Professional Documents
Culture Documents
Thesis
Thesis
Bachelor of Engineering
Honours Thesis
By
October 2000
3/1 Laurence St,
St Lucia, QLD 4067
The Dean
School of Engineering,
The University of Queensland,
St Lucia, QLD 4072.
Dear Sir,
This work was performed under the supervision of Associate Professor Marek E.
Bialkowski. I declare that the work submitted in this thesis is my own, except as
acknowledged in the text and footnotes, and have not been previously submitted for a
degree at the University of Queensland or any other institution.
Yours Faithfully,
i
Acknowledgments
ii
Abstract
The topic of this thesis is on the behaviour of the hand-held transceiver when
placed in proximity with a human body. The ante nnas that are associated with portable
transceivers are usually physically and electrically small. Because of the physical reason,
mobile phones are generally carried around and are very close to people, especially the
head and the hand of the user. There have been many studies on the performance of the
antenna when held near human body. Some studies are presented in this thesis as a
guideline to what is expected at the end of the thesis.
Several orientations for the position of the antenna are applied to the
measurements, with and without the presence of a human head model. Results show that
the presence of a human head can influence the radiation efficiency of the antenna.
More results are being elucidated in the discussion.
iii
Table of Contents
Table of Contents
CHAPTER 1 Introduction 8
1.1 A Brief Historical Perspective 8
1.2 Mobile Communications 9
1.3 Antenna Systems for Mobile Communications 10
1.4 Objective of Thesis 12
1.5 Components of Thesis 13
CHAPTER 4 Methodology 37
4.1 Introduction 37
4.2 Antenna Under Test 38
iv
Table of Contents
CHAPTER 6 Conclusion 62
REFERENCES
v
List of Figures
List of Figures
Figure 5.1 : 3D radiation pattern for the 1800 MHz radiating element 51
without the human head.
Figure 5.2 : 3D radiation pattern for the 1800 MHz radiating element 52
with the human head.
vi
List of Figures
Figure 5.3 : Position of the 1800 MHz radiating element during front and 54
back scanning.
Figure 5.4 : Near Field Vertical plane Magnitude plot for Front radiation scanning 55
without the human head.
Figure 5.5 : Near Field Vertical plane Magnitude plot for Front radiation scanning 55
with the human head.
Figure 5.6 : Near Field Vertical plane Magnitude plot for Back radiation scanning 56
without the human head.
Figure 5.7 : Near Field Vertical plane Magnitude plot for Back radiation scanning 56
with the human head.
Figure 5.8 : Near Field Horizontal plane Magnitude plot for Back radiation scanning 57
without the human head.
Figure 5.9 : Near Field Horizontal plane Magnitude plot for Back radiation scanning 57
with the human head.
Figure 5.10 : Near Field Horizontal plane Magnitude plot for Front radiation scanning 58
without the human head.
vii
Chapter 1 : Introduction
CHAPTER 1
Introduction
Towards the end of the nineteenth century, Guglielmo Marconi established the
first use of wireless communica tion, linking a land-based station and a tugboat over
several miles. By 1897, Marconi had patent and tested a complete wireless telegraph
system capable of operating over a distance of up to 7 km. In 1899, he was able to link,
continuously, contacts betwee n ships and onshore stations along the English Channel,
and by mid-December 1901, the first transatlantic transmission between England and
Canada was accomplished by Marconi and his associates, which opened up the
possibility of international wireless communications [2]. These achievements by
Marconi soon made him the father of wireless communications industry.
8
Chapter 1 : Introduction
Since then, the wireless communications industry has progress and advances
rapidly, showing no sign of slowing down. The invention of the triode vacuum tube
used in the detectors, amplifiers and oscillators circuits, which are used for the
generation and detection of the radio signal during the period from 1904 to 1915
sparked another revolution in wireless communications. During this period of
advancement in wireless communications technology, Edwin Armstrong designed and
created the superheterodyne receiver circuit, which are used for the actual voice or
music transmission [2]. This initial success of wireless communications shortly began to
be a reality and further exploration was made towards today’s booming realm of
personal communications.
In the early 20th century, mobile technology had been predominated by military
users. Before World War II, most development of mobile communications was
designated at military requirements and standards. In fact, the early mobile radio
systems were so bulky and heavy that it is essential for the equipment to fill the trunks
of those vehicles carrying it. It was not until the late 1950s, that highly reliable
transistors superseded triode vacuum tubes used in the mobile radio systems. This
transformation led to a massive reduction of almost 50% to the volume and power
consumption of the system [3]. Solid-state components later came into the limelight by
the 1960s when they were incorporated into mobile technology. It was also during this
time that the first hand-held transceivers made its introduction. In the mid-1970s,
integrated circuits took a change over from transistors, making the mobile system
smaller, lighter and less costly.
Towards the end of the 20t h century, mobile communication has been the main
focus in personal communications. There is a rapid increase in the ownership of a hand-
held transceiver (or mobile phone) by pe ople over the world [1]. With this inclination,
modern mobile phones are not only getting physically smaller, but also greater
enhancement in performance.
9
Chapter 1 : Introduction
In the last few years, the trend of the mobile phone technology has been to
dramatically decrease the weight and the size of the hand-held unit. It is because of this
trend, that the antennas used for mobile phone have to be small, light-weighted, low
profile, and have an omni-directional radiation pattern in the horizontal plane. However,
there will st ill be changes in the antennas’ performance during interaction with the
user’s head and hand, and the movement of the user during usage of the mobile phone
often lead to gain, radiation pattern and input impedance change. Therefore, antennas
used in hand-held transceivers for personal communications have been recognized as
crucial elements that can either improve or limit system performance. This is
particularly true in terms of bandwidth and efficiency. Therefore, to carefully design a
handset with superior performance, engineers need to give attention to the design of the
antenna system of the hand-held transceiver.
