materials

Article
Analysis on the Effects of the Human Body on the
Performance of Electro-Textile Antennas for Wearable
Monitoring and Tracking Application
Nurul Huda Abd Rahman 1,2, *, Yoshihide Yamada 1 and Muhammad Shakir Amin Nordin 2
1 Malaysia-Japan International Institute of Technology, Universiti Teknologi Malaysia,
Jalan Sultan Yahya Petra, Kuala Lumpur 54100, Malaysia; ndayamada@yahoo.co.jp
2 Faculty of Electrical Engineering, Universiti Teknologi MARA, Shah Alam 40450, Selangor, Malaysia;
shakiraminnordin@gmail.com
* Correspondence: nurulhuda0340@uitm.edu.my




Received: 21 March 2019; Accepted: 17 May 2019; Published: 19 May 2019


Abstract: Previous works have shown that wearable antennas can operate ideally in free space;
however, degradation in performance, specifically in terms of frequency shifts and efficiency was
observed when an antenna structure was in close proximity to the human body. These issues
have been highlighted many times yet, systematic and numerical analysis on how the dielectric
characteristics may affect the technical behavior of the antenna has not been discussed in detail. In this
paper, a wearable antenna, developed from a new electro-textile material has been designed, and the
step-by-step manufacturing process is presented. Through analysis of the frequency detuning effect,
the on-body behavior of the antenna is evaluated by focusing on quantifying the changes of its input
impedance and near-field distribution caused by the presence of lossy dielectric material. When the
antenna is attached to the top of the body fat phantom, there is an increase of 17% in impedance,
followed by 19% for the muscle phantom and 20% for the blood phantom. These phenomena correlate
with the electric field intensities (V/m) observed closely at the antenna through various layers of
mediums (z-axis) and along antenna edges (y-axis), which have shown significant increments of 29.7%
in fat, 35.3% in muscle and 36.1% in blood as compared to free space. This scenario has consequently
shown that a significant amount of energy is absorbed in the phantoms instead of radiated to the
air which has caused a substantial drop in efficiency and gain. Performance verification is also
demonstrated by using a fabricated human muscle phantom, with a dielectric constant of 48, loss
tangent of 0.29 and conductivity of 1.22 S/m.

Keywords: wearable devices; on-body antenna; wireless body area network; wireless sensing; smart
material; wearable antenna

1. Introduction
Wearable antennas are highly in demand to support various wearable technologies due to the
advantage of being flexible, lightweight and easily integrated with garments. Through wearable
technology, a wireless body sensing system is now able to support various Internet of Things (IoT)
applications such as for high speed communication, tracking, health monitoring and radio-frequency
identification (RFID) [1–3], as illustrated in Figure 1. In wearable devices, antenna is important
to support many wireless communication aspects such as for in-body communication, on-body
communication and off-body communication [4]. Many considerations should be taken into account
when designing a wearable antenna such as interconnectivity issues, consistency of performance and
comfortability of users [5–8]. Due to the importance of these issues in antenna designing, this paper

Materials 2019, 12, 1636; doi:10.3390/ma12101636 www.mdpi.com/journal/materials

copper powder [9]. For these structures, antenna substrate is normally represented by non-
conductive material such as cotton, polyester or woven, and the integration of copper as the radiating
element is done through various techniques such as embroidery, manual sewing, ironing, gluing etc.
[10]. The main drawback of the conventional or off-the-shelf conductive fabric is the sustainability
Materials 2019, 12, 1636
issue, where the antenna can be easily detached from the fabric after being washed or worn for many 2 of 17
Materials 2019, 12, x FOR PEER REVIEW 2 of 17
times [11]. In this research, a customized and more structurally practical conductive fabric as shown
in Figure 2 wasa acustomized
proposed
proposed developed
customized and analyzed.fabric,
electro-textile
electro-textile The electrical
and
fabric, characteristics
anditsitsperformance
performance of this is
consistency
consistencyfabric havewith
isanalyzed
analyzed been
withrespect
respect
determined through
totoon-body
on-bodyfactors.transmission
factors. line method; [12] however, the on-body performance has yet to be
verified. Conventional textile-based wearable antennas are normally fabricated on regular fabrics, and
integrated with copper-based radiating materials such as copper foil, copper thread, copper tape or
copper powder [9]. For these structures, antenna substrate is normally represented by non-
conductive material such as cotton, polyester or woven, and the integration of copper as the radiating
element is done through various techniques such as embroidery, GPSmanual
tracking antennas
sewing, ironing, gluing etc.
[10]. The main drawback of the conventional or off-the-shelf conductive fabric is the sustainability
issue, where the antennaCommunication
can be easily detached from the fabric after being washed or worn for many
antennas for e.g.,
times [11]. In this research,
wearable a customized
sensors, RFID, and more structurally practical
Wearable conductive
sensor e.g., heart rate fabric as shown
GPS data etc.
in Figure 2 was developed and analyzed. The electrical characteristics ofmonitor
monitor, blood pressure this fabric have been
determined through transmission line method; [12] however, the on-body performance has yet to be
verified.
Figure Figure 1. Examples
1. Examples of wearable
of wearable antennas antennas integrated
integrated with wearable
with wearable electronic electronic
devicesdevices
for many for many
applications.
applications. ThroughThrough
wearablewearable technology,
technology, variousvarious
InternetInternet of Things
of Things (IoT) applications
(IoT) applications can be can be
supported such as healthcare monitoring via wearable sensors, human tracking through
supported such as healthcare monitoring via wearable sensors, human tracking through body-centric body-centric
radio-frequency
radio-frequency identification
identification (RFID),(RFID), sportsmen
sportsmen or patient
or patient tracking
tracking via a wearable
via a wearable global positioning
global positioning
system (GPS) receiver and
system (GPS) receiver and many more.many more. GPS tracking antennas

Conventional textile-based wearable antennas are normally fabricated on regular fabrics, and
Communication
integrated with copper-based radiating materials
antennas for e.g., such as copper foil, copper thread, copper tape or
wearable sensors, RFID,
copper powder [9]. For these structures, antenna
GPS data etc.
substrate is normallyWearable sensor e.g.,
represented heart rate
by non-conductive
monitor, blood pressure monitor
material such as cotton, polyester or woven, and the integration of copper as the radiating element
is done through various techniques such as embroidery, manual sewing, ironing, gluing etc. [10].
The main drawback
Figure of the
1. Examples ofconventional or off-the-shelf
wearable antennas integrated conductive
with wearable fabric is the devices
electronic sustainability
for manyissue,
whereapplications.
the antennaThrough
can be wearable
easily detached
technology,from the fabric
various after
Internet of being
Things washed or worn can
(IoT) applications for many
be
supported
times [11]. such
In this as healthcare
research, monitoring
a customized andvia wearable
more sensors,practical
structurally human tracking through
conductive body-centric
fabric as shown in
Figureradio-frequency
2 was developed identification (RFID),
and analyzed. Thesportsmen
electricalorcharacteristics
patient trackingofvia a wearable
this fabric have global
beenpositioning
determined
throughsystem (GPS)
transmissionreceiver
line and many
method; more.
[12] however, the on-body performance has yet to be verified.
Figure 2. Warp and weft of the manufactured electro-textile fabric. The conducting part was made
from a mixture of copper and polyester threads, with the weight ratio of 83% and 17% for copper and
polyester, respectively.

In practice, wearable antennas are designed to be integrated with wearable devices and to be
worn on the human body. Thus, the effects of the human body on antenna technical performance are
significant and must be considered [4,13,14]. Previously, detuning effects were observed when an
antenna was in close proximity to the human body; however, quantitative analysis in relation to
impedance change and near-field distribution was not shown [15]. In order to minimize the detuning
effects, a slotted planar inverted-F antenna (PIFA) was proposed [16] with a very broad bandwidth
as the main feature, but in that particular work, measurement with the human body was not
Figure2.2.Warp
Figure Warpandandweftweftofofthe
the manufacturedelectro-textile
manufactured electro-textilefabric.
fabric.The
Theconducting
conductingpart
partwas
wasmade
made
conducted. Numerical analysis of input impedance and current distribution with respect to the
froma amixture
from mixtureofofcopper
copperandandpolyester
polyesterthreads,
threads,with
withthe
theweight
weightratio
ratioofof83%
83%andand17%
17%for
forcopper
copperand
and
human body’s presence were performed theoretically by [17] using Green’s function in method of
polyester,
polyester, respectively.
respectively.
moment (MoM) code. Nevertheless, the analysis was limited to dipole antenna, and not specifically
validated for textile-based
InInpractice,
practice,wearable materials.
wearable antennas
antennasSimilarly,
are as shown
aredesigned
designed totobebeby [18], thewith
integrated
integrated effect
with of antenna-body-
wearable
wearable devices
devicesandandtotobebe
worn
wornon onthe thehuman
humanbody. body. Thus,
Thus, thethe effects
effects of
ofthe
thehuman
humanbody bodyon onantenna
antennatechnical
technical performance
performance are
are significant and must be considered [4,13,14]. Previously, detuning
significant and must be considered [4,13,14]. Previously, detuning effects were observed when effects were observed whenan
an antennawas
antenna wasininclose
closeproximity
proximityto to the
the human
human body;
body; however,
however, quantitative
quantitative analysis
analysisininrelation
relationtoto
impedance
impedance change and near-field distribution was not shown [15]. In order to minimize thedetuning
change and near-field distribution was not shown [15]. In order to minimize the detuning
effects,
effects,a aslotted
slottedplanar
planar inverted-F
inverted-F antenna
antenna (PIFA) was
(PIFA) proposed
was proposed [16][16]
with a very
with broad
a very bandwidth
broad as
bandwidth
the
as main feature,
the main but in but
feature, that in
particular work, measurement
that particular with the human
work, measurement with the body was not
human conducted.
body was not
Numerical
conducted. Numerical analysis of input impedance and current distribution with respectbody’s
analysis of input impedance and current distribution with respect to the human to the
presence were performed
human body’s presence were theoretically
performed by [17] using Green’s
theoretically function
by [17] using in method
Green’s of moment
function (MoM)of
in method
code.
moment Nevertheless,
(MoM) code. theNevertheless,
analysis was the limited to dipole
analysis antenna,
was limited and not
to dipole specifically
antenna, and notvalidated for
specifically
validated for textile-based materials. Similarly, as shown by [18], the effect of antenna-body-
Materials 2019, 12, 1636 3 of 17

