IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 17, NO.

11, NOVEMBER 2018 2051

Compact and Low-Cost 3-D Printed Antennas

Metalized Using Spray-Coating Technology for 5G
mm-Wave Communication Systems
Shaker Alkaraki , Andre Sarker Andy , Yue Gao , Kin-Fai Tong , Zhinong Ying, Robert Donnan,
and Clive Parini

Abstract—This letter presents a design of two compact, light, including the 28 GHz band [2], [3]. Furthermore, the 28 GHz
rigid, and low-cost three-dimensionally (3-D) printed millimeter- band was among the highlighted bands above 24 GHz to be used
wave antennas for a fifth-generation (5G) communication system. for future mobile communications by the Federal Communica-
The proposed antennas consist of a radiating slot that is surrounded tions Commission [4]. Three-dimensional (3-D) printing, also
by a rectangular cavity and corrugations, which boost the gain per- known as additive manufacturing, is a promising technology
formance of the antennas. Furthermore, the proposed antennas
are fabricated using 3-D printing technology, and they are met-
that enables fast and cost-effective prototyping. Lately, several
alized using novel, simple, and low-cost techniques, which utilize metal-plated, plastic 3-D-printed antennas have been reported,
the commercial conducive spray-coating technology. The proposed where the plastic prototypes are covered by metal through the
antennas operate at a 28 GHz band, where the first design is fed by conventional plating or metalization techniques [5]–[8].
a waveguide to prove the performance, whereas the second design In this letter, we propose two small, compact, efficient, and
is fed by a microstrip line to demonstrate the ability to be inte- low-cost 3-D-printed antennas metalized using novel and simple
grated into a compact structure. Measurement results show a wide techniques based on electromagnetic interference called as the
impedance bandwidth, which enables the proposed antenna design radio frequency interference (EMI/RFI) conductive spray-paint
to be a strong candidate for 5G applications. technology. The proposed spray-coating technology is used to
Index Terms—Corrugations, millimeter-wave (mm-wave), slot metalize the 3-D-printed plastic antenna instead of conventional
antenna, three-dimensionally (3-D) printed antenna, 28 GHz, 5G. metalization methods. The proposed metalization method re-
duces the cost of fabrication and reduces the complexity of the
I. INTRODUCTION fabrication process.
HE fifth-generation (5G) mobile communications tech-
T nology, which is expected to be standardized by 2020, is a
promising technology, which is expected to achieve 1000 times
II. ANTENNA STRUCTURE AND FABRICATION METHOD
The first proposed antenna (Ant-1) consists of a resonant
overall system capacity, ten times data rate (i.e., peak data rate of slot surrounded by a rectangular cavity and two corrugations as
1 Gb/s for high mobility and 10 Gb/s for low mobility), 25 times shown in Fig. 1. The rectangular cavity and the two corrugations
average cell throughput, and at least 10 times spectral and en- improve the directivity and gain of the antenna. In addition, the
ergy efficiency with five times lower latency. The main target of electromagnetic (EM) energy is coupled to the antenna surface
the 5G technology is to connect the entire world and to achieve using a standard rectangular waveguide by the means of resonant
extremely reliable, robust, and energy-efficient communication slot utilizing WR-28 (Ka-band) waveguide with dimensions a ×
between anything (machine to machine, people to machine) or b = 7.112 mm × 3.556 mm. The waveguide length a is along
anybody (person to person) [1], [2]. Yet, the 5G technology the direction of slot length (SL), whereas the waveguide width
is not completely standardized, and several research activities is along the direction of slot width (SW). Hence, EM energy
are performed on different technologies, work on different fre- is coupled to the antenna surface along (SW) direction, which
quency bands in a range of frequencies between 5 and 70 GHz makes the electric field (E-field) polarized along the x-direction.
The optimized dimensions that maximize the directivity of the
Manuscript received February 19, 2018; revised April 21, 2018 and May antenna at resonance frequency can be summarized as follows:
27, 2018; accepted June 15, 2018. Date of publication June 19, 2018; date of
current version October 26, 2018. This work was supported by the Physical W 1 ≈ L1 ≈ λ (1)
Sciences Research Council (EPSRC) in the U.K. under Grant EP/R00711X/1.
(Corresponding author: Shaker Alkaraki.) λ
S. Alkaraki, A. S. Andy, Y. Gao, R. Donnan, and C. Parini are with the School K ≈ (2)
3
of Electronic Engineering and Computer Science, Queen Mary University of
London, London E1 4NS, U.K. (e-mail:, [email protected]; a.andy@ λ
qmul.ac.uk; yue.gao @qmul.ac.uk; [email protected]; c.g.parini@qmul. D≈ (3)
ac.uk).
5
K.-F. Tong is with the Department of Electronic and Electrical Engineer- where λ is the resonance wavelength. The dimensions of the
ing, University College London, London WC1E 7LE, U.K. (e-mail:, k.tong@ antenna are antenna length L = 20 mm, antenna width W =
ucl.ac.uk).
Z. Ying is with Sony Mobile Communications, Lund 22188, Sweden (e-mail:,
11 mm, slot length SL = 5.7 mm, rectangular cavity length
[email protected]). L1 = 10.8, plate thickness T = 6.4 mm, rectangular cavity
Digital Object Identifier 10.1109/LAWP.2018.2848912 width W 1 = 10.6 mm, rectangular cavity depth K = 3.2 mm,

