INVESTIGATIONS ON MICROSTRIP

PEANO LINE ANTENNA
THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE
REQUIREMENT FOR THE DEGREE OF
MASTER OF ELECTRONICS AND TELECOMMUNICATION
ENGINEERING
JADAVPUR UNIVERSITY
(MAY 2011)

BY
ARKAPROVO DAS
UNIVERSITY REGISTRATION NUMBER: 108453 OF 2009-2010
EXAMINATION ROLL NUMBER: M4ETC11-01
CLASS ROLL NUMBER: 000910702001

UNDER THE SUPERVISION OF
PROF. BHASKAR GUPTA

DEPARTMENT OF ELECTRONICS AND TELECOMMUNICATION
ENGINEERING
JADAVPUR UNIVERSITY
KOLKATA – 700032
INDIA


FACULTY OF ENGINEERING AND TECHNOLOGY
JADAVPUR UNIVERSITY

This is to certify that the dissertation titled “Investigations on
Microstrip Peano Line Antenna” has been carried out by Arkaprovo Das
(University Registration No.:108453 of 2009-10) under my guidance and
supervision and be accepted in partial fulfillment of the requirement for the
degree of Master of Electronics & Telecommunication Engineering. The
research results presented in the thesis have not been included in any other
paper submitted for the award of any degree to any other University or Institute.
CERTIFICATE


________________________
Prof. Bhaskar Gupta
Head of the Department
Electronics and Telecommunication Engineering,
Jadavpur University,
Kolkata- 700032.
___________________________________
Prof. Niladri Chakraborty
Dean
Faculty of Engineering and Technology,
Jadavpur University,
Kolkata- 700032.

___________________________________
Prof. Bhaskar Gupta
Project Supervisor
Department of Electronics and Telecommunication Engineering,
Jadavpur University,
Kolkata-700032


FACULTY OF ENGINEERING AND TECHNOLOGY
JADAVPUR UNIVERSITY


CERTIFICATE OF APPROVAL
*
The foregoing thesis is hereby approved as a creditable study of an
engineering subject carried out and presented in a manner satisfactory to
warrant its acceptance as a pre-requisite to the degree for which it has been
submitted.
It is understood that by this approval the undersigned do not endorse or
approve any statement made, option expressed or conclusion drawn therein but
approve the thesis only for the purpose for which it is submitted.
Committee on final examination
for the evaluation of the thesis

________________________
(Signature of the Examiner)



_________________________
(Signature of the Supervisor)

*
Only in case the thesis is approved



FACULTY OF ENGINEERING AND TECHNOLOGY
JADAVPUR UNIVERSITY


DECLARATION OF ORIGINALITY AND COMPLIANCE OF
ACADEMIC ETHICS
I hereby declare that this thesis titled “INVESTIGATIONS ON
MICROSTRIP PEANO LINE ANTENNA” contains literature survey and
original research work done by the undersigned candidate, as a part of his
degree of Master of Electronics and Telecommunication Engineering studies.
All information in this document have been obtained and presented in
accordance with academic rules and ethical conduct.
I also declare that, as required by these rules and conduct, I have fully
cited and referenced all material and results that are not original to this work.


Name : ARKAPROVO DAS
Examination Roll Number : M4ETC11-01
Thesis Title : INVESTIGATIONS ON MICROSTRIP
PEANO LINE ANTENNA


Signature with date :


I would like to take this opportunity to thank a few individuals, directly or indirectly
involved with me, for the successful completion of this project.
ACKNOWLEDGEMENT
First and foremost, I would like to convey my sincere gratitude towards my project
supervisor Prof. Bhaskar Gupta, Head of the Dept. of Electronics & Telecommunication
Engg., Jadavpur University for his constant guidance, support and best wishes, without which
this work would not have been initiated. Most importantly, I would like to thank him for the
faith he has shown in me to take up this work and proceed further. Working with him during
the entire tenure was indeed an absolute pleasure.
I am also grateful and thankful to my teachers at Jadavpur University, to name a few,
Prof. D.R. Poddar, Dr. Sudhabindu Roy, Prof. Asim Kar for their full support and
cooperation during my stay at the institute. Their lectures were immensely beneficial which
helped me strengthen my knowledge base for pursuing research work.
I must convey special thanks to my friend and senior, Mr. Sayantan Dhar, who has
helped me out from all sorts of difficulties that I faced while working on my project. I would
certainly like to mention that all the discussions I had with him on technical issues, helped me
grow my analytical thinking capacity to a great extent.
My whole-hearted thanks to my senior colleagues Ms. Sanghamitra Dasgupta, Mr.
Rajendra Prasad Ghosh, Mr. S. Sankaralingam, Mr. Sudipta Maity, Mr. Avra Kundu and to
my friends Santanu, Amrita, Sonali, Rahul and Ankita for providing their support and
unforgettable moments which I shall certainly cherish throughout my life. I also thank all my
other seniors and colleagues whom I could not mention in this section.
I cannot ever forget the contribution of my father Mr. Anupam Das, and my mother Mrs.
Mina Das towards my proper upbringing and for all the encouragement and support they
rendered me for pursuing higher studies and research. In addition, I would certainly like to
thank my relatives and well-wishers. Lastly, I dedicate this thesis to my grandfather Late
Prafulla Kumar Sarkar, who certainly would have been very happy on seeing the successful
completion of this dissertation.
Above all, I thank the Almighty and pray for everyone’s wellbeing.
Arkaprovo Das

CONTENTS
1. INTRODUCTION 1-2
1.1 Preface 1
1.2 Objective of the thesis 1
1.3 Organization of the thesis 2

2. What is RFID and what are the Antennas Used 3-9
2.1 Outline of the chapter 3
2.2 RFID 3
2.3 Components of RFID systems 4
2.3.1 RFID tag or transponder 4
2.3.2 RFID reader 4
2.4 RFID frequency bands and applications 5
2.5 RFID antennas 5
2.5.1 Near-field systems 6
2.5.2 Far-field systems 7
2.6 Design criteria for RFID tag antennas 8
2.7 Summary 9
References 9

3. An Introduction to Space-filling Curves 10-16
3.1 Outline of the chapter 10
3.2 Fractals 10
3.3 Space-filling curves 10
3.4 Fractal Dimension 12
3.5 Hilbert curves 13
3.6 Peano curves 14
3.7 Process of generating 2nd order Peano curves 15
3.8 Summary 16
References 16

4. Peano Line Antennas 17-24
4.1 Outline of the chapter 17
4.2 Concept of miniaturization 17
4.3 Electrically small antennas 17
4.4 Miniaturization of wire antennas 18
4.5 Peano line wire antennas 18
4.6 Miniaturization of microstrip antennas 19
4.7 Microstrip Peano line antennas 20
4.8 Summary 23
References 23

5. A Brief Review of Peano Line Antenna 25-30
5.1 Outline of the chapter 25
5.2 Advancements in miniaturization of antennas 25
5.3 Review of Peano line antennas 26
5.4 Summary 29
References 29