textile-based materials. Similarly, as shown by [18], the effect of antenna-body-separation distance

was presented, yet, the impedance change was not discussed numerically in terms of its magnitude
or trends. Furthermore, the antennas used were not based on either planar structure or full-textile
configuration, similar to the works done by [19–21]. The analysis of in-body and on-body antenna
through in-pig measurement setup has been demonstrated in [4], but the focus of the paper was
on miniaturization of the antenna design, which was not restricted to textile-based materials thus,
low-loss robust substrate e.g., RO5880 was used. The most related work was shown in [22], where the
comparison of textile antenna efficiency and gain were done with respect to power losses and field
intensity inside the body. However, as concluded by the authors, there were some inconsistencies of
test environment that should be taken into account when performing the measurement e.g., the antenna
should be positioned on the same human body at the exact same location. Therefore, as proposed in
this article, in order to reduce the uncertainties, a model of human body parts must be fabricated with
respect to the actual dielectric properties to ensure a consistent test environment. Hence, systematic
and quantitative analysis can be done accurately. Several researchers have done on-body analysis by
using human phantoms but the main subjects of their works were different, and mainly focused on
ways to enhance the antenna performance such as miniaturizing the structure, improving the gain or
reducing the specific absorption rate (SAR) of the antenna [23].
In this paper, detailed analysis was carried out by studying the antenna characteristics based on
phantoms of human body parts with various dielectric constants, εr and conductivity, σ depending
on the actual dielectric properties of the human body parts. Here, the behavior of electric field and
magnetic surface current that penetrates through a high-loss medium such as a human phantom were
demonstrated and numerically analyzed in terms of the near-field magnitude. To demonstrate the
effects of on-body condition to the electro-textile (e-textile) antenna performance, a wearable antenna
which was designed at 1.575 GHz frequency, was used. The frequency was chosen based on the
allocated spectrum for GPS tracking devices, which can be implemented for many purposes such as
patient or pet-tracking, health monitoring or locating military personnel during search and rescue
missions. All simulation works were done by using a comprehensive 3D electromagnetic software
called CST, and verification of results was done through bench and anechoic chamber measurements.
A gel-based human muscle phantom was fabricated with εr and σ similar to human muscle tissue
to demonstrate the on-body behavior. Human muscle was chosen because the conductivity σ and
dielectric constant εr were considered as average, as compared to other human body parts such as fat
(lowest εr ) and blood (highest εr ). Furthermore, it makes up the largest proportion of the human arms.

2. Methods and Processes

Figure 3 shows the overall flow of the work, starting with fabrication and characterization of
electro-textile materials, as done in [12]. Then, human muscle phantom was fabricated and the
dielectric properties such as loss tangent (tanδ), dielectric constant (εr ) and conductivity (σ) were
measured. The measured and calculated parameters were used in the design and simulation phase
to ensure optimum antenna performance in ideal conditions. To verify the performance, an electro-
textile antenna was fabricated based on its optimum design in ideal conditions. Measurement was
conducted and important parameters such as return loss (S11 ), impedance (Zin ) and radiation pattern
were recorded and analyzed. Based on the experimental results, numerical data were analyzed in
depth based on the following scenarios; ideal (free space) and on-body (with a human phantom,
for various phantom properties). By observing the near-field data (field distribution) and surface
current with respect to impedance change, clear correlation and evidence of a significant change to
antenna performance were shown. These findings may contribute towards more accurate wearable
antenna designing in the future.
Materials 2019, 12, 1636 4 of 17
Materials 2019, 12, x FOR PEER REVIEW 4 of 17

Fabrication of electro-textile material Determination of textile properties (, r and tan)

Measurement of phantom
Fabrication of human phantom
properties (r and )

Design and simulation of on-body electro-textile antenna

Fabrication of electro-
Analysis and measurement textile antenna

Figure3.3.AAflowchart
Figure flowchartshowing
showingthe theoverall
overallprocess
processofoffabrication,
fabrication,characterization,
characterization,design,
design,simulation,
simulation,
measurement and analysis of the electro-textile antenna performance in on-body
measurement and analysis of the electro-textile antenna performance in on-body conditions.conditions.

3. Characterization, Design and Fabrication of Wearable Antenna

3. Characterization, Design and Fabrication of Wearable Antenna
In this section, the design concept of the basic planar rectangular electro-textile antenna is shown.
In this section, the design concept of the basic planar rectangular electro-textile antenna is
There are three sub-sections; production process of e-textile, determination of dielectric and electrical
shown. There are three sub-sections; production process of e-textile, determination of dielectric and
properties, and design of e-textile antenna.
electrical properties, and design of e-textile antenna.
3.1. Electro-Textile Production
3.1. Electro-Textile Production
Electro-textile was manufactured through integration of conductive and non-conductive yarns.
Electro-textile
In this was manufactured
project, conductive through integration
yarns were constructed of conductive
first by twisting and non-conductive
copper threads yarns.
(0.14mm in diameter)
In this
with project,threads
polyester conductive yarns
by using were constructed
a hollow-spindle machinefirstasby twisting
shown copper
in Figure threads
4a. After (0.14mm
that, in
the process
diameter) with polyester threads by using a hollow-spindle machine as shown in Figure
continued with the production of e-textile fabric by using the ‘plain’ weaving technique. In the weaving 4a. After
that, the Sulzer
process, processTextil
continued
G6300 with
rapierthe production
weaving machine of e-textile fabric by using
(Sulzer, Winterthur, the ‘plain’
Switzerland) weaving
as shown in
technique. In the weaving process, Sulzer Textil G6300 rapier weaving machine (Sulzer, Winterthur,
Figure 4b was used. The manufactured e-textile fabric consists of 25 wefts/inch in a horizontal direction
Switzerland) as shown
and 80 warps/inch in Figure
in a vertical 4b waswhich
direction, used. was
Theselected
manufactured
based on e-textile fabric setting
the optimum consistsandof the
25
wefts/inch
machine’s in a horizontal
limitations. direction
In this fabricand 80 warps/inch
construction, in a vertical
conductive yarns direction,
were usedwhich
as the was
wefts,selected
whilst
based on the optimum setting and the machine’s limitations. In this fabric construction,
non-conductive threads (polyester) were in the warp direction as shown previously in Figure 2. conductive
yarns were used as the wefts, whilst non-conductive threads (polyester) were in the warp direction
as shown previously in Figure 2.
Materials 2019, 12, 1636 5 of 17
Materials 2019, 12, x FOR PEER REVIEW 5 of 17

Figure 4. Electro-textile production process: (a) Spindle-spinning process for yarn formation; (b)
Figure 4. Electro-textile production process: (a) Spindle-spinning process for yarn formation;
Weaving process for e-textile formation.
(b) Weaving process for e-textile formation.