1536-1225 © 2018 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution
requires IEEE permission. See http://www.ieee.org/publications standards/publications/rights/index.html for more information.
2052 IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 17, NO. 11, NOVEMBER 2018

Fig. 1. Schematics of Ant-1. (a) Top view, (b) cross section of front view, and
(c) fabricated prototype. Fig. 2. Schematic of Ant-2. (a) Cross section of front view, (b) fabricated
prototype, (c) bottom view, and (d) perspective bottom view with the con-
nector. W = 10 mm, T = 3 mm, GW = 1.5 mm,L = 17 mm, Gd = 1.9 mm,
h = 0.508 mm, W 1 = 9.5 mm, L1 = 7 mm, X = 3.2 mm, SW = 2.2 mm,
corrugation width C = 1.6 mm, corrugation depth D = SL = 6.5 mm , FW = 0.4 mm, FL = 5.4 mm, VR = 0.8 mm, and WP × WP =
2.1 mm, and the distance between cavity and corrugation X = 4.4 mm × 4.4 mm.
2.6 mm. The proposed antenna is commonly fabricated using a
metallic plate such as aluminium, and the corrugations and the
slot are drilled using computer numerical control (CNC) milling by a rectangular cavity and two corrugations. The commercial
and micromachining [7]. However, fabricating such a structure spray used for spray-coating is RS EMI/RFI 400 mL shield-
using CNC milling requires advanced machinery and skilled la- ing aerosol available commercially at a cost of ∼$50 USD per
bor and produces a relatively heavy metallic prototype. Hence, 400 mL can. Using the spray-coating technology to metalize the
to simplify the fabrication procedure, the antenna is 3-D printed proposed 3-D-printed antenna over the conventional metaliza-
in-house using a plastic material. The plastic material used is a tion techniques such as electroless plating has several advan-
transparent vero clear polyethylene, which enables fabrication tages such as a significant cost reduction as metallizing Ant-1
of a solid, rigid, and light prototype with an overall weight of and Ant-2 using the proposed technique costs less than $2 USD
1.35 g for Ant-1, which is ∼ 2–3 times lighter than an aluminum for low-volume prototyping since the aerosol has a large cover-
counterpart. A 3-D printer (Stratasys Objet30 Prime) is used to age area of ࣈ1.25 square m per 400 mL [9]. Hence, the overall
model the prototype as it prints with a layer thickness of 16 μm cost of fabrication of the antenna is very low if compared to
and a resolution of 100 μm. The prototype is 3-D printed with CNC milling or if electroless plating is chosen as an alternative
a clear finish, and the support material used in the 3-D printing metalization technique. Furthermore, using spray coating sim-
process is effectively removed by rinsing the prototype using plifies the fabrication procedure since the paint can be used at
pressurized water. The cost of vero clear transparent polyethy- room temperature, and the 3-D-printed prototypes do not need
lene is ≈ $650 USD per kg. Hence, the cost of material including any special pretreatment or posttreatment process before or after
the support material used to 3-D print the prototype is less than spraying. In addition, the proposed spray-coating technique can