6. RFID Tag Antenna Design Using Peano Curves 31-57
6.1 Outline of the chapter 31
6.2 RFID frequency bands 31
6.3 Design Goal 31
6.4 Design Procedure 32
6.5 Design prototype 32
6.6 Design approaches 34
6.6.1 Approach 1: Increase of path length of the antenna 35
6.6.2 Approach 2: Higher permittivity substrate 42
6.6.3 Approach 3: Input impedance match using quarter wave
transformer 44
6.6.4 Approach 4: Use of different dielectric materials for feed and
antenna 46
6.6.5 Approach 5: Use of stacked dielectric layers 46
6.6.6 Approach 6: Modification of the structure described in
approach 5 47
6.6.7 Approach 7: Single stub matching 51
6.7 Discussions 56
6.8 Summary 56
References 56

7. Probe Feed Optimization 58-65
7.1 Outline of the chapter 58
7.2 Antenna parameters 58
7.3 Variation in probe feed position 58
7.4 Observations from the optimization process 64
7.5 Radiation Pattern for the fundamental mode 64
7.6 Summary 65

8. Lumped Circuit Model Analysis 66-83
8.1 Outline of the chapter 66
8.2 Brief description of the antennas modeled 66
8.3 Antenna modeling procedure 67
8.3.1 Segmentation of the antenna 67
8.3.2 Modeling of line segments 68
8.3.3 Modeling of right-angled bends 71
8.3.4 Modeling of the open end 72
8.3.5 Cascading of the equivalent circuit 73
8.4 Implementation of the lumped equivalent circuit model 74
8.4.1 Meander line antenna 74
8.4.2 Peano line antenna 78
8.5 Discussions 82
8.6 Summary 83
References 83

9. Conclusion and Scope for Future Works 84-85
9.1 Conclusion 84
9.2 Scope for future research works 84
References 85










Chapter 1

Introduction
Introduction

1
Chapter 1 Introduction
1.1 Preface
In mathematical analysis, a space-filling curve is a curve, which helps in filling up a
two-dimensional space. Giuseppe Peano (1858-1932) was the first to discover one, which is
commonly known as Peano curve. A few space-filling curves, other than the Peano curves are
Hilbert curve, Moore curve, Sierpinski curve etc. These space-filling curves find immense
application in the field of antenna engineering. When the wire itself takes the shape of a
space-filling curve, and radiates in free space, they are referred to as space-filling curve wire
antennas. Again, in case of microstrip antennas, the patch can be meandered by tracing the
shape of space-filling curves, which are then referred to as space-filling curve microstrip
antennas. While extensive research work in terms of antenna performance has been
performed till date in the former category, the latter category is a relatively newer field of
study, and requires greater attention. This thesis deals with some investigations and studies
on microstrip Peano line antennas, where the patch of the microstrip antenna traces the shape
of a Peano curve. This configuration is still newer among the latter category stated above, and
hence calls for extensive research work.

1.2 Objective of the thesis
The past decade has seen phenomenal advances in portable electronics technology like
mobile phones, RFID tags and MP3 players. This has led to the development of System in
Package (SiP) which combines all the necessary components into a single package.
Miniaturization of RF circuit technology has resulted in the need for miniaturized antennas.
Antennas based on space filling curves are mainly used where compactness and
miniaturization are the key objectives. Considering that purpose, microstrip Peano line
antennas have been explored in the following chapters.
This thesis highlights the applications, where these antennas are used. It also provides a
brief theory, both from graphical point of view, as well as from the viewpoint of antennas. It
enumerates in detail the approaches towards a particular design problem, and the variation in
input impedance characteristics with varying probe feed position. Most importantly, a simple
lumped circuit equivalent model of the antenna is derived and presented, which forms the
most important part of the thesis.
Introduction

2
1.3 Organization of the thesis
This thesis is divided into nine chapters. Chapter 2 deals with an introduction to RFID,
and the antennas used for miniaturization. Here the contextual application of Peano line
antennas is also presented.
Chapter 3 gives some insight on the graphical theory of Peano curves, while the basic
theory of Peano line antenna is briefly presented in Chapter 4.
Chapter 5 highlights the historical background of Peano line antennas, and the research
work that has been carried out in this field in the recent years.
The next three chapters describe mainly the research work done during the M.E. project.
Chapter 6 highlights a particular design problem, and the problems faced therein. In
addition, the problems faced are modified with better alternatives, and are studied.
To get further idea on microstrip Peano line antennas, Chapter 7 deals with a few
studies on variation of probe feed location along the antenna, where the patch traces a second
iteration Peano curve.
In Chapter 8, a simple lumped equivalent circuit model of the antenna is derived and
presented. In addition to that, the same for microstrip meander line antenna is also presented,
and a comparison for both the cases is described.
Finally, Chapter 9 summarizes the entire project, with a concluding note on scope for
future works, and a suggestion for a way forward.

















Chapter 2

What is RFID and what are
the Antennas Used
What is RFID and what are the Antennas Used
3
Chapter 2 What is RFID and what are the Antennas Used
2.1 Outline of the chapter
This chapter presents a very brief introduction to RFID technology and the antennas used
for it. The usefulness of miniaturized antennas is also mentioned in this regard. The entire
discussion is qualitative only, and detailed mathematical formulations for this chapter are
beyond the scope of this thesis.

2.2 RFID
RFID is an abbreviation for Radio Frequency Identification, a rapidly growing
technology, which uses RF signals for automatic identification of objects. RFID systems are
short ranged- normally digital, wireless systems. Examples of applications include animal
tagging, asset tracking, electronic passports, smart cards and shop security etc [1]. A
simplified diagram of RFID system is shown in Fig. 2.1.
















Fig. 2.1 A simplified RFID system [1]

What is RFID and what are the Antennas Used
4
2.3 Components of RFID systems
There are two main components of an RFID system. They are:
• Tag or transponder
• Reader

2.3.1 RFID tag or transponder
As shown in Fig. 2.1, at one end of the system is the RFID tag or transponder, a system
that is placed on the objects to be identified. The tag contains an integrated circuit (IC), and
an antenna to receive EM waves from the reader for its identification. The tags have an
Electronic Product Code (EPC) label, which is analogous to the Universal Product Code
(UPC) format used by Bar codes. The tags are also programmable and can contain user-
specific information, i.e. they contain some user memory.
a) Passive tags: These are the tags, which needs to be powered by the RFID reader, i.e. the
fields from the reader serves as energy excitation for these tags to be operational. They are
comparatively cheaper in cost.
Types of RFID tags:
b) Semi-active tags: Semi-active tags have limited power supply, and hence allows faster
operation than the passive tags.
c) Active tags: They have inbuilt battery power supply which makes them functional,
without relying on external aid. They are used in high data rate communications.

2.3.2 RFID reader
An RFID reader or interrogator on the other hand, prompts communication with the tags.
However, contrary to its name, a reader can also write or transfer data to the tag. The reader
is usually connected to a host network or a computer.
a) Mono-static readers: In this case, the transmission and reception of EM signal uses the
same antenna.
Types of RFID readers:
What is RFID and what are the Antennas Used
5
b) Bi-static readers: In bi-static readers, the antennas and the associated circuitry are
different. The scheme is shown in Fig. 2.1, where the second optional antenna is also used.