3.2. Electrical
3.2. Electrical and Dielectric
and Dielectric Properties
Properties
ThereThere
are two areelements
two elements in theofdesign
in the design textile of textile conductive
antenna; antenna; conductive
textile as thetextile as the
radiating radiating
element
element and non-conductive textile as the non-radiating (substrate) element.
and non-conductive textile as the non-radiating (substrate) element. For the radiating element, electrical For the radiating
element, (σ)
conductivity electrical
is veryconductivity (σ) is very
important, whereas, forimportant, whereas,
the substrate, for the
dielectric substrate,
properties dielectric
such properties
as dielectric
constant (εr ) and loss tangent (tanδ) play significant roles. Table 1 shows the properties of electro-textile the
such as dielectric constant (εr ) and loss tangent (tanδ) play significant roles. Table 1 shows
properties
fabricated in this ofpaper.
electro-textile
Electricalfabricated
conductivity in this paper.
(σ) and the Electrical conductivity
fabric’s thickness (t) are(σ)
twoand the fabric’s
important
thickness
parameters that(t)shall
are two important
influence parameters
the antenna performance. The σ was
that shall influence thecalculated
antenna performance.
from resistance The(R)σ was
value, obtained through current-voltage (I-V) probe method [24] and verified through propagation and
calculated from resistance (R) value, obtained through current-voltage (I-V) probe method [24]
verified
constant through propagation
(γ) parameter obtained from constant () parameter
the newly proposed obtained
strip linefrom the newly proposed
measurement technique strip
[12]. line
measurement
The technique wastechnique
developed[12]. Theontechnique
based the two-portwas transmission
developed based line,on thethe
and σ was derived
two-port transmission
using line,
and the
Equation  was derived
(1), whereby ε is the using equation
material’s 1, whereby
permittivity and µisisthethe material’s
material’s permittivity and µ is the
permeability:
material’s permeability:
ω3 ε2 µ
σ= 2 2  3 2  (1)
α=h (2α2 + 2ω2 εµ) (1)
 2 h 2 (2 2 + 2 2 )
Table 1. The physical and electrical properties of fabricated e-textile.

TableProperties Values of fabricated e-textile.

1. The physical and electrical properties
Conductivity (σ) 2.2 × 104 S/m
Properties Values
Thickness (t) 0.48 mm
Conductivity (
Fabric composition) 2.2
83% Copper + 17% × 10 4 S/m
Polyester
Thickness (t) 0.48 mm
Fabric composition 83% Copper + 17% Polyester
Although the main component in the e-textile was copper, the calculated σ was lower than the σ
Although
of pure copper due tothethe
main component
presence in the e-textile
of non-conductive was copper,
material such asthe calculated
polyester insideσ was lower than the
the conductive
 of pure copper due to the presence of non-conductive material
thread. This has resulted in a less ideal performance of the textile-based antenna as comparedsuch as polyester inside
to the
conductive antenna,
the full-metallic thread. This which haswill
resulted in a less
be discussed ideal
later performance
in this paper. Theofcomposition
the textile-based antenna
of fabric was as
compared to the full-metallic antenna, which will be discussed later in this
calculated based on mass per unit area, according to ASTM D3776 [25]. In this research, polyester paper. The composition
fabricofhaving
fabric was
low εcalculated based on mass per unit area, according to ASTM D3776 [25]. In this research,
r and moderate tanδ was used as the substrate. This polyester material was chosen
polyester fabric having low
due to its suitability and comfortability εr andfeatures
moderate tanδ
to be wasforused
used aswear.
daily the substrate. This characteristics
The dielectric polyester material
was chosen due to its suitability and comfortability features to be used
of the polyester substrate were determined prior to performing antenna design and simulation for daily wear. The dielectric
to
ensure accurate analysis. Both εr and tanδ were calculated through the coaxial probe measurement and
characteristics of the polyester substrate were determined prior to performing antenna design
simulation
technique to ensure
by using accurate
the Keysight analysis.
85070 Both probe
dielectric εr andkittanδ were calculated
(Keysight through
Technologies, theRosa,
Santa coaxial
CA,probe
measurement technique by using the Keysight 85070 dielectric probe
USA), which consists of a vector network analyzer (VNA), a coaxial probe and software to displaykit (Keysight Technologies,
Santa Rosa,
the complex CA, USA),
permittivity datawhich consists
obtained of aS-parameter
through vector network analyzermeasured
conversion, (VNA), abycoaxial probe and
the network
software to display the complex permittivity data obtained through S-parameter
analyzer [26]. The measurement setup is illustrated in Figure 5a. A customized jig with multiple holes conversion,
measured
as shown by the
in Figure network
5b was analyzer
developed [26]. The
to allow measurement
the fabric sample to setup is illustrated
be stretched in Figure
uniformly on the5a. A
customized jig with multiple holes as shown in Figure 5b was developed
surface to ensure consistent and accurate readings. Furthermore, the height of fabric sample from to allow the fabric sample
the table’s surface must be more than λ/4 to avoid multiple reflection, hence, an additional feature as the
to be stretched uniformly on the surface to ensure consistent and accurate readings. Furthermore,
height of fabric sample from the table’s surface must be more than /4 to avoid multiple reflection,
  r =  ' − j " (2)

where the ’ and ” reflect the characteristics of a lossy material in terms of its tan, as shown by
equation 3 [26]:
Materials 2019, 12, 1636 6 of 17
"
tan  = ' (3)

shown in Figure 5c was included in this jig in which the height of sample can be adjusted according to
Based on frequency,
its operational the measurement
fo . at 1.575 GHz, the polyester substrate had a r of 1.36 and a tan of
0.031, with a thickness, h of 1.4 mm.
Keysight PNA-L
N5232A

Port 1

Keysight 85070
SUT Dielectric Probe

/4 Holding jig

(a)

Dielectric probe
connected to VNA

Adjustable
A holder to hold height
the fabric

(b) (c)

Figure5.5.Measurement
Figure Measurement of dielectric properties:
of dielectric (a) Equipment
properties: arrangement
(a) Equipment which consists
arrangement which ofconsists
dielectric
of
probe, network
dielectric analyzer
probe, networkandanalyzer
holding jig;
and(b) Customized
holding jigCustomized
jig; (b) with holes and
jigholders to ensure
with holes and consistent
holders to
readings and to control
ensure consistent fabricand
readings stiffness; (c) Adjustable
to control height feature
fabric stiffness; to allowheight
(c) Adjustable measurement
feature of
to many
allow
samples at various operational frequencies. VNA, vector network analyzer; SUT, sample-under-test.
measurement of many samples at various operational frequencies. VNA, vector network analyzer;
SUT, sample-under-test.
In the εr measurement, the ε’ (known as the real part of the εr ) and the ε” (known as the
imaginary
3.3. Antennapart of the
Design εr ) were
(Ideal measured
and On-Body by contacting the dielectric probe to the flat surface of the
Conditions)
sample-under-test (SUT). The complex permittivity is defined as follows:
For analysis, a rectangular antenna with an edge-fed transmission line was chosen. This antenna
was chosen due to the symmetrical geometry, which
εr = ε0 − makes
jε” it easier to design, fabricate and analyze.
(2)
The performance of the designed antenna should be improved further; however, in this paper, the
focus is
where onεstudying
the ’ and ε” reflect
and understanding the technical
the characteristics behavior
of a lossy of the
material antenna
in terms when
of its located
tanδ, near by
as shown the
human body.
Equation (3) [26]:Thus, based on the electrical properties of conductive material and the dielectric
characteristics of the polyester substrate, atan 1.575 ε” electro-textile antenna was designed. The
δ =GHz (3)
antenna consists of a rectangular radiating elementεwith a full ground plane, made by using the
0

Based on
fabricated the measurement
conductive at 1.575The
textile material. GHz, the polyester
geometry and fabricated had a εr of
substrate structure of1.36
the and a tanδ of
electro-textile
0.031,
antennawith
is ashown
thickness, h of 1.4
in Figure mm.integration of the top radiator (conductive textile) and the bottom
6. The
ground plane to the dielectric layer (polyester) was performed by using a fabric glue that was heated
3.3. Antenna Design (Ideal and On-Body Conditions)
up by using a hot-compress machine during assembly to ensure strong attachment of fabric layers.
For analysis, a rectangular antenna with an edge-fed transmission line was chosen. This antenna
was chosen due to the symmetrical geometry, which makes it easier to design, fabricate and analyze.
The performance of the designed antenna should be improved further; however, in this paper, the focus
is on studying and understanding the technical behavior of the antenna when located near the human
body. Thus, based on the electrical properties of conductive material and the dielectric characteristics
of the polyester substrate, a 1.575 GHz electro-textile antenna was designed. The antenna consists of a
rectangular radiating element with a full ground plane, made by using the fabricated conductive textile
material. The geometry and fabricated structure of the electro-textile antenna is shown in Figure 6.
The integration of the top radiator (conductive textile) and the bottom ground plane to the dielectric
Materials 2019, 12, x FOR PEER REVIEW 7 of 17

Materials 2019, 12, 1636 7 of 17

layer (polyester) was performed by using a fabric glue that was heated up by using a hot-compress
Materials 2019, 12, x FOR PEER REVIEW 7 of 17
machine during assembly to ensure strong attachment of fabric layers.

Figure 6. Geometry and structure of 1.575 GHz rectangular electro-textile antenna in the free space
condition (Case A), with W/L ratio of 1.1 for the conductive layer and Ws /Ls ratio of 0.9 for the non-
conductive layer. These values have been optimized based on the free space condition.