Helical antenna has a structure that behaves like a spring and is flexible. This
antenna can be constructed to be shorter than a ë/4 monopole and by choosing the size,
pitch, and the number of turns of the helical structure; a good performance on its
efficiency can be obtained. For mobile communications, the normal mode helical
antenna has numerous advantages.
10
Chapter 1 : Introduction
The Inverted-F antenna (IFA) is another antenna that is being modified from the
monopole element. Derived from the Inverted-L antenna (ILA), it has a low-profile
structure and has often been mounted onto the body of flying vehicles such as
aeroplanes and rockets. The Planar Inverted-F antenna (PIFA) [8] was later modified
from the IFA. PIFA typically consists of a rectangular planar element and is often
placed at a corner of a mobile unit. It has an omnidirectional radiation characteristic and
is less costly than other high gain antennas. The antenna structure was found to be
useful for mobile equipment since the antenna system, with inclusion of the equipment
body as part of the radiator, exhibited a reasonably high gain and broad enough
bandwidth to cover the frequency bands used for mobile systems.
Currently, there are several types of antennas used in mobile phones. Monopole
and Planar Inverted-F antenna (PIFA) are typical examples of popular types of antenna
currently used in mobile communications. It has been a tradition for mobile phone to
use a monopole type of antenna. Monopole type antenna is easy to design and it offers
omnidirectional radiation pattern that is convenient for receiving signals from vertically
polarised base station antennas. On the other hand, PIFA offers significant advantages
not only in size and efficiency, but also its ability to exhibit sensitivity to both vertically
and horizontally polarised radio waves. This reportedly improved the average received
power in urban area where the cross-polarisation coupling is relatively large. Therefore
PIFA is suitable for use with portable radio equipment in which antenna orientation is
not fixed.
Nevertheless, the presence of human body is still the primary concern, for this
obstacle can significantly perturb antenna performance in terms of efficiency and
pattern characteristics.
11
Chapter 1 : Introduction
2. Perform the characteristic test of the antenna with and without the presence
of a human head using the newly acquired software package, FEKO.
The first goal allows the author to acquire an in-depth understanding of the
problem mention in the beginning of this section. Studies on past researches and papers
by antenna designers and engineers around the world are to be reviewed. Additionally,
the selection of the antenna to be tested must also be addressed. Furthermore, the
criterion to the second goal is to perform simulations on the characteristic of the chosen
antenna, with and without the presence of a human head, using FEKO, the recently
acquired software by the University of Queensland. However, much of the work will be
concentrated on the testing and measurement of this antenna in the University of
Queensland’s Near Field Testing Range, which is the final and most crucial goal of this
thesis.
12
Chapter 1 : Introduction
Chapter 2 of this thesis will review some of the relevant articles regarding
human interactions with mobile base antenna. The influence of the human on the
antenna gain, radiation pattern, and the input impedance are some of the issues, which
were being discussed throughout several papers for the past decade. Reviews of these
papers are presented in brief summaries.
Chapter 4 will review the method used for the simulations of results and the
experiment set up in the University of Queensland Near Field Testing Range. Included
in this chapter is a brief introduction to the newly acquired software, FEKO. Different
orientations and set ups for the experiment in the anechoic chamber are illustrated with
top view drawings.
The early part of Chapter 5 presents the results produced from the experiment
described in Chapter 4. The experimental results are presented in three dimensional
colour plots. Seven magnitude plots are displayed and the results are discussed in the
end part of the Chapter.
13
Chapter 2 : Review of Relevant Articles
CHAPTER 2
2.1 Introduction
• In “Effects on Portable Antennas of the Presence of a Person” [4], Jorn Toftgard and
colleagues brought upon two commonly asked questions on the interactions of
handset antenna and human. They are, (1) the possibility of health hazard for the
user, and (2) whether the person will seriously degrade the antenna performance.
The main topic to be investigated is, however, on how antenna parameters are
affected by the presence of the operator. The parameters discusse d here are the input
impedance, far-field radiation pattern, radiation efficiency, and the magnitude of the
near field.
The technique used for the simulated results of these antenna characteristics is the
finite -difference time-domain (FD-TD) method. This flexible numerical tool was
14
Chapter 2 : Review of Relevant Articles
chosen because of its applicability to operate with both metallic and lossy dielectric
structures, when simulations on models of the combination of the human body and
the hand-held portable telephone have to be made. The dimensions of the simulated
model of the human head and hand are closely constructed to the size of a real
person, with the modelled hand wrapping around the handset with the antenna. The
ë/4 monopole antenna was used throughout the experiment, which was carried out in
914 MHz and 1890 MHz.
Radiation at 914 MHz, with the hand-head model of the operator, is distributed
between Ö and è polarization. The authors found that less power is radiated in the
direction where the person’s head is located. The measured pattern agrees well with
the simulated one, but there is a difference in level, indicating that more power is
absorbed in the real body than in the model used for simulations.
The resonance frequency, on the other hand, is in reasonable agreement with the
measured results for the handset only. But when these impedance curves are
compared with those for handset and operator, it is seen that the resonance
frequency is shifted 10% lower. When measuring the radiation efficiency of the
handset with person, it is discovered that almost half the power is lost in the persons
hand and head; 96% of the absorbed power is lost in the person’s head and only 4%
is lost in the hand. This suggest that the further the head is away from the handset,
the better the efficiency.
Jorn Toftgard and colleagues concluded that at higher frequency, good qualitative
agreement with measurements for gain, radiation pattern, and impedance was
obtained. It is proven that the head and the hand undeniably caused significant
changes to the radiation pattern and efficiency of the handset’s antenna, with 45% of
the power being lost in the human.