2.4 RFID frequency bands and applications [2]
• 125-134.2 KHz and 140-148.5 KHz (LF) – can be used globally without a license,
often used in vehicle identification.
• 6.765-6.795 MHz (MF) – inductive coupling.
• 13.553-13.567 MHz (HF) - often called 13.56 MHz band, used for electronic
ticketing, garment tracking, access control, contactless payment etc.
• 26.957-27.283 MHz (HF) – Inductive coupling, used for special applications, bio-
medical applications.
• 433 MHz (UHF) – Backscatter coupling, used for remote car keys.
• 858-930 MHz (UHF) – Restricted usage, used for asset management, container
tracking, baggage tracking, work in progress tracking etc.
• 2.4-2.483 GHz (SHF) – Backscatter coupling, only used in USA/Canada.
• 2.446-2.454 GHz (SHF) – Backscatter coupling, used for long-range tracking and
with active tags, AVI (Automatic Vehicle Identification).
• 5.725-5.875 GHz (SHF) – Backscatter coupling, not so widely used.

2.5 RFID antennas
In general, the antenna for the RFID tag is a balanced one, to suit the differential inputs
of the tag IC. It is a linearly polarized electrically small antenna.
The reader antenna is generally electrically larger than the tag antenna, and may be either
balanced or unbalanced. Reader antennas are often circularly polarized when the orientation
of the tag is unknown. It is because; a circularly polarized wave can be decomposed into two
mutually orthogonal, vertical and horizontal linearly polarized components, and thus
eliminates any likelihood of complete polarization mismatch at the tag. This is important in
tagging applications where high reliability is required. However, there is a 3-dB loss between
What is RFID and what are the Antennas Used
6
circularly and linearly polarized antennas and hence maximum power transfer between the
reader and tag cannot be achieved.
LF and HF systems use near field coupling between the reader and the tag antennas.
Generally, the coupling is inductive; however, it may also be capacitive. When the coupling
is inductive, multi-turn coils are used as tag and reader antennas. The effect is more like that
of a transformer, than that of an antenna, with the coil in the reader serving as the primary
winding, while that in the tag serving as the secondary. The coupling between the antennas is
dependent on their separation, and always occurs in the near-field region. This near-field
coupling method is used in applications, where the tag can be placed on a lossy dielectric
such as an animal or a solid object etc.
Far-field coupling is used generally for UHF systems, where the transfer of energy
between the reader and the tag is determined by Friis equation [3].
2.5.1 Near-field systems
Near-field systems use coil/loop antennas for communication between the reader and the
tag. The reader and tag loop antennas are shown in Fig. 2.2. The reader and tag antennas with
radii a & b respectively are separated by a distance d.














Fig. 2.2 Simplified Reader and Tag loop antennas [1]
What is RFID and what are the Antennas Used
7
The magnetic field produced due to the coil antennas is greatest along the axis of the loop, i.e.
along the z-axis. The tag antenna size and geometry is governed by the application. For
instance, credit card dimensions are often used for smart cards. Multi-turn loops are also
often employed in order to increase the mutual inductance. A typical rectangular tag antenna
is shown in Fig. 2.3, while Fig. 2.4 depicts the photograph of an example of a commercial
coil antenna- used in a European passport.







The antenna in Fig. 2.4 has five turns using thin wires. The IC is connected to the strap prior
to being connected to the coil.
2.5.2 Far-field systems
Systems with carrier frequencies greater than 100 MHz generally operate by transferring
power in the far field. ISM band centered around 2.45 GHz, also uses this method of RFID
communication. A typical far field system is shown in Fig. 2.5. A wide variety of antennas is
possible in the far field systems, though coil antennas are very rarely used. Dipoles, PIFAs
and patches are among the options at the reader. Patches are often used to provide circular
polarization (shown in Fig. 2.5). Modified dipoles are commonly used as tag antennas.








Fig. 2.3 Rectangular Tag coil [1]

Fig. 2.4 Coil antenna used within passport [1]

Fig. 2.5 Typical far field reader and tag antennas [1]

What is RFID and what are the Antennas Used
8
The transfer of energy between the reader and the tag is governed by Friis transmission
equation as stated below [3]
R T
T
R
G G
d P
P
2
4






=
π
λ
(2.1)
where, P
R
is the power received by the tag, P
T
is the power transmitted by the reader, G
T
&
G
R
are the gains for reader (Transmitter) and tag antennas (Receiver) respectively, separated
d distance apart, λ being the wavelength corresponding to the frequency of signal transmitted.
Among the wire and microstrip antenna configurations, meander line antennas, or other
antennas based on space-filling curves are being used lately, in order to facilitate
miniaturization. Those RFID systems, which are used for biomedical applications, require
miniaturized RFID circuitry, and hence affirm the need for miniaturized antennas.

2.6 Design criteria for RFID tag antennas
The requirements that largely determine the design of an RFID tag antenna are
enumerated below [4]:
1. Frequency band: the desired frequency band depends on the regulations formulated by
the country, where the tag is to be used.
2. Size and form: The size and form of the tags depend on the applications, where it is to be
used.
3. Read range: The maximum distance at which the RFID reader can detect the
backscattered signal from the tag is defined as the read range. The read range depends on:
• EIRP: Effective Isotropic Radiated Power is determined by local country regulations.
• Objects: Tag performance changes when placed on different objects.
• Orientation: Read range depends on antenna orientation. Some applications require
tags to have a specific directive pattern, such as omnidirectional or hemispherical
coverage.
4. Applications with mobility: In case of mobile tag applications, Doppler shift must be
accounted for, in the computations.
What is RFID and what are the Antennas Used
9
5. Cost: The antennas should be of low cost, so as to reduce the total cost of the RFID tag.
This puts restriction on the choice of antenna structure, and the choice of materials used for
its construction.

2.7 Summary
A brief account on RFID technology has been portrayed in this chapter. The antennas
used for this technology were also illustrated briefly, and their characteristics were discussed.
As presented in earlier sections, it is evident that, this technology is a growing technology,
due to the applications on which it is used. As this technology does not require Line of Sight
(LOS) propagation, it is replacing the optical Bar code technology very rapidly.

References
[1] Y.Huang and K.Boyle, Antennas from Theory to Practice, John Wiley & Sons Ltd,
2008
[2] http://www.radio-electronics.com/info/wireless/radio-frequency-identification-
rfid/low-high-frequency-bands-frequencies.php
[3] H.T.Friis, “A note on a simple transmission formula”, Proceedings IRE, 41, pp.254-
256, 1946
[4] K.V.Seshagiri Rao, P.V.Nikitin, S.F.Lam, “Antenna Design for UHF RFID Tags: A
Review and a Practical Application”, IEEE Transactions of Antenna and
Propagation, Vol. 53, No. 12, December 2005



















Chapter 3

An Introduction to Space-
filling curves
An Introduction to Space-filling Curves
10
Chapter 3 An Introduction to Space-filling Curves
3.1 Outline of the chapter
This chapter presents a very brief description of the space-filling curves in general. Few
of the space-filling curves are also explained in the text. However, Peano curves constitutes
the most important part of this chapter, since the antennas described in the later chapters, are
patterned on this curve. Process of generating a standard 2
nd
iteration Peano space-filling
curve is also explained.