$2 USD. Therefore, Objet30 offers cost-effective solutions to be applied directly to open geometry structures such as Ant-1
3-D print small structures with high resolution for millimeter- and Ant-2, but it is not possible to be applied to complex and
wave applications. Then, the 3-D-printed plastic prototype is closed geometry structures unless they are disassembled as in
metalized using a commercial and low-cost conductive spray [10]. One of the main disadvantages of the proposed metaliza-
paint that has superior scratch resistance characteristics with tion technique is that it is not possible to effectively control
very strong adhesive properties on a wide range of substrates its thickness uniformity across the whole structure. Therefore,
[9]. This conductive metal spray consists of metallic particles two coats were primarily applied to make sure that the interior
and acrylic resin along with compressed air. Two coats of the sides of the slot and corrugations are painted, which makes the
conductive spray paint are applied directly to the surface of the measured paint thickness on the surface of the prototype vary
antenna at room temperature with less than 1 h separation time in the range of 45–65 μm, which is higher than the skin depth
between them. The paint typically dries within 5 min, and it of copper ∼0.39 μm at 28.5 GHz. Hence, more expensive and
delivers maximum conductivity after 24 h [9]. complex plating techniques such as electroless plating [10] are
The schematic and the prototype of the second antenna recommended as a reliable alternative plating method, if the
(Ant-2) are shown in Fig. 2. Ant-2 is a smaller and a more thickness of the metalization layer is highly critical.
compact version of Ant-1 due to a special microstrip feeding
structure. Therefore, Ant-2 consists of two layers: The first layer
is a feeding structure designed to feed the second layer, which III. RESULTS AND ANALYSIS
is the 3-D-printed radiating structure. As shown in Fig. 2(c),
the feeding layer is designed to accommodate the PE 44489 A. Operating Principles
mini-surface mount package (SMP) connector. It consists of a Ant-1 and Ant-2 share the same operating principles with a
transmission line, mini-SMP pad (ground plane), and vias fabri- difference in the feeding method and the overall size, as Ant-2
cated on Rogers RO4003C substrate using conventional printed has a smaller and a more compact size than Ant-1 with an overall
circuit board fabrication techniques to guarantee good perfor- thickness of 3 mm instead of 6.4 mm for Ant-1. For Ant-1 and
mance. RO4003C is a hydrocarbon ceramic laminate, which has Ant 2, the EM energy is coupled to the surface of the 3-D-printed
a dielectric constant of 3.38 and thickness of h = 0.508 mm. The structure via a resonant slot, where the resonance frequency of
second layer of Ant-2 is the radiating structure, and it is similar the slot is inversely proportional to its length (SL ≈ λ2 ), as shown
to Ant-1 as it contains a 3-D-printed resonant slot surrounded in Fig. 3(a).
ALKARAKI et al.: COMPACT AND LOW-COST 3-D PRINTED ANTENNAS METALIZED USING SPRAY-COATING TECHNOLOGY 2053