In wearable GPS tracking devices, electro-textile antenna should operate on the human body
and be Figure
Figure 6. Geometry
6. Geometry
specifically placedandand structure
atstructure
any of of
body 1.575
1.575 GHz
locationsGHz rectangular
rectangular
such electro-textile
electro-textile
as the arm, antenna
wrist,antenna
shoulder in inorthe
the freefree
leg. space
space
For that
condition
condition (Case
(Case A),
A), with
with W/LW/L ratio
ratio of
of 1.1
1.1 for
for the
the conductive
conductive layer and
and WW
reason, the effects must be extrapolated and analyzed. In order to understand the impact of on-body /L
ss/L s sratio
ratioof 0.9
of 0.9for the
for non-
the
conductive layer.
non-conductive layer.These
These values
valueshave
have been
been optimized
optimized based
basedononthe free
the freespace
spacecondition.
condition.
condition to a wearable antenna, two situations were simulated, demonstrated and studied in this
article:
In In
wearable
wearable GPS tracking
GPS trackingdevices, electro-textile
devices, antenna
electro-textile should
antenna operate
should on theon
operate human body and
the human body
1. Case A: Antenna in free space (without phantom)—Figure 6.
beand
specifically placed at
be specifically any body
placed at anylocations such as the
body locations sucharm, wrist,
as the shoulder
arm, or leg. Fororthat
wrist, shoulder leg.reason,
For that
2. Case B: Antenna on a human phantom—Figure 7.
thereason,
effects the
must be extrapolated
effects and analyzed.
must be extrapolated andInanalyzed.
order to understand the impact ofthe
In order to understand on-body
impactcondition
of on-body
For Case B, a rectangular shaped dielectric material (phantom) that has similar properties with
to condition
a wearabletoantenna, two situations were simulated, demonstrated and studied
a wearable antenna, two situations were simulated, demonstrated and studied in this in this article:
the corresponding human body part was attached as another structure to the ideal antenna. Here,
article:
important
1. Caseproperties
A: Antenna ofin
body
freetissues which are
space (without r and  were set.6.For this simulation, the r and  at
phantom)—Figure
1. Case A: Antenna in free space (without phantom)—Figure 6.
1.575
2. GHz Case were variedon
B: Antenna forathree
human cases; body fat (r = 12,
phantom—Figure 7.  = 0.17 S/m), muscle (r = 48,  = 1.22 S/m)
2. Case B: Antenna on a human phantom—Figure 7.
and blood (r = 64,  = 1.9 S/m) [27].
For Case B, a rectangular shaped dielectric material (phantom) that has similar properties with
the corresponding human body part was attached as another structure to the ideal antenna. Here,
important properties of body tissues which are r and  were set. For this simulation, the r and  at
1.575 GHz were varied for three cases; body fat (r = 12,  = 0.17 S/m), muscle (r = 48,  = 1.22 S/m)
and blood (r = 64,  = 1.9 S/m) [27].

Figure
Figure7.7.Configuration
Configurationfor
foron-body
on-bodycondition
conditionanalysis
analysis(Case
(CaseB).
B).During
Duringsimulation,
simulation,the
theantenna
antennaisis
attached to the human phantom with various dielectric constant,  r and conductivity,  to represent
attached to the human phantom with various dielectric constant, εr and conductivity, σ to represent the
the exact
exact properties
properties of human
of human body
body fat,fat, muscle
muscle andand blood.
blood.

4. Fabrication
For Caseof B, Human Muscle
a rectangular Phantom
shaped dielectric material (phantom) that has similar properties with
the corresponding human body part wascondition
Figure 7. Configuration for on-body attachedanalysis (Casestructure
as another B). Duringtosimulation,
the ideal the antennaHere,
antenna. is
A human
attachedphantom is a phantom
to the human physical with
model having
various similar
dielectric properties
constant, r andasconductivity,
human biological tissues
 to represent
important properties of body tissues which are εr and σ were set. For this simulation, the εr and σ at
[28]. Forthe
antenna performance
exact properties of humanvalidation,
body fat,measurement
muscle and blood.of scattering parameters (S11) for on-body
1.575 GHz were varied for three cases; body fat (ε = 12, σ = 0.17 S/m), muscle (εr = 48, σ = 1.22 S/m)
condition can be done by using a real human bodyr part; however, due to the complexity of antenna
and blood (εr = 64, σ = 1.9 S/m) [27].
three-dimensional (3D) radiation
4. Fabrication of Human Muscle Phantom pattern measurement, phantom fabrication was proposed to
represent a human-like
4. Fabrication of Human material. Phantom
The 3D radiation pattern measurement for on-body wearable
A human phantomMuscleis a physical model having similar properties as human biological tissues
antenna must be conducted inside an anechoic chamber. Due to radiation, health concerns and
[28]. For antenna
A human phantom performance validation, measurement of scattering parameters (S11)tissues
for on-body
mounting complexity, itisisa physical modelfor
not advisable having similar
a real humanproperties
body to as be
human biological
included as part [28].
of the
Forcondition
antenna can be done
performance by using a
validation,real human body
measurement part;
of however,
scattering due to the
parameters complexity
(S ) for of antenna
on-body
measurement setup. In this paper, in order to validate the analysis, human phantom 11 fabrication was
three-dimensional
condition can be done (3D) radiation
by using a real pattern
human measurement,
body phantom
part;proportion
however, due tofabrication
the was of
complexity proposed
antenna to
simplified to human muscle which makes up the largest of the human arm. Thus, based
represent a
three-dimensional human-like material.
(3D) radiation patternThe 3D radiation
measurement, pattern
phantomwas measurement
fabrication for on-body
wasFabrication
proposed towas wearable
represent
on the specified parameters, a suitable mixing ratio of chemicals suggested. done
a antenna must
human-like be conducted
material. The 3D inside an
radiation anechoic
pattern chamber. for
measurement Due to radiation,
on-body wearablehealth
antennaconcerns
must beand
several times to ensure that the required dielectric properties were achieved. Finally, after several
mounting complexity, it is not advisable for a real human body to be included as part of the
measurement setup. In this paper, in order to validate the analysis, human phantom fabrication was
simplified to human muscle which makes up the largest proportion of the human arm. Thus, based
on the specified parameters, a suitable mixing ratio of chemicals was suggested. Fabrication was done
several times to ensure that the required dielectric properties were achieved. Finally, after several
 Materials 2019, 12, 1636 8 of 17

conducted inside an anechoic chamber. Due to radiation, health concerns and mounting complexity,
it is not advisable for a real human body to be included as part of the measurement setup. In this
paper, in order to validate the analysis, human phantom fabrication was simplified to human muscle
which makes up the largest proportion of the human arm. Thus, based on the specified parameters,
a suitable mixing ratio of chemicals was suggested. Fabrication was done several times to ensure that
Materials 2019, 12,dielectric
the required x FOR PEER REVIEW were achieved. Finally, after several trials, a good muscle phantom
properties 8 of 17
was developed by using the chemical composition as shown in Table 2. After the fabrication, the εr of
trials, a good muscle
the phantom phantom was
was determined developed
by using by using
the coaxial thetechnique
probe chemical composition as shown
as shown in Figure 8a. in Table
The data
2.from
Afterfive
themeasurements
fabrication, the r of the phantom was determined by using the coaxial probe technique
were taken and the average εr and σ were calculated, as shown in Figure 8b.
asThe
shown in are
Figure 8a. Theε data from five measurements were taken and the average εr and σ were
results as follows; r = 48.2, σ = 1.23 and tanδ = 0.292. There were only slight variations between
calculated,
the measured as shown
data asincompared
Figure 8b.toThe
the results are as
theoretical withrthe
follows;
values, = 48.2,  = 1.23standard
calculated and tandeviation
= 0.292. There
(σx ) of
were only slight variations between
0.3 for εr , 0.12 for σ and 0.03 for tanδ. the measured data as compared to the theoretical values, with
the calculated standard deviation (x) of 0.3 for r, 0.12 for  and 0.03 for tan.
Table 2. Chemical composition of the human muscle phantom.
Table 2. Chemical composition of the human muscle phantom.
Ingredients Purpose Quantity
Ingredients Purpose Quantity
Deionized water Main component 420 mL
Deionized water Main component 420 mL
Agar Polyethylene powder 20 g
Agar
Sodium dehydroacetate
Polyethylene powder
As a preservative
20 g0.25 g
Sodium dehydroacetate
Xanthan gum As As
a preservative
a thickener 0.25 6.25
g g
Sodium Xanthan
chloridegum
(NaCl) As a thickener
To control the σ of phantom 6.25 g4.1 g
Sodium chloride
Polyethylene (NaCl) To
powder Tocontrol theεof
controlthe r ofphantom
phantom 4.1 g29 g
Polyethylene powder To control the r of phantom 29 g

50 3
Dielectric probe
connected to VNA 49
Dielectric constant, r

Conductivity, 
48 2

47

46 1
Human muscle
45 r
Er Mean
r (mean)
Er
phantom

sigma Mean
(mean)
sigma
44 0
1 2 3 4 5
Test points
(a) (b)
Figure
Figure8.8.(a)(a)
Measurement
Measurementofofdielectric
dielectricproperties
propertiesofofthe
themuscle
musclephantom.
phantom.TheThereadings
readingswere
weretaken
takenatat
five different locations on the phantom’s surface to ensure the accuracy of the data; (b) Graph showing
five different locations on the phantom’s surface to ensure the accuracy of the data; (b) Graph showing
the measuredrεand
themeasured , and comparison with the mean values.
r and σ, and comparison with the mean values.