3.2 Fractals
A fractal is defined as “a rough or fragmented geometric shape that can be split into
parts, each of which is (at least approximately) a reduced-size copy of the whole," [1] a
property called self-similarity. In other words, fractal forms a part of non-Euclidean
geometrical shapes, that is repeated at even-smaller scales to produce irregular shapes and
surfaces that cannot be represented by classical geometry. The term fractal was coined by
Benoit Mandelbrot in 1975 and was derived from the Latin word fractus meaning "broken" or
"fractured" [2]. A fractal shape is generated using recursive algorithms. Two important
properties of fractal shapes are self-similarity and scale-invariance. Fractals consist of
identical and similar elements repeated in different magnifications, orientations and positions,
often in inter-connected fashion to obtain the final structure. One important feature of fractal
geometry in 2-dimension is, after infinite numbers of recursive iterations, the length of the
curve traced becomes infinite, yet the curve occupies a finite area. This point is explained
later.

3.3 Space-filling curves
Space-filling curves are subset of a broader class of fractals. As the name suggests, these
curves effectively aid in filling up a particular space (in 2-D), or a particular volume (in 3-D),
as the iteration order increases. In simpler words, the property of a space-filling curve may be
explained as: if a particular geometrical curve occupies a definite area in 2-dimensional space
in its 1
st
iteration, then in higher iterations, it occupies the same area, with reduced
magnification, and many repetitive patterns. Thus, as the iteration order increases, the curve
An Introduction to Space-filling Curves
11
traces more path, yet confined in a definite area, and helping in compression. The repetitive
pattern brings in the concept of self-similarity as described earlier.
Examples: Some of the common types of space-filling curves are Moore curves, Peano
curves, Hilbert curves, Wunderlich curves etc. The first six iterations for Hilbert curve are
shown in Fig. 3.1.












Fig 3.1 Six iterations of the Hilbert curve construction [3]

Construction: The first iteration is the unit cell, the repetition of which generates the higher
iterations. As seen from the figure above, the first iteration curve is used in the second
iteration with reduced size, either rotated by 90 degrees clockwise/anticlockwise, or keeping
it intact. The unit cell of first iteration is repeated 4 times (2- rotated, 2- intact) to fill the
same space in second iteration. Additional line segments are added in order to ensure
continuity of the curve. Now, this second iteration Hilbert curve is the unit cell for the third
iteration. This process is a recursive one, and goes on with increasing iteration orders.



1
st
iteration 2
nd
iteration 3
rd
iteration
4
th
iteration 5
th
iteration 6
th
iteration
An Introduction to Space-filling Curves
12
3.4 Fractal Dimension
Fractal dimension is an important parameter to quantify the space-filling ability of a
space-filling curve. For a space-filling curve to effectively fill a 2-D space, the fractal
dimension must approach 2. Similarly, the same parameter must approach a value of 3 so as
to effectively fill a 3-D volume.
To illustrate the above point, some examples are considered.
Let us consider a line segment of length l. When the line is equally halved, each of the
segments is of length l/2. Alternatively, each of those smaller line segments when magnified
by a factor 2, yields the original line segment. Similarly, if a line segment is divided into
three parts, each of them may be magnified by a factor of 3 to get the original segment.
Therefore, in general, if a line is divided equally into ‘n’ self-similar smaller line segments,
each one of them when magnified by a factor ‘n’ yields the original line segment.
Let us now consider the case of a square, with sides of length l. If we now assume a smaller
square with each side of length l/2, then 4 such smaller squares are required to fill in the
space by the original bigger square. Thus, the bigger square can be divided into 4 smaller
self-similar squares, each with magnification 2. Similarly, if l/3 be the length of each side of
the smaller squares, then it would require 9 such squares to fill the bigger square of length l.
For this case, we can say each of the smaller 9 squares have a magnification factor 3, i.e. if
the sides of any of the 9 smaller squares be trebled, it would produce the bigger square.
Similar examples may be taken for the case of cubes.
For the first case of a straight line, it can be divided equally into any numbers of smaller self-
similar segments. Let it be n. Hence, it can be expressed as unit power of n,
i.e. n = n
1
,
or, more generally, n = N
D
, where for this case, N = n, and D = 1.
In the above expressions, n is the number of smaller line segments, N is the magnification
factor, and D is the dimension.
D = 1 indicates, a straight line is one-dimensional.
The second case of a square suggests that, the bigger square can be divided equally into n
number of smaller self-similar squares, n being a perfect square number. As seen above, it
can be divided into 4, 9, 16 etc. number of smaller squares.
An Introduction to Space-filling Curves
13
So, if a square is divided into 9 smaller squares, then
9 = 3
2
, where n = 9, N = 3, D = 2
Which clearly suggests, the dimension of a square is 2, i.e. it occupies a definite area in 2-
dimensional plane.
Therefore, for any general D-dimensional space, the following relation holds true:
n = N
D
(3.1)
For regular geometries, D is the dimension, while for fractal geometries; D is defined as
‘Fractal Dimension’.
Taking logarithm on both sides of (3.1), we get,
log n = D log N,
or, D = log n / log N (3.2)
Therefore, dimension or fractal dimension is defined as the ratio of the logarithm of number
of self-similar pieces to the logarithm of the magnification factor.
For a Koch fractal curve, the maximum attainable fractal dimension D = 1.26, while for a
Hilbert curve, maximum value of D = 2. This clearly suggests that, as the iteration number of
a Hilbert curve increases, it fills the 2-D space more effectively than a similar order Koch
curve.

3.5 Hilbert curves
The basic pictorial representation of Hilbert curves is depicted in Fig. 3.1. In this section,
the values of the dimensions of its line segments are provided (Fig. 3.2).
Dimensions:







(a) Iteration - 1


(b) Iteration - 2

Fig. 3.2 Hilbert curve


l
d = l
d=l/3
l
An Introduction to Space-filling Curves
14
Let l be the side dimension of the Hilbert curve, with d being the length of each smallest unit
of line segment. Let the total length of the curve be S, and n be the iteration order.
Then, for a Hilbert curve,
1 2 −
=
n
l
d ; and
S = ( 2
2n
– 1 ) d = ( 2
n
+ 1 ) l

3.6 Peano curves
The first space-filling curve was discovered by Giuseppe Peano in the year 1890. It was
named after him as Peano curve. There are different types of Peano curves – the standard first
Peano curve, minimal N-shaped curve etc. The first Peano curve is the most commonly used
one for antenna design purposes, and it is shown in Fig. 3.3. A Peano curve has a fractal
dimension of the value 2.