15
Chapter 2 : Review of Relevant Articles
• The team of Sato, Nishikawa, Suzuki and Ogawa work on the previous
accomplishment by [4] and investigate the effects of the human body on the
antennas using the famous Finite -Difference Time-Domain (FDTD) method. The
FDTD method will be further elaborated in section 2.3.
The authors claimed that the effects of the shape and arrangement of the human
body model were not examined by Toftgard et al. [4]. They believed that the
characteristics of the antennas are largely affected by variation of the shape and
arrangement of the human body model. Therefore, two analytical models were
constructed for FDTD analysis. One model had the shape of a rectangula r prism,
while the other is an ellipsoid of rotation. Also take note that the combination of the
human head and hand are considered as the standard human body model. For the
model to express the effects of an actual human body, physiological saline solution
is filled in the model. The calculated results agreed with the measured results within
3 dB, and the author proved that the dummy could express the effects of an actual
human body.
Calculations were made for the planar inverted-F antenna mounted on portable
equipment near the human body model. The radiation efficiency decreases to 14 %
and 36 % at 900 MHz and 1900 MHz respectively for the human head and hand
model. Without the hand, the radiation efficiency is 50% and 62% at 900 MHz and
1900 MHz respectively. The ratio of power dissipated in the human head model to
the whole dissipated power is 32% and 52% at again, 900 MHz and 1900 MHz,
respectively. The authors thus proved that the human hand model affects the
radiation efficiency strongly and that radiation efficiency decreases as the frequency
decreases. On the other hand, the radiation efficiency calculated using the ellipsoid
model is about 5% larger than that calculated using the rectangular prism model.
Hence confirming that the radiation efficiency varies little with the shape of the
human head.
16
Chapter 2 : Review of Relevant Articles
Sato and colleagues also discovered that the radiation efficiency decreases as the
frequency decreases, because the size of the antenna is relatively large and the
planar of the antenna with the lower frequency is closer to the hand. When the hand
model covers the antenna element, the radiation efficiency drops to 13%. It is later
confirmed that when the distance between the lower edge of the planar element and
the hand model is 0.1 wavelength or more, the radiation efficiency is over 40%.
17
Chapter 2 : Review of Relevant Articles
field is found in the same manner. The new value of the H field is dependent on the old
value of the H field (hence difference in time), and also dependent on the difference in
the E field on either side of the H field point. This description holds true for 1D, 2D and
3D, FDTD techniques. However, when multiple dimensions are considered, the
difference in space must be considered in all appropriate dimensions.
Nonetheless, every modelling technique has some strengths and some weakness.
Some types of models with a given technique will excel, and other types of models with
the same technique would have difficulty performing rapidly and accurately. FDTD is a
very versatile modelling technique. FDTD is a time domain technique, and when a time-
domain pulse (Gaussian pulse) is used as the source pulse, then a wide frequency range
is solved with only one simulation.
18
Chapter 3 : The Planar Inverted-F Antenna
CHAPTER 3
3.1 Introduction
It is eminent that, with the reduction in the size of an antenna, the efficiency
tends to degrade as well, and the bandwidth becomes much narrower. For a size -
reduced antenna, matching to the load becomes equally essential, so as to make the
antenna practically useful. One of the effectual ways to achieve efficient matching is to
reach self -resonance of the antenna, since a self-resonant antenna is purely resistive at
the frequency of opera tion and no conjugate circuit between the antenna and the load is
necessary. For that reason, a physically constrained antenna is required and the planar
inverted-F antenna (PIFA) is most suited for this application. Although the planar
inverted-F antenna has been around for sometimes, its theory and operation are still
presented here so as to provide a good understanding of the principles of their operation.
19
Chapter 3 : The Planar Inverted-F Antenna
The origin of the planar inverted-F antenna can be traced all the way back to the
Inverted-L antenna (ILA). The antenna has a structure as shown in Figure 3.1, and may
be categorized as a physically constrained small antenna. It consists of a short monopole
as a vertical element and a wire horizontal element attached at the end of the monopole.
horizontal element
vertical element
ground plane
driving point
The ILA is basically a low profile structure, due to the fact that the height of the
vertical element is usually constrained to a fraction of the wavelength. The horizontal
element is not necessarily very short, and usually has a length of about a quarter
wavelength. The ILA has inherently low impedance, since the antenna is essentially a
vertical short monopole loaded with a long horizontal wire at the end of the monopole.
The input impedance is comparable to that of the short monopole, in addition the
reactance of the horizontal wire, which is closely placed to the ground plane. Hence to
increase the radiation impedance, another inverted-L element is attached to the end of
the vertical element and this is how the inverted-F antenna (IFA), as depicted in Figure
3.2, is formed from the ILA.
20
Chapter 3 : The Planar Inverted-F Antenna
horizontal element
vertical element
ground plane
driving point
This adjustment can be imperative because the input impedance of an IFA can
be set to have an appropriate value to match the load impedance, without using any
additional circuit between the antenna and the load. For this reason, the IFA, rather than
the ILA, has been used practically and applied often to moving bodies such as rockets
and aircraft due to its low profile structure. In addition to the above-mentioned features,
its performance with two polarizations would be useful for urban environmental use.
This is especially true for application on mobile equipment like the hand-held
transceiver.
21
Chapter 3 : The Planar Inverted-F Antenna
horizontal planar
element
vertical plate
element
ground plane
feed source
Figure 3.3 : The early planar inverted-F antenna (PIFA) modified from IFA.
22
Chapter 3 : The Planar Inverted-F Antenna
The the ory and operation of the Planar inverted-F Antenna can be explained
briefly by means of a three dimensional electro-magnetic field time-domain analysis
method known as the Spatial Network Method (SNM). The application of this method is
meant to provide the reader with a good understanding of the principles of their
operation and is not actually used by the author in the work of this thesis. For detailed
study into the Spatial Network Method, it is suggested to look into [7] as it is more
thoroughly explained.
