTDC). patch antenna with high isolation,” IEEE Transactions on
Antennas and Propagation, vol. 67, no. 7, pp. 4453–4462, 2019.
[16] O. Sokunbi, H. Attia, A. Hamza, A. Shamim, Y. Yu, and
References A. A. Kishk, “New self-isolated wideband MIMO antenna
[1] U. Rafque, S. Khan, M. M. Ahmed et al., “Uni-planar MIMO system for 5G mm-wave applications using slot characteris-
antenna for sub-6 GHz 5G mobile phone applications,” tics,” IEEE Open Journal of Antennas and Propagation, vol. 4,
Applied Sciences, vol. 12, no. 8, p. 3746, 2022. pp. 81–90, 2023.
[2] S. H. Kiani, A. G. Alharbi, S. Khan, M. Marey, H. Mostafa, and [17] H. W. Tian, X. G. Zhang, W. X. Jiang et al., “Programmable
M. A. Khan, “Wideband three loop element antenna array for controlling of multiple spatial harmonics via a nonlinearly
future 5G mm wave devices,” IEEE Access, vol. 10, phased grating metasurface,” Advanced Functional Materials,
pp. 22472–22479, 2022. vol. 32, no. 31, Article ID 2203120, 2022.
[3] Y. Luo, Y. Shen, X. Cai et al., “Substrate integrated coaxial line [18] L. Bao, X. Fu, R. Y. Wu, Q. Ma, and T. J. Cui, “Full-space
design for mm wave antenna with multilayer confguration,” manipulations of electromagnetic wavefronts at two fre-
International Journal of RF and Microwave Computer-Aided quencies by encoding both amplitude and phase of meta-
Engineering, vol. 32, no. 5, Article ID e23090, 2022. surface,” Advanced Materials Technologies, vol. 6, no. 4, 2021.
[4] W. Hong, “Solving the 5G mobile antenna puzzle: assessing [19] M. Singh and M. S. Parihar, “Gain improvement of Vivaldi
future directions for the 5G mobile antenna paradigm shift,” MIMO antenna with pattern diversity using Bi-axial aniso-
IEEE Microwave Magazine, vol. 18, no. 7, pp. 86–102, 2017. tropic metasurface for millimeter-wave band application,”
[5] A. Desai, T. Upadhyaya, and R. Patel, “Compact wideband IEEE Antennas and Wireless Propagation Letters, vol. 22, no. 3,
transparent antenna for 5G communication systems,” pp. 621–625, 2023.
18 International Journal of Antennas and Propagation

[20] D. A. Sehrai, M. Abdullah, A. Altaf et al., “A novel high gain

wideband MIMO antenna for 5G millimeter wave applica-
tions,” Electronics, vol. 9, no. 6, p. 1031, 2020.
[21] S. Tariq, S. I. Naqvi, N. Hussain, and Y. Amin, “A
metasurface-based MIMO antenna for 5G millimeter-wave
applications,” IEEE Access, vol. 9, pp. 51805–51817, 2021.
[22] M. E. Munir, S. H. Kiani, H. S. Savci et al., “A four element
mm-wave MIMO antenna system with wide-band and high
isolation characteristics for 5G applications,” Micromachines,
vol. 14, no. 4, p. 776, 2023.
[23] M. A. Abbas, A. Allam, A. Gaafar, H. M. Elhennawy, and
M. F. A. Sree, “Compact UWB MIMO antenna for 5G
millimeter-wave applications,” Sensors, vol. 23, no. 5, p. 2702,
2023.
[24] A. Desai, C. D. Bui, J. Patel, T. Upadhyaya, G. Byun, and
T. K. Nguyen, “Compact wideband four-element optically
transparent MIMO antenna for mm-wave 5G applications,”
IEEE Access, vol. 8, pp. 194206–194217, 2020.
[25] A. Patel, A. Desai, I. Elfergani et al., “UWB CPW fed 4-port
connected ground MIMO antenna for sub-millimeter-wave
5G applications,” Alexandria Engineering Journal, vol. 61,
no. 9, pp. 6645–6658, 2022.
[26] Y. F. Tsao, A. Desai, and H. T. Hsu, “Dual-band and dual-
polarization CPW Fed MIMO antenna for ffth-generation
mobile communications technology at 28 and 38 GHz,” IEEE
Access, vol. 10, pp. 46853–46863, 2022.
[27] I. U. Din, S. Ullah, S. I. Naqvi et al., “Improvement in the gain
of UWB antenna for GPR applications by using frequency-
selective surface,” International Journal of Antennas and
Propagation, vol. 2022, pp. 1–12, 2022.
[28] I. U. Din, W. Ullah, N. A. Abbasi et al., “A novel compact
ultra-wideband frequency-selective surface-based antenna for
gain enhancement applications,” Journal of Electromagnetic
Engineering and Science, vol. 23, no. 2, pp. 188–201, 2023.
[29] I. U. Din, S. Ullah, N. Mufti, R. Ullah, B. Kamal, and R. Ullah,
“Metamaterial-based highly isolated MIMO antenna system
for 5G smartphone application,” International Journal of
Communication Systems, vol. 36, no. 3, p. e5392, 2023.