Dimensions: For a Peano curve, the dimensions of the curves are given as follows:
1 3 −
=
n
l
d ; and
S = ( 3
2n
– 1 ) d = ( 3
n
+ 1 ) l
Comments: If we observe the length dimensions for the Hilbert and Peano curves, we find
that for a definite footprint area and fixed iteration order, the path length S covered by a
Peano curve, is greater than that covered by a Hilbert curve. This suggests that, a Peano curve
offers more compression and space-filling property than a similar Hilbert curve. This

(a) First iteration (b) Second iteration (c) Third iteration
Fig. 3.3 Peano curves [3]

An Introduction to Space-filling Curves
15
contrasting property is beneficial, especially in case of antenna design, which asserts Peano
line antennas as better candidates for miniaturization.

3.7 Process of generating 2
nd
order Peano curves
Step 1: First, consider the first iteration, and its mirror image. They are depicted below:





Step 2: Reduce the scale of the size of first iteration curve and its mirror image
Step 3: Divide the square space to be filled (i.e. space containing the first iteration curve) into
9 equal squares.






Step 4: Go on filling the smaller squares in the numbered order stated above, alternatively
with the reduced Peano 1
st
iteration curve and its mirror image. The square number 1 contains
reduced 1
st
iteration curve, square number 2 contains its mirror image, and so on, till we
arrive at square number 9





Original Curve Mirror image







1
2
3 4
5
6
9
8
7

An Introduction to Space-filling Curves
16
Step 5: Add extra line segments to ensure the continuity of the curve. The generated curve is
the second iteration Peano curve. The same process, taking the second order Peano curve and
its mirror image for a unit cell, can be repeated to produce a third order Peano curve. This
process is a recursive one, and can be applied to generate any higher iteration.







3.8 Summary
This chapter presented an overview of the space-filling curves in general and its
concepts, while emphasizing mainly on the Hilbert and Peano curves. Also the process of
generating a second order Peano curve has been discussed, that can be applied to generate
any higher order of the same curve.

References
[1] B.B. Mandelbrot, The Fractal Geometry of Nature, New York: W.H. Freeman & Co.,
1983
[2] http://en.wikipedia.org/wiki/Fractal
[3] http://en.wikipedia.org/wiki/Space-filling_curve


















Chapter 4

Peano Line Antennas
Peano Line Antennas

17
Chapter 4 Peano Line Antennas
4.1 Outline of the chapter
In chapter 3, a brief idea on Peano space-filling curves has been presented. This chapter
presents the application of the same curve in antenna engineering. In that context, the basic
theory regarding miniaturization of antennas is also discussed.

4.2 Concept of miniaturization
Miniaturization of an antenna can be thought of from two different viewpoints. Although
they mean the same thing, yet they may be considered separately:
• Frequency Reduction: Compared to a particular conventional antenna design, if
any other design can reduce the resonant frequency of the antenna, yet not
exceeding the total physical size, then the new design can be considered to be as
a miniaturized one.
• Size miniaturization: The above concept can be rephrased in a different manner.
Since the design of a resonant antenna depends on the effective wavelength,
hence a lower frequency antenna design requires the physical size of the antenna
to be much larger. However, as explained in last point, if the antenna can realize a
lower resonant frequency, without exceeding the antenna size, then it certainly
aids in size reduction, as compared to the conventional design at lower frequency,
where in terms of effective wavelength, the size is much larger.

4.3 Electrically small antennas
An antenna is said to be small when its size is much smaller than the operating
wavelength. Small Antenna Limit (SAL) determines the criteria for an antenna to be
electrically small. It is defined as the upper frequency boundary, at which an antenna may be
considered as electrically small [1]. In the context of SAL, Wheeler defined an electrically
small antenna as one whose maximum dimension is less than λ/2π [2]. This relation is often
expressed as:
ka < 1 (4.1)
Peano Line Antennas

18
where, k = 2π/λ, λ is the free space wavelength, a is the radius of a sphere circumscribing
the maximum physical dimension of the antenna.
A sphere of radius a = 1/k = λ/2π is defined as the radiansphere.
Therefore, when an antenna can be enclosed into a radiansphere of radius a, it is said to be an
electrically small antenna.
From (4.1), it can be inferred that, an antenna is electrically small for a particular frequency
of interest. It follows directly that, a frequency reduction, causes an antenna to be electrically
small, thus helping in miniaturization.

4.4 Miniaturization of wire antennas
Wire antennas can be miniaturized if the antenna length of a conventional dipole or
monopole design is extended following a particular meandering pattern, called meander line
antenna, or by following any other space-filling curve. The compression property of space-
filling curves has been discussed in chapter 3. Thus, when
the wire traces a space-filling curve, the current path over it
follows an extended path, and hence causes a reduction in
the resonant frequency of the antenna, without exceeding
the physical size limits. One particular simple
configuration of a meander line wire antenna is shown in
Fig. 4.1. In the figure, the wire takes meandering turns to
reduce the resonant frequency, and the whole antenna is
placed over a ground plane in a monopole like fashion.

4.5 Peano line wire antennas
In Fig. 4.1, if the meander pattern of the wire is replaced by a Peano curve pattern, then it
forms a Peano line wire antenna. As shown in chapter 3, the Peano curve may be of any
iteration order. As the iteration order increases, the space-filling property of the curve
increases, and hence it results in more frequency reduction for the antenna, leading to
miniaturization. Though these space-filling wire antennas result in miniaturization, yet they
suffer from certain drawbacks. When used as a vertical monopole antenna, just like that

Fig. 4.1 Meander line wire antenna
Ground
Plane

Peano Line Antennas

19
shown in Fig. 4.1, these antennas require a large horizontal ground plane, which eventually
increases the total size of the package, and thus hampers our original goal for miniaturization.
Secondly, the antenna is fed at the end as a monopole, and hence, the input impedance cannot
be changed by changing the feed position. Therefore, the problem of impedance matching is a
serious concern. For these reasons, we switch over to microstrip configuration antennas,
where the space-filling property of the curves is utilized for miniaturization.

4.6 Miniaturization of microstrip antennas
With growing demands for miniaturized circuitry in today’s world, smaller and cheaper
antennas are required to be integrated into electronic devices. This calls for the introduction
of miniaturized planer antennas in most of the applications. The microstrip patch antennas
have a significant number of advantages over conventional antennas. Nevertheless, the length
of the patch required for an operation at lower frequency range at its fundamental mode is too
high, since the length is of the order of half of the effective wavelength. In such cases,
different methods are to be employed to reduce the size of the antennas. They are:
• Use of high permittivity substrate: Rectangular patch antennas are operated at a
resonant frequency f
r
, where f
r
is given by [3]:
r
r
L
c
f
ε 2
= (4.2)
where c is the speed of light in free space, L is the length of the patch and ε
r
is the
permittivity of the substrate used. Thus (4.2) implies that the resonant length L is
proportional to
r
ε
1
at a fixed resonant frequency. Hence, a high permittivity
substrate results in smaller antenna length.
• Use of shorting pins: Use of edge shorted patch, or shorting pins at the vicinity
of a coaxial feed are well known techniques for size reduction of a microstrip
antenna [4-7]. The shorting post or pin adds an extra path for the current to flow,
thus increasing the magnetic field associated with it, which effectively increases
the inductive effect. This reduces the resonant frequency of the antenna.
Peano Line Antennas

20
• Insertion of slots: Inserting suitable slots in the radiating patch is also an
important technique for reducing the dimensions of the patch antenna [8-9]. The
slots introduce parasitic capacitance, and extended path for surface current, which
tends to reduce the resonant frequency of the antenna.
• Use of space-filling curve: Miniaturization can also be effected by the
introduction of space-filling curves [10-12]. If the patch of the microstrip antenna
traces a space-filling curve without increasing the original physical footprint area,
then it greatly helps in frequency reduction, and hence miniaturization. This
technique is used in the thesis, and the space-filling curve that is implemented in
microstrip form is the 2
nd
order Peano curve.