Fig. 5. Power flow for Ant-1 at (xz) cutting plane and E-field distribution on
the surface of the antenna at 28.5 GHz. (a) Ant-1 power flow and (b) Ant-1
E-field.
Fig. 3. Effect of Ant-1 parameters on S11 . (a) SL, (b) T , (c) L1, and (d) K .
the performance of the antenna, as shown in Figs. 3(d) and 4(c).
For instance, the directivity sees a major improvement once K
is larger than quarter-wavelength (K = λ4 = 2.6 mm) with a di-
rectivity of 12.7 dBi, and the maximum directivity of 13.2 dBi
is obtained once K is in range of 3λ 10
≤ K ≤ 4λ10
, as shown in
Fig. 4(c); hence, in the proposed design of Ant-1, K is set to
3λ
10
= 3.2 mm at 28.5 GHz. In addition, the two corrugations
contribute to further ∼ 3 dBi improvement in the directivity of
the antenna due to excitation of the EM waves on the surface
of the antenna, as analyzed in [11]–[15]. Furthermore, the main
tradeoff of adding the cavity to the surface of Ant-1 with a depth
of K ≈ λ3 is that the overall thickness of the antenna has to be
increased. However, even with the cavity, Ant-1 is compact, and
it has a superior aperture efficiency performance compared to
Fig. 4. Effect of Ant-1 parameters on the directivity. (a) L1, (b) C, (c) K , and the dimensions of other antennas with similar feeding structure
(d) D. such as PE 9850-15 standard horn or if compared to 5G antennas
presented in [16] and [17].
The antennas offer high gain performance due to combined
B. Experimental and Numerical Results
radiation from the slot, the rectangular cavity, and the corruga-
tions. For instance, in Ant-1, the used WR28 waveguide operates Ant-1 and Ant-2 are numerically analyzed using CST Mi-
in the dominant transverse electric (TE10 ) mode, and it couples crowave Studio software, and the prototypes are fabricated
the EM energy to the surface of the antenna, where part of this and tested. Both structures were simulated as a perfect elec-
energy is radiated by the slot, and the rest of the coupled energy tric conductor (PEC) with an extremely smooth surface. The
is converted to a traveling wave that propagates in the x-direction reflection coefficient is experimentally measured using a vector
on the surface of the antenna and ultimately lost and leaked at network analyzer, whereas the gain of the antennas has been
antenna edges. Hence, Ant-1 has a directivity of 5.6 dBi once measured using the standard horn gain comparison method
the slot is solely present on the surface of the 3-D-printed struc- described in detail in [18]. The measured results show that
ture, and this is due to conventional space-wave radiation from Ant-1 resonates at 28.5 GHz with a measured bandwidth of
the slot. However, the directivity of the antenna is boosted by 2.5 GHz (8.8%) compared to a 2.3 GHz bandwidth in the
an extra ∼3 dBi once the corrugations are added, and further by simulation as shown in Fig. 6(a). The proposed antenna has
4.4 dBi once the rectangular cavity is added. In fact, both the a peak measured gain of 12.5 dBi at 28.7 GHz and a gain of
rectangular cavity and the corrugations excite the traveling wave 12.4 dBi at 28.5 GHz, as shown in Fig. 6(b), compared to a
on the surface of the antenna, resulting in an improvement in simulated gain of 13.2 dBi at 28.5 GHz. It is also observed,
the directivity and gain performance of the antenna resulting in from Fig. 6(b), that the measured gain is more than 10 dBi over
combined radiation in the boresight. Hence, the peak directivity the entire bandwidth of the antenna, and the gain is particularly
performance is obtained once the dimensions of the cavity and higher than 12 dBi over 800 MHz of the operating bandwidth.
the corrugations are optimized, as shown in Fig. 4. Moreover, Besides, Ant-1 has a half-power beamwidth (HPBW) of 30°
the reason behind the gain improvement in the proposed antenna with sidelobe level of –12 dB in the E-plane and HPBW of 54°
after the introduction of the rectangular cavity on the surface of in the H-plane and with front-to-back (F/B) ratio of 20.1 dB.
the antenna is the radiation caused by the cavity due to strong In addition, one of the main figures of merit of Ant-1 is
excitation and resonance of the dominant (TM12 ) mode inside its compact size and high measured aperture efficiency (Ae ).
the cavity, as shown in Fig. 5(b). Hence, the length of the cavity Ae is defined as the ratio of the maximum effective antenna
L1 has a major effect on the resonance frequency of the antenna, aperture area to its physical aperture area, and it determines
as shown in Fig. 3(c). Also, L1 has an effect on the directivity how efficiently the physical area of the antenna is utilized [19].
of the antenna as the peak in the directivity is obtained once L1 The outstanding Ae performance of Ant-1 is due to its high gain
is equivalent to λ. Furthermore, the depth of the cavity K affects performance compared to its small size. Ant-1 has an Ae of 70%
2054 IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 17, NO. 11, NOVEMBER 2018