A model of the PIFA used in the SNM is depicted in Figure 3.4. It has the
dimensions of L1 = L2 = 16 d and H = 4 d, where d is the distance between similar
nodes of the SNM, and the width of the sort circuit plate varies with 2 d, 4 d, 8 d,
12 d and 16 d. Conductive and dielectric losses are neglected during the application of
the SNM, and a TEM transmission line feeds the radiating patch of the PIFA from the
back of the ground plane.
Figure 3.4 : The PIFA model as used in the Spatial Network Method.
23
Chapter 3 : The Planar Inverted-F Antenna
To determine the electrical field and current distribution, the SNM has to ensue
on the following procedure. First, a pulse wave is inserted into the TEM transmission
line in order to view the reflectio n wave. After that, this wave, which is in the time
domain, is Fourier transformed into the frequency domain so as to determine the
reflection coefficient. Analysing the resonant frequency and the bandwidth
characteristics of the antenna can then be easily done by determining the site of the feed
point, which the minimum reflection coefficient is to be obtained. And with the
introduction of a sinusoidal wave (with the resonant frequency) into the TEM
transmission line, the electric field distribution at resonance can be acquired from the
amplitude of the electric node. Correspondingly, the magnetic node determined the
current distribution. Both the electric field and current distributions are represented in
Figure 3.5 and 3.6 respectively.
24
Chapter 3 : The Planar Inverted-F Antenna
Figure 3.5 : The distribution of the electrical fields E x, Ey and Ez on the x-y plane.
25
Chapter 3 : The Planar Inverted-F Antenna
26
Chapter 3 : The Planar Inverted-F Antenna
The width of the short circuit plate of the PIFA played a very important part in
governing its resonant frequency [7]. Figure 3.7 shows the graph of the frequencies
change when the width of the short circuit plates is varied. The resonant frequency
decreases with the decrease in short circuit plate width; W. However, approximately 3%
difference is reflected in the results, which should be caused by the discrete positioning
of the feed point that must be located at the electric nodes with the interval of d [7].
Figure 3.7 : Frequencies change with varying widths of the short circuit plates.
27
Chapter 3 : The Planar Inverted-F Antenna
Conversely, changes in the width of the planar element can too affect the
determination of the resonant frequency. The graph in Figure 3.8 shows the
characteristics of resonant frequency with respect to the size ratio of the planar element,
L1/L2. Take note that only L1 is varied, so that the overall size ratio can change
accordingly.
Figure 3.8 : Frequencies change with varying widths of the planar element.
28
Chapter 3 : The Planar Inverted-F Antenna
Traditionally, antennas used on portable radio units have been whips, sleeve
dipoles or monopole. However, due to the rising demand of portability and compactness
of hand-held transceivers, it is desirable that the antenna be built-in. Among the existing
built-in antenna schemes, the planar inverted-F antenna is one of the most promising
candidates considering its compactness. In addition to portability, the PIFA exhibits
sensitivity to both vertically and horizontally polarized radio waves and therefore is
suitable for use with hand-held transceiver in which the antenna orientation is not fixed.
This section will briefly elucidate the characteristics of the PIFA when mounted on
conducting boxes, which is an introduction to what the author will investigate on in this
thesis.
The characteristics of the planar inverted-F antenna on the infinite ground plane
are expected to be similar to those of the linear inverted-F antenna. However, the
characteristics of a PIFA mounted on portable conducting cases (or boxes) are different
from those of a PIFA on the infinite ground plane. They depend on the positioning of
the antenna on the conducting boxes, and also the dimensions of the conducting boxes,
which induce surface current [8]. Some typical configurations of the PIFA mounted on
the conducting boxes are shown in Figure 3.9. These three configurations of the PIFA
consist individually of a rectangular planar radiation element, a short pin between the
radiation element and the surface of the box, and the feed line. Figure 3.9 (a) depict the
PIFA mounted on the lateral side of the conducting box, Figure 3.9 (b) on the top of the
box, and Figure 3.9 (c) on the back of the box. By carefully selecting an appropriate
distance between the feed point on the radiation element and the short pin, the input
impedance of the PIFA elements, including the conducting box, can be matched.
29
Chapter 3 : The Planar Inverted-F Antenna
The wire-grid model is used for analysing the characteristics of the antenna
mounted on the conducting boxes with dimensions comparable to the wavelength [8].
Figure 3.10 shows how the wire-grid model is applied for the configurations A and B of
the PIFA with conducting box. Again, this method is not use in the study of this thesis.
Four parallel wires perpendicularly intersecting five parallel loops replace the
conducting surfaces of the rectangular radio case, and all wire cross-points are
connected (see Figure 3.10). In addition, a rectangular loop is used to replace the PIFA
planar radiator; the feed point in Figure 3.10 (a) is placed at the corner of the top loop of
the radio case, and that in Figure 3.10 (b) is placed at a 0.0418ë offset point from the
corner of the radiator loop [8]. The length, width, and depth of the conducting box is
represented by L, W, and D respectively, and l, w and h are the length, width, and height
of the antenna radiator.
30
Chapter 3 : The Planar Inverted-F Antenna
The radiation pattern of the antenna is the relative distribution of radiated power
as a function of direction in space. It is an important component in the determination of
the characteristics of an antenna. This is especially so when the radiation pattern is
essential in determining the amount of power radiated by the antenna being absorbed by
a human body in the later part of this thesis.
The radiation patterns for Figure 3.9 (a) are shown in Figure 3.11. In the XY
plane, the Eè component shows an approximately omnidirectional pattern, which is
caused by the surface currents on the conducting box. The E Ö components in the XY
and YZ planes are mainly the radiation from the feed line because the amplitude of the
current on the feed line is much higher than that on other wires [8]. However, the most
effective parameter for the radiation pattern is the length L of the conducting box.