4.7 Microstrip Peano line antennas
Fig. 4.2 shows the geometry of a microstrip Peano line antenna, while the photograph of
a fabricated Peano line antenna is shown in Fig. 4.3.







The patch of the microstrip antenna traces a second order Peano curve. The fabricated
photograph shows a Peano line antenna fabricated on a Duroid substrate and separated from
the ground plane using foam.
Frequency Reduction:
The convoluting nature of the patch as shown in figures above suggests that the surface
current that flows over it follows a greatly meandered path. Due to the extended path over
which the surface current flows, as compared to its path on a rectangular patch with similar
footprint area, extra inductive effect comes into picture. In addition, the closely placed

Fig. 4.2 Geometry of Microstrip
Peano line antenna [13]
© IEEE 2007

Fig. 4.3 A fabricated 2
nd
order Peano line antenna [13]
© IEEE 2007
Peano Line Antennas

21
adjacent lines facilitates electromagnetic coupling between them, which accounts for extra
capacitive effect as well. As the inductive and capacitive effects increases, the resonant
frequency (f
r
) of the antenna falls, following the relation:
LC
f
r
π 2
1
= (4.3)
provided, the effective inductance L and capacitance C of the antenna equivalent circuit can
be represented as a series resonant circuit.
Radiation Mechanism:
The radiation from the antenna mainly occurs due to the discontinuities inherent within
the structure itself. As the current over the antenna traverses a convoluted path, the electrons
or the charges encounter sharp discontinuities, where they are accelerated or decelerated. This
causes radiation to occur from the discontinuities. In this context, it is noteworthy that the
fringing fields from both the open ends contribute very little to the effective radiation. It is
because, the width of the microstrip line forming a Peano line antenna is very small, and thus
provides very small radiating edges for the fringing fields to occur. This is the main reason
for these antennas not complying with the conventional design rule of l = λ
eff
/2, where l is the
length of the antenna, and λ
eff
the effective wavelength. The above rule is valid only for
microstrip rectangular patch antenna, where the resonance occurs for an antenna length equal
to half the effective wavelength. When the length of the patch follows the above rule, the
fringing fields from both the open ends is exactly oppositely directed. The vertical
components of those fields cancel out each other, leaving only the horizontal components as
effective radiating slots separated by a distance of half the effective wavelength, resulting in
maximum radiation in the broadside direction. But in case of Peano line antennas, radiation
occurs from the discontinuities and not due to the fringing fields, and therefore the design
rule is not followed. This fact is verified in Chapter 6 by the simulation results. The radiation
pattern of this antenna is depicted in Chapter 7.
Comparison with Hilbert antennas:
In Chapter 3, it has been mentioned that since the path length of the curve as occupied by
a 2
nd
order Peano curve is greater than that covered by a 2
nd
order Hilbert curve, therefore
Peano curves provide better compression properties. Thus when the patch follows a 2
nd
order
Peano curve, the surface current on it traverses more path length than that caused due to a 2
nd

Peano Line Antennas

22
order Hilbert curve within the same footprint area. Therefore, microstrip Peano line antennas
provide better miniaturization and more frequency reduction properties than a similar Hilbert
antenna.
Comparison with Meander line antennas:
As far as miniaturization is concerned, both the antennas almost provide similar
miniaturization characteristics. However, a comparison may be drawn considering the
structure of the two. The structure of only the patches for microstrip meander line antenna,
and microstrip Peano line antenna has been
shown in Fig. 4.4. Both the antennas cover
similar path lengths, and occupy equal
footprint area. The 3-dimensional figures are
not shown for simplicity. As evident from the
figures, there are more numbers of
discontinuities in Peano line antennas, than in
meander line antennas. So if the radiation
from the discontinuities add up
constructively, Peano line antenna can
provide more radiation gain than the corresponding meander line antenna of same antenna
length and same footprint area. The work of Fukusako et al. [11] reported such an
observation.
Another important point in this regard is that in case of meander line antennas, the
currents in the parallel adjacently placed line segments being oppositely directed helps in
cancellation of radiation due to them at the far field, while the smaller line segments
comprises of the effective radiating elements. However, by virtue of its structure, complete
cancellation of radiation due to parallel adjacent line segments is impossible in case of Peano
line antennas.
Quality factor, Bandwidth & Radiation efficiency:
• Due to closely placed line segments, there is a huge amount electromagnetic
coupling between them, which helps in storage of a lot of energy in the near field
of the antenna. This enhances the quality factor (Q) of the antenna

(a)

(b)
Fig. 4.4 (a) Microstrip meander line antenna
(b) Microstrip Peano line antenna
Peano Line Antennas

23
• As bandwidth is inversely proportional to the quality factor, therefore microstrip
Peano line antenna suffers from lower bandwidth. This is the case with all
antennas based on space-filling curves.
• Application of space-filling curves, effectively increases the antenna path length,
hence increases the conductor loss of the antenna. This enhances the loss
resistance of the antenna, which results in reduction of antenna radiation
efficiency.

4.8 Summary
This chapter explained the basic principles of miniaturization, with an emphasis on
electrically small antennas. The focus has been put mainly on Peano line antennas for
miniaturization, and in that context the basic principles of Peano line antennas has been
discussed. As described in the chapter, Peano line antennas are advantageous in terms of
frequency reduction as compared to Hilbert antennas, and in terms of radiation gain, as
compared to meander line antennas. However, they suffer from drawbacks, which is common
to all antennas based on space-filling curves.