Fig. 6. Measured and simulated results of Ant-1. (a) S11 , (b) gain, (c) E-plane, Fig. 8. Measured and simulated results of Ant-2. (a) S11 , (b) gain, (c) H-plane,
and (d) H-plane at 28.5 GHz. and (d) E-plane at 28.5 GHz.

TABLE I gain. Furthermore, fabrication tolerances and nonuniformity of

COMPARISON OF THE PROPOSED ANTENNAS PERFORMANCE WITH PE-HORN
AND 5G ANTENNAS AT 28.5 GHZ
the thickness of the paint might affect the S11 performance of the
antenna as the paint thickness was not counted for in the simula-
tion process, especially that the resonance frequency is sensitive
to the dimensions of the slot as shown in Fig. 3(a). However, the
surface roughness of the paint is hardly responsible for any gain
losses. The measured root mean square (RMS) roughness of the
paint at antenna surface is 4.92 μm along a path of 3 mm. How-
ever, it has been reported that a surface roughness of 25.9 μm
(RMS) has no effect on the gain of a Ku-band horn antenna at
15 GHz, whereas a measured ∼ 0.7 dB loss in the horn gain is
noticed once it has an extreme surface roughness of 39.7 μm
(RMS), which is noticeably far larger than the proposed antenna
surface roughness.
Furthermore, the measured and simulated reflection coef-
ficient of Ant-2 is shown in Fig. 8(a), and it shows that
the measured −10 dB bandwidth of the antenna is 3.7 GHz
(12.9%) that ranges from 26.9 to 30.6 GHz. Furthermore, Ant-2
has a simulated peak gain of 9.6 dBi at 29.5 GHz and a simu-
Fig. 7. Effect of conductivity on (a) the gain and (b) the S11 of Ant-1. lated gain of 9.4 dBi at 28.5 GHz. However, the measured peak
gain is 8 dBi at 28.5 GHz and 7.9 dBi at 29.5 GHz, and the
antenna measured gain is higher than 7.6 dBi over 2 GHz band-
at 28.5 GHz, and Ae approaches a peak of 71.7% at 28.7 GHz. width that ranges from 27.7 to 29.7 GHz, as shown in Fig. 8(b).
In comparison to an Ae of 59.5% for PE 9850-15 horn, 8.3% to The simulated and measured far-field radiation patterns of the
the 28.5 GHz tilted beam 5G antenna fed by the waveguide and antenna at 28.5 GHz are shown in Fig. 8. The proposed antenna
to an Ae of 30% at for 5G Franklin array has been presented has a measured HPBW of 60◦ in the H-plane and 53◦ in the E-
recently in [17]. A detailed comparison between the proposed plane, whereas the F/B ratio of the proposed antenna is 9.2 dB.
antennas performance with other antennas is shown in Table I. The discrepancies between the simulated and measured results
Furthermore, Ant-1 is compact once compared to the Ka-band of Ant-2, besides the conduction losses as discussed in Ant-1
PE 9850-15 horn as the horn is 7.4 times higher than Ant-1 with case, are due to insertion losses as well as matching losses due
an overall height of 4.5λ at 28.5 GHz as compared to only 0.61λ to mini-SMP connector soldering and misalignment with the
in case of Ant-1. Generally, the measured gain of Ant-1 is in 0.4 mm wide transmission line and fabrication tolerances of the
range of 0.7 to 2 dB lower than the simulated gain, as shown in feeding layer.
Fig. 6(b), whereas the measured bandwidth is 200 MHz wider
than the simulated one.
In fact, the lower gain and wider bandwidth are mainly due
to the conduction losses introduced by the paint as shown IV. CONCLUSION
in Fig. 7(a) and (b). The paint has a surface resistivity of This letter presents designs of efficient, high-gain, and com-
Rs = 0.3 − 0.7 / at 50 μm thickness [9], which makes pact 3-D-printed antennas metalized using low-cost EMI/RFI
its conductivity σ = 0.29 × 105 − 0.67 × 105 S/m. This con- spray-coating technology. Using EMI/RFI conductive paint re-
tributes to gain losses of 0.4∼ 0.8 dB at 28.5 GHz com- duces the production cost of the antenna and the complexity of
pared to PEC as the simulation results show in Fig. 7(a). the fabrication procedure while delivering suitable performance
Hence, at 28.5 GHz, the antenna simulated gain is 13.2 dBi at 28 GHz band. Finally, the novel fabrication method employed
for PEC, 12.7 dBi for σ = 0.67 × 105 S/m, and 12.5 dBi for in this investigation offers the proposed antenna to be a potential
σ = 0.29 × 105 S/m, which is 0.1 dBi higher than the measured candidate for low-cost 5G mm-wave applications.
ALKARAKI et al.: COMPACT AND LOW-COST 3-D PRINTED ANTENNAS METALIZED USING SPRAY-COATING TECHNOLOGY 2055