Figure 3.12 depicts the variation of radiation patterns for the length L. The radiation
pattern reflects that the longer the length L, the more the Eè component patterns in the
YZ plane face slightly downward. All characteris tics found for Figure 3.9 (a) are the
same for Figure 3.9 (b).
31
Chapter 3 : The Planar Inverted-F Antenna
Figure 3.11 : Radiation patterns of PIFA mounted on the side of the conducting
box (D = 0.17ë, W = 0.11ë, L = 0.55ë, l = 0.14ë, w = 0.09ë, h = 0.03ë, f = 920 MHz).
32
Chapter 3 : The Planar Inverted-F Antenna
In the case of the antenna configuration in Figure 3.9 (b), the radiation pattern is
different because the planar element of the PIFA is facing the top and the surface area of
the conducting box under the antenna is much smaller, in this case, equal. Figure 3.13
shows the radiation pattern for this top configuration of the PIFA. The Eè component
patterns in the YZ plane face slightly more downwards as the length of the conducting
box, which is the depth in this configuration, is shorter than that of Figure 3.9 (a). The
radiation level of the EÖ components for the top mounted configuration in the XZ and
YZ planes is 10 dB less than that of the side mounted one, in the YZ plane [8]. Because
under practical use conditions the hand-held transceiver is almost often inclined, the
sensitivity for vertically polarized waves changes, from the Eè component pattern in the
XZ plane to the EÖ component pattern in the XZ or YZ plane [8].
Figure 3.13 : Radiation patterns of PIFA mounted on the top of the conducting box
(D = 0.17ë, W = 0.11ë, L = 0.52ë, l = 0.15ë, w = 0.09ë, h = 0.03ë, f = 920 MHz).
33
Chapter 3 : The Planar Inverted-F Antenna
3.4.2 Bandwidth
The bandwidth of the PIFA depends mainly on the antenna height h. Therefore
the required bandwidth can easily be obtained by increasing the antenna height. The
graph of relative bandwidth versus the antenna height is shown in Figure 3.14. On the
other hand, if the height of the antenna is increased, the volume of antenna will increase
too, same with the hand-held transceiver as a whole. It is therefore desirable to develop
an effective method to expand the bandwidth of the PIFA without increasing the volume.
For the top configuration of the PIFA, the antenna height is constant at 0.03ë. A
maximum bandwidth can be obtained with a length L of approximately 0.42ë, and the
relative bandwidth for a VSWR 2 is greater than 10%. This maximum bandwidth is
very wide despite antenna compactness, since the antenna height h is about 0.04ë. To
obtain a bandwidth larger than 10%, the antenna height in Figure 3.14 must be greater
than 0.08ë. Accordingly, further improvements in antenna bandwidth can be obtained
by selecting appropriate dimensions for the conducting box.
34
Chapter 3 : The Planar Inverted-F Antenna
Ever since the planar inverted-F antenna was introduced, it has found many
applications particularly as a substitute for an external antenna. Because of its
practically small size, it has become very popular with mobile phone manufacturer as
the demand for small mobile system on a rise. With this rapid increase in the ownership
of a personal mobile handset, there is a requirement to bring on an additional frequency
band. Initially, it operates in one frequency, which is the GSM system at 900 MHz. The
additional band used in Europe operates in 1800 MHz, and is known as the DCS1800
system. The introduction of a dual-band system suggests that there is a need for a dual-
frequency handset antenna.
Researches have been done over the past decade by antenna designers all over
the world in search for a small, yet efficient antenna, with the ability to operate in two
frequencies. It is well known that the single-frequency planar inverted-F antenna,
described in early part of this chapter, has good radiation characteristic. Not only it is
compact, it can also exhibits sensitivity on dual polarization, which makes it desirable
as a built-in antenna in most hand-held units where the user holds no fixed orientation
of the handset.
Hence, Liu et al. [9] came out with the design of the dual- frequency PIFA in
1997. Their work has produced a dual-band antenna almost the sa me size as a single -
band PIFA. Basically, it consists of two separate radiating elements, as depicted in
Figure 3.15, with the smaller rectangular radiating element for 1800 MHz and the L-
shaped radiating element for 900 MHz. The mutual coupling between the two radiating
elements is found to be good and isolation acquired is less than –17 dB at both resonant
frequencies [9]. The bandwidths of the antenna are close to those required for the GSM
and DCS1800 systems, at 63 MHz at 900 MHz and 110 MHz at 1760 MHz for VSWR
2 [9]. Radiation patterns of the dual-frequency PIFA are approximately
omnidirectional at both resonant frequencies too.
35
Chapter 3 : The Planar Inverted-F Antenna
36
Chapter 4 : Methodology
CHAPTER 4
Methodology
4.1 Introduction
The second and third goals of this thesis are to perform the characteristic tests of
the selected antenna, with and without the presence of a human head, using the
simulation software, FEKO, and in the University of Queensland Near Field Testing
Range respectively.
The planar inverted-F antenna has been reviewed in Chapter 3 and is chosen as
the antenna under test (AUT). Design and construction of the PIFA is not required as
the University already has the PIFA built by past undergraduate, M.Reidy [10]. A brief
review of his work and measurement on the antenna will be addressed in section 4.2.
The rest of this Chapter will build on the newly acquired software, FEKO, and the
testing of the PIFA in the University’s Near Field Testing Range.
37
Chapter 4 : Methodology
The planar inverted-F antenna is chosen as the antenna under test due to its
desirable features of robustness and superior characteristic, and most importantly, it
suits the criterion of this thesis as the antenna for a hand-held transceiver. Providentially,
the PIFA has been studied by one of the past undergraduate of the University of
Queensland. He designed and built the PIFA as part of the work of his final year thesis
project. As a result, the author need not build another PIFA but use the work of this
student. However, it is still essential to understand the design and performance of this
AUT because it is practically the one to be used for testing in the Near Field Testing
Range.