References
[1] S.R. Best, “A Discussion on the Properties of Electrically Small Self-Resonant Wire
Antennas”, IEEE Antennas and Propagation Magazine, Vol. 46, No. 6, December
2004
[2] H.A. Wheeler, “Fundamental Limitations of Small Antennas”, Proceedings IRE,
pp.1479-1484, December 1947
[3] R. Garg, P. Bhartia, I. Bahl, A. Ittipiboon, Microstrip Antenna Design Handbook,
Arctec House Inc., 2001
[4] N. Fayyaz, E. Shin, S. Safavi-Naeini, “A novel dual-band patch antenna for GSM
band”, Antennas and Propagation for Wireless Communications, pp. 156–159, 1998
[5] R. Waterhouse, “Small microstrip patch antenna”, Electron. Lett. 31, 604–605, April
13, 1995
[6] S. Dey and R. Mittra, “Compact microstrip patch antenna”, Microwave Opt. Technol.
Lett. 13, 12–14, Sept. 1996
[7] M. Sanad, “Effect of the Shorting Posts on Short Circuit Microstrip Antennas”, IEEE
International Symposium on Antennas and Propagation, Vol. 2, pp.794-797, 1994
[8] G. Kosiavas, A. Papiernik, J.P. Boisset, M. Sauvan, ‘‘The C-Patch: A Small
Microstrip Element’’, Electronics Letters, Vol. 25, pp. 253–254, 1989
[9] K. L. Wong and K. P. Yang, “Small dual-frequency microstrip antenna with cross
slot”, Electron. Lett. Vol. 33, pp.1916–1917, Nov 1997
Peano Line Antennas

24
[10] H.Y. Wang and M.J. Lancaster, “Aperture-Coupled Thin-Film Superconducting
Meander Antennas”, IEEE Transactions on Antennas and Propagation, Vol.47, No.5,
May 1999
[11] T. Fukusako, T. Terada, K. Iwata, “Design and comparative study on planar small
antennas using meander and peano line structure”, International Symposium on
Antennas and Propagation Society, pp. 2451-2454, 2007
[12] X. Chen, S.Safavi-Naeini, Y. Liu, “A Down-Sized Printed Hilbert Antenna for UHF
Band”, IEEE Antennas and Propagation Society International Symposium, Vol. 2,
pp.581-584, 2003
[13] J. Mc Vay and A. Hoorfar, “Miniaturization of top-loaded monopole antennas using
Peano-curves”, Radio and Wireless Symposium, IEEE, pp. 253-256, 2007


















Chapter 5

A Brief Review of Peano Line
Antennas
A Brief Review of Peano Line Antenna

25
Chapter 5 A Brief Review of Peano Line Antenna
5.1 Outline of the chapter
This chapter briefly highlights the research works that has been carried out in the field of
Peano line antennas over the years. As mentioned in Chapter 1, this field is a relatively newer
field of study in the branch of antenna engineering. The chronological advancements in the
development of Peano line antennas have been elucidated in the following sections.

5.2 Advancements in miniaturization of antennas
The methods for miniaturization of antennas have been discussed in the previous chapter.
The electrical size of an antenna has been an important field for research work, right from
1947, when H.A.Wheeler published a paper on small antennas [1]. In the very next year, L.J.
Chu in his journal formulated a fundamental limit, popularly known as Chu limit [2]. Also,
there has been several other researchers in this field over the years [3-4]. In addition, a lot of
thought process on miniaturization techniques has taken place. Some of them have been
discussed in Chapter 4, and the review of those works is not the focus of attention in this
thesis.
For miniaturization purposes, the concept of fractals is also introduced in the field of
antennas. The wire may follow different fractal shapes like Koch, Minkowski etc. [5] to
reduce the resonant frequency of the antenna. In addition, fractal geometry can also be
applied to microstrip configuration, where the edges of the patch are protruded in the form of
the fractal shapes, to result in miniaturization.
Nevertheless, the most popularly used antenna for these purposes are the meander line
antennas [6-7]. They may be in wire configuration, or in microstrip form. This concept paved
the way for the introduction of graph theory in antenna engineering. Therefore, space-filling
curves like Hilbert curves, Moore curves, Wunderlich curves, Peano curves etc. were thus
studied in greater detail. Although, Hilbert curve antennas [8-9] both in wire configuration
and in microstrip form have attracted several researchers due to its simplicity, the others are a
bit less explored. Hence, it calls for more and more studies on the other space filling curve
antennas. The advancements in the field of Peano line antennas is described next, while the
review of other space filling curve antennas are beyond the scope of this thesis.
A Brief Review of Peano Line Antenna

26
5.3 Review of Peano line antennas
Although Giuseppe Peano discovered the Peano curve a long time earlier (1890), yet it
was a late introduction in the field of antenna engineering. Though there were a few papers
on broadband arrays using Peano-Gosper curve, yet the standard first Peano curve variant
was not yet implemented.
It was in the year 2003 that Jose M. Gonzalez-Arbesu et al. investigated the effectiveness
of space-filling curves as small antennas [10]. The importance of space-filling geometries, as
optimal or efficient curves for small antenna design was assessed in this work using bi-
dimensional wire monopoles. Several variants of space-filling curves were taken in the form
of monopole wires in his comparative study, which included three different variants of Peano
monopoles in first two iterations. The important conclusion was, though they were suitable
for reaching higher miniaturization ratios than conventional quarter-wave monopoles, they
store a lot of energy in the near field of the antenna and have higher ohmic losses, resulting in
high quality factor and lower radiation efficiencies.
In 2004, J.Zhu et al. [11] investigated the characteristics of a single antenna made of thin
wire, patterned after Peano space-filling curve. The same team had earlier worked on similar
investigation for Hilbert curve wire antennas. This study mainly focused on the feed point
effects, current distribution pattern, cross polarization, impedance bandwidth and other
radiation characteristics. The study revealed that when the feed point was placed at the point
of symmetry, the real part of input impedance was very low. However, it was possible to
locate an off-centre feed, where there was a perfect 50-ohm or 75-ohm input impedance
match. Though the size reduction was substantial, yet the impedance bandwidth was quiet
poor as compared to similar order Hilbert antenna. The magnitude of the current distribution
was reported to be almost symmetric, while the phase varied very slowly along the wire
length. The radiation pattern of the antenna at its fundamental resonant frequency in the
planes of Φ = 0° and Φ = 90° almost resembled the patterns of a linear dipole. Other than
frequency reduction, the main advantage of this antenna compared to Hilbert antenna was
lower cross polarization level. Lastly, the radiation efficiency was reported to be decreasing
with increasing iteration order.
In the same year, X.Chen et al. reported the first work on printed Peano line antenna
[12]. This work was a generalized one on a few space-filling printed antennas, the Peano
antenna being one of them. The input impedance and current distribution of all the antennas
A Brief Review of Peano Line Antenna

27
were calculated using Method of Moment (MOM) with Mixed Potential Integral Equation
(MPIE) formulation. The resonant frequencies were obtained by searching for the roots of the
imaginary part of input impedance. The feed structure was modeled as a delta-gap voltage
source. Among the antennas modeled, the schematic structure of the patch of Peano line
antenna, which was taken for analysis, is shown in Fig. 5.1a. The magnitude and phase of the
current distribution along the same antenna is depicted in Fig. 5.1b.






Fig. 5.1 (a) Printed 2
nd
order Peano antenna (b) Current distribution along the antenna [12]
© IEEE 2004

J. Mc Vay and A. Hoorfar in their paper [13] in 2006 investigated the characteristics of
utilizing planer metallic path in the form of Peano curve and Hilbert curve to provide top
loading properties to electrically short monopole antenna elements. This antenna focused on
planer space-filling curve configuration with a much lower profile, and utilized probe feed
and shorting posts so that it could be matched to a 50-ohm source over a wide bandwidth.
Fig. 5.2 shows the schematic diagram of the antenna, VSWR characteristics and radiation
pattern.