5.5.Results
Results
Forvalidation,
For validation,two
twomeasurements
measurementswere werecarried
carriedout;
out;return
returnloss
loss(S
(S1111
) )and
andradiation
radiationpattern
pattern
measurement. Case A and Case B conditions were tested for both measurements, as illustrated inin
measurement. Case A and Case B conditions were tested for both measurements, as illustrated
Figure9a,b,
Figure 9a,b,respectively. TheSS1111which
respectively.The whichindicates
indicatesthe
theimpedance
impedancematching
matchingperformance
performanceofofan anantenna
antenna
wasmeasured
was measuredby byusing
using VNA;
VNA; meanwhile
meanwhile the measurement
measurement of of the
theradiation
radiationpattern
patternwas
wasconducted
conductedin
inanananechoic
anechoicchamber.
chamber. In In
3D3Dradiation pattern
radiation measurement,
pattern measurement,the antenna was rotated
the antenna on a turn-table
was rotated on a turn-in
thetain(θ)
table and(θ)
theta phi (ϕ)phi
and () and
axis, axis,the
andradiated field field
the radiated was was
measured
measured at each angle.
at each Figure
angle. 10 10
Figure shows
showsthe
measurement
the measurement configuration.
configuration.
 For validation, two measurements were carried out; return loss (S11) and radiation pattern
measurement. Case A and Case B conditions were tested for both measurements, as illustrated in
Figure 9a,b, respectively. The S11 which indicates the impedance matching performance of an antenna
was measured by using VNA; meanwhile the measurement of the radiation pattern was conducted
in an anechoic chamber. In 3D radiation pattern measurement, the antenna was rotated on a turn-
Materials 2019, 12, 1636 9 of 17
table in theta (θ) and phi () axis, and the radiated field was measured at each angle. Figure 10 shows
the measurement configuration.

Figure 9. Measurement setup for return loss (S11): (a) In free space (Case A); (b) In the presence of the
Figure 9. Measurement setup for return loss (S11 ): (a) In free space (Case A); (b) In the presence of the
muscle phantom (Case B).
Materials 2019, 12,
muscle x FOR PEER
phantom (CaseREVIEW
B). 9 of 17

Figure
Figure10.
10. (a)
(a)Measurement
Measurementconfiguration
configurationfor
forthe
theradiation
radiationpattern,
pattern,showing
showingthethestarting
startingpoint
pointof
ofθθ
andϕ;
and ; (b)
(b)Antenna
Antennaconfiguration
configurationin
infree
freespace
space(Case
(CaseA);
A);(c)
(c)Antenna
Antennaconfiguration
configurationininthe
thepresence
presenceof
of
the
themuscle
musclephantom
phantom(Case
(CaseB).
B).Readings
Readingswere
were recorded
recorded at
at every
every 2
2◦ for both θand
for both andϕorientation.
orientation.

Comparison
Comparison of of SS1111between
betweenmeasured
measured andand simulated
simulated results
results are shown
are shown in Figurein11.
Figure 11. As
As predicted,
predicted, thefrequency
the resonant resonant shifted
frequency shifted
slightly slightly
to the to thethe
right when right whenwas
antenna the tested
antenna waspresence
in the tested inofthe
the
presence of the muscle
muscle phantom, and the phantom, andsignificant
effect was the effect during
was significant duringThis
measurement. measurement. Thisdetuning
effect is called effect is
called
effect,detuning effect, and
and is calculated is calculated
based on Equationbased
(4):on equation 4:

Detuning 
Detuning∆
Detuning
Detuning%%= = 
× 100
100 (4)(4)
ffoo
By comparing
By comparing the the amount
amount of of frequency
frequency detuning
detuning to to the
the desired
desired resonant
resonant frequency
frequency through
through
simulationresults,
simulation results,the
thephantom
phantomwith withthe lowestεr rand
thelowest andσ,, which
whichisisbody
bodyfat
fathas
hasthe
thesmallest
smallestdetuning
detuning
percentage (1.14%)
percentage (1.14%) as
as compared
compared to to the
the phantoms
phantoms with higher εrrand
with higher andσ(e.g.,
(e.g.,muscle
muscleand
andblood)
blood)which
which
give1.4%
give 1.4%each.
each.This
This effect
effect is is
duedueto to
thethe changes
changes in the
in the antenna’s
antenna’s internal
internal impedance,
impedance, Zin when
Zin when it wasit
was placed
placed on a on a material
material having
having highhigh
lossloss (tanδ)
(tanδ) andand significant
significant  σvalue.
value.The
Thecorrelation
correlationbetween
betweenthethe
presence of lossy dielectric
presence dielectric materials
materialsfor variousεrεrtotoZZ
forvarious ininwill
willbe bediscussed
discussedininthe next
the nextsection.
section.Figure
Figure12
shows
12 showsthe the
radiation pattern
radiation of theof
pattern e-textile antennaantenna
the e-textile for bothforcases in E-plane
both cases inand H-plane,
E-plane and respectively.
H-plane,
Based on theBased
respectively. graphs,onslight deformation
the graphs, slightindeformation
radiation pattern was observed
in radiation patternwhich indicates which
was observed that in
terms of radiation
indicates behavior,
that in terms the on-body
of radiation behavior,condition did not
the on-body give much
condition didinfluence
not give as
muchcompared to the
influence as
free space environment.
compared to the free space However, based onHowever,
environment. the efficiency
baseddata,
on significant losses
the efficiency were
data, observed losses
significant due to
the reflection
were observedofdue signal or reflection
to the impedance ofmismatch, which will mismatch,
signal or impedance be clarifiedwhich
throughwillnear-field observation
be clarified through
in the nextobservation
near-field section. in the next section.
12 shows the radiation pattern of the e-textile antenna for both cases in E-plane and H-plane,
respectively. Based on the graphs, slight deformation in radiation pattern was observed which
indicates that in terms of radiation behavior, the on-body condition did not give much influence as
compared to the free space environment. However, based on the efficiency data, significant losses
were observed
Materials 2019, 12, due
1636 to the reflection of signal or impedance mismatch, which will be clarified through
10 of 17
near-field observation in the next section.

1.575 GHz
1.575 GHz
(a) (b)
Figure 11. (a) Simulated return loss (S11) of the antenna; (b) Measured return loss (S11) of the antenna.
Figure 11. (a) Simulated return loss (S11 ) of the antenna; (b) Measured return loss (S11 ) of the antenna.
Small deviations are observed for both cases, where the resonant frequency has shifted to the right
Small deviations are observed for both cases, where the resonant frequency has shifted to the right due
due 2019,
Materials to the12,changes
x FOR in input
PEER impedance (Zin) when the phantom is added.
REVIEW 10 of 17
to the changes in input impedance (Zin ) when the phantom is added.
MEASUREMENT PHI 0 MEASUREMENT PHI 90
0 0
5 5
30 330 30 330
0 0

-5 -5

-10 -10
60 300 60 300
-15 -15

-20 -20

-25 -25

90 -30 270 90 -30 270

5 0 -5 -10 -15 -20 -25 -30 -25 -20 -15 -10 -5 0 5 5 0 -5 -10 -15 -20 -25 -30 -25 -20 -15 -10 -5 0 5
-25 -25

-20 -20

-15 -15
120 240 120 240
-10 -10

-5 -5

0 0
150 210 150 210
5 5
180 180

Case A Case A
Case B Case B

(a) (b)
Figure 12.
Figure 12. Measured
Measured radiation
radiationpattern:
pattern:(a)
(a)Electric
Electricfield
fieldplane (E-plane);
plane (b)(b)
(E-plane); Magnetic field
Magnetic plane
field (H-
plane
plane). Slight deformation is observed due to the loss in radiated power that is absorbed by
(H-plane). Slight deformation is observed due to the loss in radiated power that is absorbed by the the lossy
medium.
lossy medium.