The AUT is mounted onto the side of a conducting box shown in Figure 4.1. The
dimensions of the AUT are: L1 = 15.3 mm, L2 = 33.0 mm, L3 = 29.9 mm, L4 = 59.6
mm, and g = 2mm. The conducting box is constructed with dimensions of a = 55mm, b
= 136mm, and c = 41mm. Additionally, the height of the PIFA above the ground plane
of the conducting box is measured at h = 8mm and εr = 1.05. The shorting plates are
measured at 8.0 mm each, and the feed pins for the 900 MHz and 1800 MHz radiating
plates are located differently from the one proposed by Liu et al. [9]. Figure 4.2 depicts
the close-up view of the AUT.
For simulation of the antenna, a ground plane is used as the handset instead of
the conducting box, at the dimensions of 41.0 mm by 136.0 mm. The author of [10]
used the commercial software package, Ensemble, for simulating the PIFA. The
magnitude or return loss results of the dual-frequency PIFA is obtained and shown in
Figure 4.3.
38
Chapter 4 : Methodology
Shorting Plates
Conducting Box
Short plates
L3
g
L1
1800 MHz feed
L2
L4
39
Chapter 4 : Methodology
Ä = s[1,1], í = s[2,2]
The result acquired from the simulation reflects that the dual-frequency antenna
radiates at both 900 MHz and 1800 MHz. For each of the frequencies, the return losses
are found to be –36 dB for 900 MHz and –28 dB for 1800 MHz. This results in relative
bandwidths of 11% for the 900 MHz element and 7% for the 1800 MHz.
From far field scanning, the gain radiation patterns for both frequencies are
acquired and are depicted in Figure 4.4. These results are similar to those obtained
individually by Reidy [10]. Nonetheless, from these results attained, it reflects clearly
that the dual-frequency PIFA is indeed feasible and is therefore chosen to be the AUT
for this thesis.
40
Chapter 4 : Methodology
Figure 4.4 : The simulated 2D far field results for the dual-frequency PIFA.
41
Chapter 4 : Methodology
The MoM, which is a full-wave solution that suffers from poor scalability with
frequency, has been hybridised with two asymptotic high frequency techniques known
as the physical optics (PO) and the uniform theory of diffraction (UTD). This
hybridisation enables the solution of much larger (in terms of wavelengths) problems.
The hybridisation will be right if the coupling between the MoM and the PO, or the
UTD, is maintained in the solution. This is necessary if the effect of the PO or UTD
structure on the antenna is to be taken into account.
It is difficult to visualize the operation of this software, but things are not always
that complicated. An example for its application is the changing of the input impedance
of a dipole when it is in close proximity to a large structure. The implementation and
validation of the hybrid approach provides major computational advantages for various
classes of electromagnetic problems.
42
Chapter 4 : Methodology
43
Chapter 4 : Methodology
The graphical user interface allows the author to design the PIFA easily. Using
EditFEKO to input the data of the PIFA dimensions, the design can be created and be
viewed with WinFEKO. Figure 4.5 shows the graphical user interface of WinFEKO and
the final design of the PIFA. Once the design has been completed, a large range of
simulation options are provided on the WinFEKO interface, including s-parameters, 2D
and 3D far fields, 2D and 3D near fields, as well as the frequency range of interest.
Figure 4.5 : An example of WinFEKO and the design of the PIFA using EditFEKO.
44
Chapter 4 : Methodology
45
Chapter 4 : Methodology
The testing of the PIFA was done in the University of Queensland Near Field
Testing Range. The objective of the experiments is to determine the relative power (in
dB) of the AUT mounted on conducting box with and without the presence of a human
head and in different orientations. Because of the lengthy testing time involved and
because people tend not to remain motion free for the duration of test, a model of the
human head has to be constructed. Hence, the human head that was being modelled here
was basically made of a plastic sphere filled with water at a diameter of 250 mm.
Scans were taken at seven different orientations, with and without the human
head model. Front scanning made with the human head model had the AUT mounted on
conducting box placed between the head and the testing probe. Only the back scanning
was done with the human head model positioned between the test probe and the AUT
mounted on the conducting box. An example of the orientation of the back scanning
with human head model is shown in Figure 4.6. Since in practical situation, the user of a
mobile phone tends to hold the transceiver at an inclination angle or sometimes even
horizontally, the horizontal scans were taken too for the back scans. The test probe was
constructed in the form of wire dipoles that was backed by a circular short circuit plate
at a quarter-wavelength from the dipole plane.
During the measurements, the AUT was positioned at 400 mm from the test
probe. Whenever the human head model was required, it should be placed 10 mm from
the AUT. All measurements were accomplished with a Pentium PC, which controlled
the xy-movement of the test probe, and a HP8510 vector network analyser that
measures the transmission coefficient. Every measured data was stored in the Pentium
PC for further processing. The 3D relative power plots were completed by another
powerful software, MATLAB.
46
Chapter 4 : Methodology
Human Head
Model
10 mm
250 mm 140 mm
400 mm
Figure 4.7 : Vertical back scanning with the human head model (top view).
Human Head
Model
400 mm
10 mm
Figure 4.8 : Vertical front scanning with the human head model (top view).
47
Chapter 4 : Methodology
Human Head
Model
10 mm
250 mm 140 mm
400 mm
Figure 4.9 : Horizontal back scanning with the human head model (top view).
48
Chapter 5 : Results and Discussion
CHAPTER 5
5.1 Introduction
This chapter will present the accomplishments of the second and third goals
stated in the obje ctive of the thesis. With the theory of the planar inverted-F antenna
already discussed in the previous chapters, it is now left to the practical examination of
this antenna in accordance with the goals. The first two sections of this chapter provide
the results obtained from software simulation, as well as experimental results from the
University of Queensland Near Field Testing Range. During the experiment, different
orientations of the antenna under test were erected, either on a horizontal plane or a
vertical plane, and not to overlook, with and without the presence of the human head
model, so that comparison can be made between the different results achieved. Also
included in this chapter is the discussion of the results.