REFERENCES [10] M. D’Auria et al., “3-D printed metal-pipe rectangular waveguides,” IEEE
Trans. Compon., Packag. Manuf., Technol., vol. 5, no. 9, pp. 1339–1349,
[1] Ericsson.com, “Mobility report,” 2015. [Online]. Available: http://www. Sep. 2015.
ericsson.com/res/docs/2015/ericsson-mobility-report-june-2015.pdf. Ac- [11] S. Alkaraki, Y. Gao, and C. Parini, “High aperture efficient antenna at
cessed: Jan. 5, 2018. Ku band,” in Proc. Int. Workshop Electromagn.: Appl. Student Innov.
[2] NTT Docomo Inc., “NTT DOCOMO’s views on 5G,” 2014. [Online]. Competition, London, U.K., 2017, pp. 60–62.
Available: http://johannesbergsummit.com/wp-content/uploads/sites/6/ [12] D. Y. Na, K. Y. Jung, and Y. B. Park, “Transmission through an annular
2013/11/Nakamura-Johnnesberg-Summit-NTT-DOCOMOs-Views-on- aperture surrounded with corrugations in a PEC plane,” IEEE Antennas
5G_rev.pdf. Accessed: Jan. 5, 2018. Wireless Propag. Lett., vol. 14, pp. 179–182, 2015.
[3] S. F. Jilani and A. Alomainy, “Millimetre-wave T-shaped MIMO antenna [13] S. Alkaraki, Y. Gao, and C. Parini, “Dual-layer corrugated plate an-
with defected ground structures for 5G cellular networks,” Microw., An- tenna,” IEEE Antennas Wireless Propag. Lett., vol. 16, pp. 2086–2089,
tennas Propag., vol. 12, no. 5, pp. 672–677, 2018. 2017.
[4] FCC, “Use of Spectrum Bands above 24 GHz for Mobile Radio Services, [14] M. B. Diaz et al., “Dual-band low-profile corrugated feeder antenna,”
GN Docket No. 14-177, Notice of Proposed Rulemaking,” 15 FCC Record IEEE Trans. Antennas Propag., vol. 54, no. 2, pp. 340–350, Feb. 2006.
138A1. Oct. 23, 2015. [15] S. Alkaraki, Y. Gao, and C. Parini, “Small and high gain millimetre wave
[5] G. P. Le Sage, “3D printed waveguide slot array antennas,” IEEE Access, corrugated grooves antenna,” in Proc. IEEE Int. Symp. Antennas Propag.
vol. 4, pp. 1258–1265, 2016. USNC/URSI Nat. Radio Sci. Meeting, Vancouver, BC, Canada, 2015,
[6] J. A. Gordon et al., “An all-metal, 3-D-printed CubeSat feed horn: An pp. 2091–2092.
assessment of performance conducted at 118.7503 GHz using a robotic [16] J. S. Park, J. B. Ko, H. K. Kwon, B. S. Kang, B. Park, and D. Kim, “A
antenna range,” IEEE Antennas Propag. Mag., vol. 59, no. 2, pp. 96–102, tilted combined beam antenna for 5G communications using a 28-GHz
Apr. 2017. Band,” IEEE Antennas Wireless Propag. Lett., vol. 15, pp. 1685–1688,
[7] U. Beaskoetxea, S. Maci, M. Navarro-Cı́a, and M. Beruete, “3-D-printed 2016.
96 GHz bull’s-eye antenna with off-axis beaming,” IEEE Trans. Antennas [17] S. F. Jilani and A. Alomainy, “A multiband millimeter-wave 2-D array
Propag., vol. 65, no. 1, pp. 17–25, Jan. 2017. based on enhanced franklin antenna for 5G wireless systems,” IEEE An-
[8] Y. Tawk, M. Chahoud, M. Fadous, J. Costantine, and C. G. Christodoulou, tennas Wireless Propag. Lett., vol. 16, pp. 2983–2986, 2017.
“The miniaturization of a partially 3D printed quadrifilar helix antenna,” [18] C. A. Balanis,“ Antenna measurement,” in Antenna Theory, 3rd ed. Hobo-
IEEE Trans. Antennas Propag., vol. 65, no. 10, pp. 5043–5051, Oct. 2017. ken, NJ, USA: Wiley, 2005, ch. 17, sec. 4.2, pp. 1033–1034.
[9] RS EMI/RFI Shielding Aerosol with Bronze colour 400mL data [19] C. A. Balanis, “Aperture antennas,” in Antenna Theory, 3rd ed. Hoboken,
Sheet. [Online]. Available: https://docs-emea.rs-online.com/webdocs/ NJ, USA: Wiley, 2005, ch. 12, sec. 5.3, p. 682.
1583/0900766b8158361e.pdf. Accessed: Apr. 20, 2018.