6.6. Discussion
Discussion and
and Analysis
Analysis
Electromagnetic
Electromagnetic radiation
radiation of
of the
the antenna
antenna waswas expected
expected to
to be
be affected
affected by
by the
thehuman
humanbody bodyandand
may
maycause
causealteration in the
alteration antenna’s
in the performance.
antenna’s Furthermore,
performance. humanhuman
Furthermore, flesh has σ and
highhas
flesh high  and
different
dielectric
different properties as compared
dielectric properties as to a free space
compared to aenvironment, thus this might
free space environment, have
thus influenced
this might have its
performance.
influenced itsThrough simulation
performance. Throughand simulation
measurement, andthe antenna’s Zinthe
measurement, significantly
antenna’s Zchanged when
in significantly

itchanged
was in when
close proximity withproximity
it was in close the human withbody. This section
the human body. presents the presents
This section analysis the
on analysis
the effects
on
of
thethe phantom’s
effects of the phantom’s e.g., εr to e.g.,
propertiesproperties overall
εr toantenna performance,
overall antenna throughthrough
performance, investigation of the
investigation
simulated near-field
of the simulated distribution.
near-field The near-field
distribution. computations
The near-field were conducted
computations to observetothe
were conducted electrical
observe the
characteristics of each case which have affected performance such as impedance and
electrical characteristics of each case which have affected performance such as impedance and efficiencies.
efficiencies.
6.1. Correlation Between On-Body Condition with Input Impedance, Zin
6.1. Correlation BetweenZOn-Body
Input impedance Condition with Input Impedance, Zin
in is an important parameter because it determines how well the power is
radiated from
Input the antenna
impedance Zin (described by radiation
is an important parameter becauseRit
resistance, r ) determines
or removed howas heat
well(due
thetopower
losses).
is
When the dielectric material is not lossless, some power will be converted as heat,
radiated from the antenna (described by radiation resistance, Rr) or removed as heat (due to losses). defined by load
When the R
resistance, L . Reactance,
dielectric X onisthe
material notother hand
lossless, is related
some powertowill
reactive near-fieldasthat
be converted oscillates
heat, definednear the
by load
antenna. Figure 13 shows the changes in
resistance, RL. Reactance, X on the other hand Z in when ε is
is related
r varied from 1 (free space) to 64 (human blood),
to reactive near-field that oscillates near the
calculated based on both resistance (R , R ) and reactance
antenna. Figure 13 shows the changes in Zin when εr is varied
r L (X).from
Based on this
1 (free diagram,
space) the value
to 64 (human of εr
blood),
calculated based on both resistance (Rr, RL) and reactance (X). Based on this diagram, the value of εr
only does not justify precisely (in magnitude) the amount of changes in Zin because in these cases,
phantoms with higher εr such as blood has a higher tanδ due to higher conductivity (σ = 1.9 S/m) as
compared to phantoms with lower εr such as body fat (σ = 0.17 S/m). Hence, more than one
parameters are varied. However, it can be observed that the presence of lossy dielectric on the
Materials 2019, 12, 1636 11 of 17

only does not justify precisely (in magnitude) the amount of changes in Zin because in these cases,
phantoms with higher εr such as blood has a higher tanδ due to higher conductivity (σ = 1.9 S/m) as
compared to phantoms with lower εr such as body fat (σ = 0.17 S/m). Hence, more than one parameters
are varied. However, it can be observed that the presence of lossy dielectric on the antenna has made
significant changes to its Zin , mainly due to the increase in RL and X. For example, when the antenna is
put on top of the body fat phantom, there is an increase of 17% in Zin , followed by 19% for muscle and
20% for blood. The increase of Zin is also expected due to the fringing effect that has significantly taken
place when a phantom was added, which resulted in the changes of electrical size of the antenna’s
radiating element. The phantom added capacitive effect to the antenna thus, the capacitive reactance,
Xc was higher and the resonant frequency was shifted to a higher value. Due to the impedance change,
mismatch occurred, which resulted in degradation of radiated power, Prad by the antenna as shown by
the graph where the Prad is reduced by 9% for fat, and 11% for muscle and blood. This was proven
through efficiency performance of the antenna which has reduced significantly especially when a
Materials 2019,
phantom with12, a
x FOR PEER
higher εr REVIEW
was added. 11 of 17

Figure 13.
Figure 13. Input
Inputimpedance
impedance andand radiated
radiated power
power of antenna
of the the antenna
whenwhen
locatedlocated near phantoms
near phantoms of
of various
various ε r. It shows that Zin is highly affected by the lossy medium, which has resulted in degradation
εr . It shows that Zin is highly affected by the lossy medium, which has resulted in degradation of
of efficiency
efficiency duedue to lower
to lower P P.rad.
rad

6.2. Correlation Between On-Body Condition with Near-Field Distribution

Distribution
The previous section has shown that antenna performance is highly affected when it is is located
located
very
very near to the
the human
humanbody body(phantom).
(phantom).According
According to to [18],
[18], detuning
detuning effects
effects willwill be higher
be higher when when
the
the separation
separation distance
distance between
between an antenna
an antenna and anand anisobject
object lower isduelower due to significant
to significant alteration
alteration to its near-
to itsdistribution
field near-field distribution
caused by the caused
object.byInthe
thisobject. In this
paper, the paper, the
separation separation
distance distance
is almost is almost
zero. Therefore,
zero. Therefore,
in order to assessinthis
order to assess
situation, thethis situation,
electric the electric
and magnetic andstrengths
field’s magneticinfield’s strengths
the reactive in the
near-field
reactive near-field
region are analyzed region are analyzed in this section. The reactive near-field
in this section. The reactive near-field region, RNF is defined region, RNF is defined
as RNF
p
3 /λ, where D is the aperture size of the antenna. Figure 14 shows the electric field
 0.62 D /  , where D is the aperture size of the antenna. Figure 14 shows the electric field
as R NF ≤ 0.62
3 D
behavior along the z-axis of antenna and phantom layers as described in Figure 7. There is a significant
behavior along the z-axis of antenna and phantom layers as described in Figure 7. There is a
amount of fields propagating inside antenna layers when a phantom is attached to the antenna. Due to
significant amount of fields propagating inside antenna layers when a phantom is attached to the
high loss characteristics in the phantom, some power was absorbed through the dielectric layers of the
antenna. Due to high loss characteristics in the phantom, some power was absorbed through the
antenna which reduced its efficiency.
dielectric layers of the antenna which reduced its efficiency.
region are analyzed in this section. The reactive near-field region, RNF is defined as RNF
 0.62 D 3 /  , where D is the aperture size of the antenna. Figure 14 shows the electric field
behavior along the z-axis of antenna and phantom layers as described in Figure 7. There is a
significant amount of fields propagating inside antenna layers when a phantom is attached to the
Materials
antenna. Due 12,high
2019,to 1636 loss characteristics in the phantom, some power was absorbed through the
12 of 17
dielectric layers of the antenna which reduced its efficiency.

Figure
Figure 14. Graph
14. Graph of electric
of electric fieldfield intensities
intensities along
along antenna
antenna and and phantom
phantom layers
layers (z-axis),
(z-axis), which
which shows
shows
the difference in E-field magnitude on the antenna surface for various εr . As compared to the free space
the difference in E-field magnitude on the antenna surface for various r. As compared to the free
(no phantom) condition, the amount of E-field intensity (V/m) inside the antenna layer is higher in the
presence of a phantom, which is expected due to the increase in Zin as observed in Figure 13.

Figure 15a illustrates the electric field (E-field) distribution along non-radiating edges, L of the
antenna. The antenna’s radiating element behaves as a perfect electric conductor at the surface;
therefore, based on the graph, the electric-field density is zero at the patch surface. Towards the edge,
the E-field is increasing due to the rise of Zin . By comparing the cases when a phantom is added,
the E-field increases with respect to the εr and σ. For example, the maximum E-field increased to 29.7%
in fat, followed by 35.3% in muscle and 36.1% in blood. Based on theoretical current-voltage relation,
this E-field observation correlates to the increment of impedance, Zin as explained in Figure 13, which
indicates that more energy is trapped inside the antenna layer when a phantom is added instead of
radiated to the air. Figure 15b,c show the E-field distribution in two-dimensional views, showing
higher magnitude when a phantom is added. The fringing fields in Figure 15c are clearly observed,
which are predicted to be one of the factors in performance degradation. The increase in E-field
intensity is also observed from the colors of arrows in this diagram.
Figure 16 shows the magnetic surface currents along radiating edges, W of the antenna. The currents
are almost uniformly distributed throughout the antenna surface, with an average reading of 12.1 A/m
and a standard deviation of only 0.27. It is important to ensure the constant behavior of current in
order to validate the correlation between E-field intensity and Zin as previously explained through
current-voltage relation. Based on comparison to electric field magnitude shown in Figures 14 and 15,
there is a clear correlation in how the E-fields are responsible for radiation performance, as illustrated
by gain and total efficiency in Figure 17. As the phantom becomes more lossy (from fat to blood),
the total efficiency has dropped by −1.74% (in fat) and −2% (in blood) in comparison to free space
performance, and thus the gain has also decreased e.g., −0.21 dB (fat), followed by −0.37 dB (muscle)
and −0.39 dB (blood). Further design optimizations are needed in this antenna to increase the gain and
efficiency, such as by increasing the σ of the e-textile material (currently, σ = ~104 S/m is considered as
low); however for a comparison study, clear observation on the factors that may affect the antenna
performance in free space and on-body has been achieved through detailed analysis of near-field
computation. Table 3 shows the detailed performance deviation of the antenna when located near a
human phantom.
in fat, followed by 35.3% in muscle and 36.1% in blood. Based on theoretical current-voltage relation,
this E-field observation correlates to the increment of impedance, Zin as explained in Figure 13, which
indicates that more energy is trapped inside the antenna layer when a phantom is added instead of
radiated to the air. Figure 15b,c show the E-field distribution in two-dimensional views, showing
higher magnitude when a phantom is added. The fringing fields in Figure 15c are clearly observed,
Materials 2019, 12, 1636 13 of 17
which are predicted to be one of the factors in performance degradation. The increase in E-field
intensity is also observed from the colors of arrows in this diagram.