49
Chapter 5 : Results and Discussion
After being introduced to the newly acquired software package FEKO , heaps of
work has been done in procuring the knowledge of this new software. As this software
is still a fresh asset of the University, no prior work has been done with it. Theref ore,
many simulations had to be produced with the conventional method known as the trial
and error method. Because running a simulation may take several hours to several days,
huge amount of time has been consumed on waiting. Also, many results came out of the
unexpected and much of it on the error side of the mentioned method, worst if the
simulation came out of several days.
Designs of the simulations were later done more carefully with experience
gradually acquired using the EditFEKO of the software package. Still, results were only
obtained from the simulation of the 1800 MHz radiating element of the computer
generated dual-frequency planar inverted-F antenna. Figure 5.1 and 5.2 depict the 3D
radiation pattern for the 1800 MHz radiating element without the human head and with
the human head respectively.
50
Chapter 5 : Results and Discussion
Figure 5.1 : 3D radiation pattern for the 1800 MHz radiating element
without the human head.
51
Chapter 5 : Results and Discussion
Figure 5.2 : 3D radiation pattern for the 1800 MHz radiating element
with the human head.
52
Chapter 5 : Results and Discussion
The objective of the experiments is to determine the relative power (in dB) of
the planar inverted-F antenna mounted on a conducting box with and without the
presence of a human head. The experiments were carried out in the University of
Queensland Near Field Testing Range. Scans were taken at seven different orientations,
with and without the human head model. This includes scans taken for the horizontal
and vertical plane. Also, all scans were only taken for the 1800 MHz resonant
frequency, while the 900 MHz feed was terminated. This is due to the lack of a viable
reference antenna (the test probe) for the 900 MHz scanning. The measurements taken
from the anechoic chamber were processed by another software package, MATLAB.
This brilliant software has the ability to process and plot three-dimensional colour-
imaging graphs. Hence, the Magnitude (in dB) versus Position plots from the scans
were obtained and shown over the next three pages.
Figure 5.3 describes the way the antenna under test is positioned during the
scans. Shown on the Figure is the positioning of the planar inverted-F antenna for the
front and back scanning. This might not be the actual position of the PIFA, but just a
rough guide for the reader to visualize the orientation during the observation of the
results. The scans were taken from –200 mm to 400 mm on the horizontal x-axis and
from –800 mm to –200 mm on the vertical y-axis. Take note of the positioning of the
1800 MHz radiating plates for a clearer picture and better understanding of the 3D
Magnitude plots.
The magnitudes of the radiation field in the vertical plane during front scanning
of the PIFA, without and with the human head model are shown in Figure 5.4 and 5.5
respectively. When comparing the two magnitude plots, it can be observed significantly
that the introduction of the human head model had affected the radiated field. The dent
in the central portion of the plot in Figure 5.5 seems to be some obstruction of the
radiated field. Obviously, this obstruction only occurs when the human head model is
placed in the scan.
53
Chapter 5 : Results and Discussion
-200
Horizontal front scanning Horizontal back scanning
-400
1800 MHz
radiating plate
y (mm)
-600
-800
x (mm)
Figure 5.6 to 5.9 demonstrate the difference of the magnitude plots for the back
radiation scan. The vertical scans without and with the human head model are shown in
Figure 5.5 and 5.6 respectively. The results do exhibit dissimilarity but the discrepancy
is not huge. Both plots shows a depression in the middle, which is due to the obstruction
of the conducting box, where direct radiation from the PIFA is not possible in that area.
However, the depression for the presence of the human head model seems to be slightly
greater in magnitude. For horizontal scanning of the back radiation, comparison can be
made from Figure 5.8 and 5.9. Once again, the areas of the magnitude difference
between the two plots can be observed. A horizontal plane scan for the front radiation
orientation was acquired too for comparison with the vertical orientation. More will be
discussed in section 5.4.
54
Chapter 5 : Results and Discussion
-40
-60
Magnitude (dB)
-80
-100
-120
-200
400
-400 300
200
-600 100
0
-100
y (mm) -800 -200 x (mm)
Figure 5.4 : Near Field Vertical plane Magnitude plot for Front radiation scanning
without the human head.
-40
-50
Magnitude (dB)
-60
-70
-80
-90
-200
400
-400 300
200
-600 100
0
-100
y (mm) -800 -200 x (mm)
Figure 5.5 : Near Field Ve rtical plane Magnitude plot for Front radiation scanning
with the human head.
55
Chapter 5 : Results and Discussion
-30
-40
-50
Magnitude (dB)
-60
-70
-80
-90
-100
-200
400
-400 300
200
-600 100
0
-100
y (mm) -800 -200 x (mm)
Figure 5.6 : Near Field Vertical plane Magnitude plot for Back radiation scanning
without the human head.
-30
-40
Magnitude (dB)
-50
-60
-70
-80
-90
-200
400
-400 300
200
-600 100
0
-100
y (mm) -800 -200 x (mm)
Figure 5.7 : Near Field Vertical plane Magnitude plot for Back radiation scanning
with the human head.
56
Chapter 5 : Results and Discussion
-40
-50
Magnitude (dB)
-60
-70
-80
-90
-200
400
-400 300
200
-600 100
0
-100
y (mm) -800 -200 x (mm)
Figure 5.8 : Near Field Horizontal plane Magnitude plot for Back radiation
scanning without the human head.
-40
-50
Magnitude (dB)
-60
-70
-80
-90
-100
-200
400
-400 300
200
-600 100
0
-100
y (mm) -800 -200 x (mm)
Figure 5.9 : Near Field Horizontal plane Magnitude plot for Back radiation
scanning with the human head.