Materials 2019, 12, x FOR PEER REVIEW 13 of 17

the gain and efficiency, such as by increasing the  of the e-textile material (currently,  = ~104 S/m is
considered as low); however for a comparison study, clear observation on the factors that may affect
the antenna performance in free space and on-body has been achieved through detailed analysis of
Near-field electric y-axis (non-radiating
Figure
near-field 15.
Figurecomputation.
15. Near-field Table 3field
electric field intensity:
intensity:
shows (a) Distribution
(a) Distribution graph
the detailed graph along
performance along y-axis
deviation (non-radiating edge);
edge);
of the antenna when
(b) 3D
3D view
(b)near
located view (without
(without
a human
phantom);
phantom); (c)
phantom. (c) 3D
3D view
view (with
(with muscle phantom, εεrr =
muscle phantom, 48).
= 48).

Figure 16 shows the magnetic surface currents along radiating edges, W of the antenna. The
currents are almost uniformly distributed throughout the antenna surface, with an average reading
of 12.1 A/m and a standard deviation of only 0.27. It is important to ensure the constant behavior of
current in order to validate the correlation between E-field intensity and Zin as previously explained
through current-voltage relation. Based on comparison to electric field magnitude shown in Figures
14 and 15, there is a clear correlation in how the E-fields are responsible for radiation performance,
as illustrated by gain and total efficiency in Figure 17. As the phantom becomes more lossy (from fat
to blood), the total efficiency has dropped by −1.74% (in fat) and −2% (in blood) in comparison to free
space performance, and thus the gain has also decreased e.g., −0.21 dB (fat), followed by −0.37 dB
(muscle) and −0.39 dB (blood). Further design optimizations are needed in this antenna to increase

Figure 16. Magnetic surface currents along the x-axis (radiating edge). As the values of surface
Figure 16. Magnetic surface currents along the x-axis (radiating edge). As the values of surface currents
currents are similar for all cases, this factor can be omitted. It can be observed that when surface
are similar for all cases, this factor can be omitted. It can be observed that when surface current is
current is constant, high E-field intensity (V/m) for higher r and  case is mainly due to the increase
constant, high E-field intensity (V/m) for higher εr and σ case is mainly due to the increase in Zin in
in Zin in lossy medium.
lossy medium.
 Figure 16. Magnetic surface currents along the x-axis (radiating edge). As the values of surface
currents are similar for all cases, this factor can be omitted. It can be observed that when surface
Materialscurrent
2019, 12, constant, high E-field intensity (V/m) for higher r and  case is mainly due to the increase
is1636 14 of 17
in Zin in lossy medium.

Figure17.17.Efficiency
Figure Efficiencyand
andgain
gainofofe-textile
e-textileantenna
antennafor
forvarious
variousphantom
phantomproperties.
properties.From
Fromananideal
ideal
antenna perspective, the efficiency is very low, which is due to the low conductivity and thickness of
antenna perspective, the efficiency is very low, which is due to the low conductivity and thickness of
the developed material; however, through observation on the trend, it is clear that higher loss medium
the developed material; however, through observation on the trend, it is clear that higher loss medium
will result in lower gain.
will result in lower gain.
Table 3. Changes in the antenna’s performance variables when located on a human phantom in
Table 3. Changes
comparison in thecondition.
to a free space antenna’sTheperformance
significancevariables when located
of each parameter on a human
is explained phantom in
in the paragraph.
comparison to a free space condition. The significance of each parameter is explained in the
paragraph. Fat Muscle Blood
Condition
(εr = 12) (εr = 48) (εr = 64)
Fat Muscle Blood
Detuning %
Condition 1.14 1.4 1.4
∆Zin (%) ( 
17 ↑ r = 12) (  r = 48)
19 ↑ ( r = 20
64)↑
∆PradDetuning
(%) % 9 ↓1.14 11 1.4
↓ 1.411 ↓
Zin (%)
∆E-field intensity (%) 29.717 ↑ 19 ↑
35.3  ↑
2036.1
Prad (%)
∆Total efficiency % 1.749↓ 2 ↓
11 11 2 ↓
E-field(dB)
∆Gain intensity (%) 0.21 29.7↓  35.3 ↓
0.37 36.1  ↓
0.39
 Total efficiency % 1.74  2  2 
Note: ↑ indicates positive change in the variables and ↓ indicates negative change in the variables.
Gain (dB) 0.21  0.37  0.39 
Note:  indicates positive change in the variables and  indicates negative change in the variables
6.3. Scope and Limitations
The scope of the project is limited to on-body investigation, the main focus is to provide detailed
investigation on how the antenna performance and behavior are changed with respect to the phantom’s
parameters. In the context of wearable technology, the antenna will be located very close to the
human body, thus a detailed study on the field strength at reactive near-field region is very important.
For validation, human muscle which makes up the largest proportion of the human arm was chosen
for fabrication and testing. The shape of the muscle phantom was simplified to a cuboid shape and the
size was about the same as the size of the antenna’s radiator. All experiments in this paper were not
conducted with actual human participants to ensure consistency of test environment and to reduce
other uncertainties such as human body movement. Furthermore, during far-field radiation pattern
measurement, high power electromagnetic waves were transmitted to the antenna-under-test at a very
close range, thus by using a human phantom, safety and health precautions were ensured. Near-field
simulation results in this paper were used to analyze and validate the behavior of the antenna in
two conditions; free space and on-body. Actual near-field measurement was not conducted due to
the unavailability of a near-field probe which is very costly and requires expensive chamber setup
(~millions in USD). Other conditions such as bending or stretching are not included in this paper to
maintain the consistency of the discussion and analysis.
In the future, based on the analysis presented in this paper, optimization works can be done
to improve the performance of the antenna in both free space and on-body conditions. It is clearly
observed that the task of designing antennas located near human body parts with different properties
Materials 2019, 12, 1636 15 of 17

(εr , σ, tanδ) requires more considerations as compared to free space. Due to the complexity of the
environment that can significantly affect the antenna’s performance, more optimization works are
needed. One solution is by designing the antenna with Zin that is lower than the system impedance so
that mismatch loss can be minimized in on-body condition. A high bandwidth antenna can also be
designed to reduce frequency detuning effects. Other than that, based on the data shown in this paper,
the usage of a ‘buffer-like layer’ can also be proposed.

7. Integration with GPS Tracking Device

In order to verify the ability of the antenna for tracking application, the most optimum antenna
prototype (despite the performance change) was incorporated and tested in a GPS tracking system test
bench, as illustrated in the basic block diagram in Figure 18. To demonstrate the receiving capability,
a GPS module was integrated with the e-textile antenna operating at 1.575 GHz and connected to
Arduino board that was attached to a Wi-Fi/3G antenna for data transmission to the server. From the
server, the precise location of the e-textile antenna was displayed, as shown in Figure 18a. Based on
the data received at the computer, the GPS wearable antenna developed in this research worked well
in receiving GPS data with no data loss within the required time interval, which indicates that the
technical performance of the antenna in terms of its resonant frequency, impedance and gain was
sufficient to act
Materials 2019, 12,as an important
x FOR sub-system to the GPS tracking device.
PEER REVIEW 15 of 17

(a)

(b)

Figure18.18.(a)
Figure (a)Block
Blockdiagram
diagramofofGPS
GPStracking
trackingsystem
systemtest
testbench;
bench;(b)
(b)Measurement
Measurementsetupsetupofofthe
the
integration of e-textile antenna with GPS tracking module. During operation, precise and
integration of e-textile antenna with GPS tracking module. During operation, precise and consistentconsistent
real
real time
time GPSGPSdata data (location)
(location) were received
were received at the
at the server server without
(computer) (computer) without
any losses any losses or
or intermittence.
intermittence.

8. Conclusions
The correlation between a phantom’s dielectric properties to antenna technical performance
such as impedance, electric field and gain was analyzed systematically in this paper. As validated
through simulation and measurement, the presence of a phantom added some capacitive elements
in the antenna structure, thus resulted in the shift of resonant frequency due to higher capacitive
reactance in the input impedance value. Investigation through near-field distribution has been done,
Materials 2019, 12, 1636 16 of 17

8. Conclusions
The correlation between a phantom’s dielectric properties to antenna technical performance such
as impedance, electric field and gain was analyzed systematically in this paper. As validated through
simulation and measurement, the presence of a phantom added some capacitive elements in the
antenna structure, thus resulted in the shift of resonant frequency due to higher capacitive reactance in
the input impedance value. Investigation through near-field distribution has been done, which agrees
very well with the theoretical current-voltage relation that shows that as the impedance is increasing,
mismatch occurs. Thus, at 1.575 GHz, the electric field intensity inside the antenna is increased and
less power is radiated by the antenna to its surroundings, which can affect the key performance of the
antenna such as gain and efficiency. The flow of the study and parameters evaluated in this paper can
be established as good methods in analyzing and validating the reason for performance degradation
when an antenna is located near the human body. These findings are very useful to antenna designers,
especially in determining the best location for their wearable antennas.