57
Chapter 5 : Results and Discussion
-45
-50
Magnitude (dB)
-55
-60
-65
-200
400
-400 300
200
-600 100
0
-100
y (mm) -800 -200 x (mm)
Figure 5.10 : Near Field Horizontal plane Magnitude plot for Front radiation
scanning without the human head.
58
Chapter 5 : Results and Discussion
5.4 Discussion
The results of the experiment were very promising and formed the basis towards
the goals of this thesis. In contrast, although the results obtained from the simulation
were satisfactory, the number of simulations is not sufficient for comparison with the
experimental results. Due to this reason, much emphasis will be based on the
experimental results alone.
During the software simulations, the PIFA was generated with a ground plate
instead of the conducting box. The conducting box was actually designated to be a
component of the simulation, and the author did initially include the conducting box for
simulation. However, problems arose during the simulations and the results turn out to
be disastrous. The incorrect applications of program cards used in the EditFEKO of the
FEKO software or the erroneous input of the codes may have contributed to the
complications. In order to save time in applying the trial and error method, the author
decide to proceed with the ground plate. Nonetheless, the simulated results still prove to
be valuable in comparing between the simulations with the head and without the head.
Figure 5.1 shows the 3D radiation pattern of the PIFA without the presence of a head.
Radiation originates from the 1800 MHz radiating element as we can observe from the
“V” shape curve on the plot. The radiation pattern is unobstructed and has a good
weighting of gain. When a head phantom is introduced into the path of the radiation as
shown in Figure 5.2, it causes a change in the radiation pattern. There is a portion of the
simulated head where the radiation is completely diminished. The gain of the radiation,
however, increases towards the sides of the simulated head and under the ground plate.
It is suggested that the power of the radiation is absorbed in the head, which caused the
barren portion.
The results acquired from the simulations are not as good as it suppose to be.
The lack of skill in performing the software may prove to be costly in the work of this
thesis project. But, on the other hand, the author found out the usefulness of this
software in the aspect of Specific Absorption Rate (SAR) measurement and on the areas
59
Chapter 5 : Results and Discussion
of micro-strip antennas where the Green’s functions can be fully utilized for the analysis
of the multi-layered dielectric and metallic media.
Following the FEKO simulations, huge attention was shifted to the University of
Queensland Near Field Testing Range. Significant works have been done in positioning
the antenna under test (AUT) and the human head model as accurately as possible for
all orientations, which is, for example, the same position of the AUT from the test probe
when measured with or without the head model. This measurement of distance between
each object (see Figure 4.7 to 4.9 for illustration) has to be made cautiously in order to
acquire the best results for comparison. In fact, eleven scans were taken at different
orientations, with and without the human head model. But only seven are chosen to be
included in this thesis. The other four plots, which are discarded, proved purposeless.
The head model were positioned between the AUT and the test probe in the front
radiation scanning, which is impractical because in normal hand-held transceiver, the
antenna is usually placed on the side, back or top of the handset rather than between the
user and the handset. Therefore, it will be pointless to include these plots in this
discussion.
In the front radiation scans or forward radiation scans, as some people call it, the
vertical orientation was positioned, with and without the head model. Figure 5.5
indicates that the presence of the head model did detriment the radiating pattern of the
AUT with the depression shown in the central section when compared with the result
from Figure 5.4. There is also a slight power drop in the presence of the head. The
horizontal scan was taken too but without the presence of the head model. Depicted in
Figure 5.10, the most reddish portion of the plot is the actual position of the 1800 MHz
radiating element. To assist the reader to visualize the positioning of the AUT, a small
patch of redness can be observed from the –600 mm of the y-axis as well, which is the
position of the 900 MHz feed line. Refer to Figure 5.3 for a rough guide on the position
of the 1800 MHz radiating element.
In the back radiation scans (or backward radiation scans), two different
orientation of the AUT and four different scans were made. Figure 5.6 and 5.7 exhibit
60
Chapter 5 : Results and Discussion
the vertical scanning without and with the head model respectively. Without the head
model, the power distribution is almost equivalent mainly on the left side of the x-axis.
The depression may be due to the conducting box, which happens to stand between the
AUT and the test probe. However, that will not do much destruction the radiation
pattern. A drop in relative power can be observed in Figure 5.7 when the head in
included, but with a slight amount. The green dip indicates the significance of the head
model while the depression goes further. This proves that the human head has an effect
on the relative power of the antenna. Because the user does not always hold the handset
vertically during operation, therefore, there is a need to measure the radiation power of
the antenna in an inclination. The author demonstrated this behaviour by performing a
horizontal scan, which is 90 degree to the vertical orientation. After the AUT is shifted
to the horizontal alignment, the AUT was again scanned with and without the head
model. From Figure 5.8 and 5.9, the magnitude versus position plots show that the
presence of the human head model does not provide great detrimental effect to the
relative power of the AUT. On the other hand, increase in relative power can be
observed towards the positive end of the x-axis on the plot with the presence of the
human head. The author suspects the possibility of the feed, for the 1800 MHz radiating
element, touching the conducting box, causing the box to radiate as well. Nevertheless,
the main issue here is that the planar inverted-F antenna radiates soundly in both
orienta tions and the presence of the human head model did little effect to the relative
power of the AUT.
61
Chapter 6 : Conclusions
CHAPTER 6
Conclusions
The rapid increase in the number of mobile phone users contributes to the
expansion of wireless communications especially in the antenna design field. As
antennas are getting physically smaller due to popular market demand, their
performances are not neglected. Human body affects the characteristics of antennas
mounted on portable handset. Therefore, it is indispensable when in designing the
antenna to take the influence of the human body into consideration.
62
References
REFERENCES
63