Author Contributions: Conceptualization, N.H.A.R.; Methodology, N.H.A.R. and Y.Y.; Fabrication, N.H.A.R. and
M.S.A.N.; Simulation, N.H.A.R. and M.S.A.N.; Measurement, N.H.A.R. and M.S.A.N.; Formal Analysis, N.H.A.R.
and Y.Y.; Investigation, N.H.A.R. and Y.Y.; Writing—Original Draft Preparation, N.H.A.R.; Writing—Review &
Editing, N.H.A.R and Y.Y.; Supervision, N.H.A.R. and Y.Y.; Project Administration, N.H.A.R. and Y.Y.; Funding
Acquisition, Y.Y.
Funding: This research was funded by Universiti Teknologi Malaysia through the Professional Development
Research University grant (PY/2017/02232) and Malaysia Ministry of Education through the Fundamental Research
Grant Scheme (FRGS/1/2018/TK04/UITM/03/5).
Acknowledgments: Authors would like to thank all researchers of Antenna Research Centre, Faculty of Electrical
Engineering, Universiti Teknologi MARA, Shah Alam, Selangor, Malaysia and Textile Research Group, Faculty of
Applied Sciences, Universiti Teknologi MARA, Shah Alam, Selangor, Malaysia for supporting this project.
Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the
study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to
publish the results.

References
1. Wang, Z.; Zhang, L.; Volakis, J.L. Textile antennas for wearable radio frequency applications. TLIST 2013, 2,
105–112.
2. Amendola, S.; Lodato, R.; Manzari, S.; Occhiuzzi, C.; Marrocco, G. RFID technology for IoT-based personal
healthcare in smart spaces. IEEE Internet Things J. 2014, 1, 144–151. [CrossRef]
3. Lee, H.; Tak, J.; Choi, J. Wearable antenna integrated into military berets for indoor/outdoor positioning
system. IEEE Antennas Wirel. Propag. Lett. 2017, 16, 1919–1922. [CrossRef]
4. Moradi, E.; Koski, K.; Bjorninen, T.; Sydanheimo, L.; Rabaey, J.M.; Carmena, J.M.; Rahmat-Samii, Y.;
Ukkonen, L. Miniature implantable and wearable on-body antennas: Towards the new era of wireless
body-centric systems. IEEE Antennas Propag. Mag. 2014, 56, 271–291.
5. Yahya, R.; Kamarudin, M.R.; Seman, N. Effect of Rainwater and Seawater on the Permittivity of Denim
Jean Substrate and Performance of UWB Eye-Shaped Antenna. IEEE Antennas Wirel. Propag. Lett. 2014, 13,
806–809. [CrossRef]
6. Kiourti, A.; Volakis, J.L. Conductive textiles for wearable electronics. In Proceedings of the International
Symposium on Electromagnetic Theory, Espoo, Finland, 14–18 August 2016; 2016; pp. 518–519.
7. Haghi, M.; Thurow, K.; Stoll, R. Wearable devices in medical Internet of Things: Scientific research and
commercially available devices. J Healthc Inform Res. 2017, 23, 4–15. [CrossRef] [PubMed]
8. Ferreira, D.; Pires, P.; Rodrigues, R.; Caldeirinha, R.F.S. Wearable Textile Antennas: Examining the effect of
bending on their performance. IEEE Antennas Propag. Mag. 2017, 59, 54–59. [CrossRef]
9. Christina, G.; Rajeswari, A.; Lavanya, M.; Keerthana, J.; Ilamathi, K.; Manoranjitha, V. Design and development
of wearable antennas for tele-medicine applications. In Proceedings of the IEEE International Conference on
Communication and Signal Processing, Melmaruvathur, India, 6–8 April 2016; pp. 2033–2037.
10. Lilja, J.; Salonen, P.; Kaija, T.; Maagt, P.D. Design and manufacturing of robust textile antennas for harsh
environments. IEEE Trans. Antennas Propag. 2012, 60, 4130–4140. [CrossRef]
Materials 2019, 12, 1636 17 of 17

11. Roshni, S.B.; Jayakrishnan, M.P.; Mohanan, P.; Surendran, K.P. Design and fabrication of an E-shaped
wearable textile antenna on PVB-coated hydrophobic polyester fabric. Smart Mater. Struct. 2017, 26, 1–8.
[CrossRef]
12. Abd Rahman, N.H.; Yamada, Y.; Amin Nordin, M.S. Reliability of Strip Line Method for Determination of
Conductivity for Lossy Conductive Materials. IEEE Access 2018, 6, 64630–64638. [CrossRef]
13. Melia, G. Electromagnetic Absorption by the Human Body from 1–15 GHz. Ph.D Thesis, University of York,
York, UK, August 2013.
14. Seneviratne, S.; Hu, Y.; Nguyen, T.; Lan, G.; Khalifa, S.; Thilakarathna, K.; Hassan, M.; Seneviratne, A.
A survey of wearable devices and challenges. IEEE Commun. Surv. Tutor. 2017, 19, 2573–2620. [CrossRef]
15. Pinapati, S.P.; Chen, S.J.; Ranasinghe, D.; Fumeaux, C. Detuning effects of wearable patch antennas. In
Proceedings of the Asia Pacific Microwave Conference, Kuala Lumpur, Malaysia, 13–16 November 2017;
pp. 162–165.
16. Soh, P.J.; Vandenbosch, G.A.E.; Ooi, S.L.; Rais, N.H.M. Design of a Broadband All-Textile Slotted PIFA.
IEEE Trans. Antennas Propag. 2012, 60, 379–384. [CrossRef]
17. Keshmiri, F. Body-Area-Network Antennas: Green’s Functions, Numerical Analysis and Design. Ph.D
Thesis, Université Catholique de Louvain, Louvain-la-Neuve, Belgium, February 2012.
18. Kellomaki, T. Effects of the Human Body on single layer wearable antennas. Ph.D Thesis, Tampere University
of Technology, Tampere, Finland, March 2012.
19. Chowdhury, S.R.; Ali, K. Effects of human body on antenna performance: A quantitative study. In Proceedings
of the International Conference on Computer and Information Technology (ICCIT), Florence, Italy, 11–12
April 2016; IEEE: Dhaka, Bangladesh, 2016; pp. 108–112.
20. Kumar, V.; Gupta, B. On-body measurements of SS-UWB patch antenna for WBAN applications. AEU-Int. J.
Electron. Commun. 2016, 72, 184–191. [CrossRef]
21. Fujita, K.; Yoshitomi, K.; Yoshida, K.; Kanaya, K. A circularly polarized planar antenna on flexible substrate
for ultra-wideband high-band applications. AEU-Int. J. Electron. Commun. 2015, 69, 1381–1386. [CrossRef]
22. Boyes, S.J.; Soh, P.J.; Huang, Y.; Vandenbosch, G.A.E.; Khiabani, N. Measurement and performance of textile
antenna efficiency on a human body in a reverberation chamber. IEEE Trans. Antennas Propag. 2013, 61,
871–881. [CrossRef]
23. Abdullah, A.S.; Ahmed, A.G.; Nikolay, D.; Nikolay, A. On-body investigation of a compact planar antenna
on multilayer polymer composite for body-centric wireless communications. AEU-Int. J. Electron. Commun.
2017, 82, 20–29.
24. Heaney, M.B. Electrical Conductivity and Resistivity. In Electrical Measurement, Signal Processing, and Displays;
Webster, J.G., Ed.; CRC Press: Boca Raton, FL, USA, 2003; pp. 7–11.
25. ASTM D3776—Standard test methods for mass per unit area (weight) of fabric; ASTM International: West
Conshohocken, PA, USA, 2009.
26. Keysight Technologies, Basics of Measuring the Dielectric Properties of Materials—Application Note, 2017.
Available online: http://literature.cdn.keysight.com/litweb/pdf/5989-2589EN.pdf (accessed on 2 January 2019).
27. Hasgall, P.A.; Gennaro, F.D.; Baumgartner, C.; Neufeld, E.; Lloyd, B.; Gosselin, M.C.; Payne, D.;
Klingenböck, A.; Kuster, N. Data from: IT’IS database for thermal and electromagnetic parameters of
biological tissues V4 [Dataset]; IT’IS Foundation: Zurich, Switzerland, 2018; Available online: https:
//doi.org/10.13099/VIP21000-04-0 (accessed on 28 December 2018).
28. Cavallari, R.; Martelli, F.; Rosini, R.; Buratti, C.; Verdone, R. A Survey on Wireless Body Area Networks:
Technologies and Design Challenges. IEEE Commun. Surv. Tutor. 2014, 16, 1635–1657. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).