Fig. 5.2 (a) Geometry of the antenna (b) VSWR characteristics (c) Gain patterns [13] © IEEE 2006


(a)

(b)

(a)

(b)

(c)
A Brief Review of Peano Line Antenna

28
The role of shorting posts was to reconfigure the structure into radiating like a matched top-
loaded monopole with vertical polarization.
In 2007, E.El-Khouly et al. [14]
proposed a high directivity antenna using
Peano space-filling curve. This was the first
instance of a broadband, high gain Peano line
antenna, but at the cost of greater size
dimensions. The antenna was proposed in
microstrip configuration, with three different
designs. In the three designs, gains of 15.7dB,
17.2dB and 17.5dB were obtained. The
antenna used six 180°-phase shifters to orient the current distribution throughout the antenna,
over same direction. As all the line segments had current distribution oriented over same
direction, the directivity of the antenna increased substantially. The proposed design is shown
in Fig. 5.3.
In the same year, J.Mc Vay and A. Hoorfar again published a paper on Peano line
antenna and Hilbert curve antenna [15], this time the work being based on microstrip
configuration with dielectric substrates, unlike their previous work in 2006. The Peano
antenna geometry and the fabricated photograph of the antenna are shown in Fig. 4.2 and Fig.
4.3 respectively in chapter 4. A gain of about 4.6dBi was obtained using this design. The
VSWR characteristics and the radiation pattern for the Peano line antenna are shown in Fig.
5.4.









Fig. 5.3 Proposed antenna design [14] ©IEEE 2007

(a) (b)
Fig. 5.4 (a) VSWR characteristics (b) Radiation pattern [15] ©IEEE 2007
A Brief Review of Peano Line Antenna

29
However, the first RFID
design specific work in the field
of microstrip Peano line antenna
was reported in the year 2007,
by a team of Fukusako et al.
from Kumamoto University,
Japan [7]. The work reported a
design of meander line as well
as Peano line antenna for a
centre frequency of 1GHz, for
RFID applications. The most
interesting aspect in this design was the application of a second order Peano curve, but in a
modified form. Normal second order Peano curve was not used; instead, an extended version
to the second order curve was implemented. This enhanced the miniaturization of the antenna
largely, which occupied a footprint area of as small as λ/50 X λ/100. The antenna structure
and its measured VSWR characteristics are shown in Fig. 5.5. The details of this antenna are
considered in Chapter 6, where it is used as the design prototype for a particular design
problem.
A similar design was proposed by a team of T.Terada et al. [16] from Kumamoto
University, Japan in the year 2008. The team fabricated and measured the impedance and
radiation characteristics of a small and low profile printed antenna using Peano line. An
antenna gain of -7.9dBi was reported, and the polarization was investigated to be elliptical.

5.4 Summary
In this chapter, a brief historical background of Peano line antennas has been presented.
Although Hilbert curve antenna has been explored by many researchers, very few papers
have been reported in the literature, which are devoted to Peano line antennas. Both from
design as well as analysis point of view, this branch is quiet open for the researchers willing
to make a contribution in this field.

References

(a)

(b)
Fig. 5.5 (a) Antenna structure (b) VSWR characteristics [16]
©IEEE 2007
A Brief Review of Peano Line Antenna

30
[1] H.A. Wheeler, “Fundamental Limitations of Small Antennas”, Proceedings IRE,
pp.1479-1484, December 1947
[2] L. J. Chu, “Physical limitations on omni-directional antennas,” J. Appl. Phys., vol. 19,
pp. 1163–1175, Dec. 1948
[3] H.A. Wheeler, “Small Antennas”, IEEE Transactions on Antennas and Propagation,
Vol. AP-23, No. 4, pp.462-469, 1975
[4] R.C. Hansen, “Fundamental limitations in Antennas”, Proceedings IEEE, Vol.69,
No.2, Feb.1981
[5] J.P. Gianvittorio, Y.R-Samii, “Fractal Antennas: A Novel Antenna Miniaturization
Technique, and Applications”, IEEE Antenna’s and Propagation Magazine, Vol. 44,
No. 1, February 2002
[6] H.Y. Wang and M.J. Lancaster, “Aperture-Coupled Thin-Film Superconducting
Meander Antennas”, IEEE Transactions on Antennas and Propagation, Vol.47, No.5,
May 1999
[7] T. Fukusako, T. Terada, K. Iwata, “Design and comparative study on planar small
antennas using meander and peano line structure”, International Symposium on
Antennas and Propagation Society, pp. 2451-2454, 2007
[8] X. Chen, S.Safavi-Naeini, Y. Liu, “A Down-Sized Printed Hilbert Antenna for UHF
Band”, IEEE Antennas and Propagation Society International Symposium, Vol. 2,
pp.581-584, 2003
[9] J. Zhu and A. Hoorfar, “Bandwidth, Cross-Polarization, and Feed-Point
Characteristics of Matched Hilbert Antennas”, IEEE Antennas and Wireless
Propagation Letters, Vol.2, 2003
[10] J. M. Gozalez-Arbesu, S. Blanch, and J. Romeu, “Are space-filling curves efficient
small antennas?”, IEEE Antennas Wireless Propagat. Lett., vol. 2, pp. 147–150, 2003
[11] J. Zhu and A. Hoorfar, “Peano Antennas”, IEEE Antennas and Wireless Propagation
Letters, Vol.3, 2004
[12] X. Chen, Y. Liu, S. Safavi-Naeini, “Printed Plane-Filling Fractal Antennas for UHF
Band”, IEEE International Symposium on Antennas and Propagation Society, vol. 4,
pp. 3425-3428, 2004
[13] J. McVay and A. Hoorfar, “A Miniaturized Planar Space-filling Curve Antenna with
Wideband Monopole-like Radiation Characteristics”, IEEE Antennas and
Propagation Society International Symposium, pp.3723-3726, 2006
[14] E.El-Khouly, H. Ghali, S.A. Khamis, “High Directivity Antenna Using a Modified
Peano Space-Filling Curve”, IEEE Antennas and Wireless Propagation Letters,
Vol.6, 2007
[15] J. Mc Vay and A. Hoorfar, “Miniaturization of top-loaded monopole antennas using
Peano-curves”, Radio and Wireless Symposium, IEEE, pp. 253-256, 2007
[16] T. Terada, K. Ide, K. Iwata, T. Fukusako, “Radiation characteristics of small and low
profile print antenna using peano line”, Asia Pacific Microwave Conference, pp.1-4,
2008

















Chapter 6

RFID Tag Antenna Design
Using Peano Curves
RFID Tag Antenna Design Using Peano Curves
31
Chapter 6 RFID Tag Antenna Design Using Peano Curves
6.1 Outline of the chapter
In the previous chapters, the relevant theory and the concepts regarding Peano line
antennas are discussed in detail. In this chapter, the focus is shifted towards a particular
design problem using microstrip Peano line antenna. As discussed earlier, Peano line
antennas in microstrip configuration find immense applications in RFID chips. Design of
such a chip antenna is taken into consideration and the necessary design steps are discussed
in detail. Moreover, the difficulties faced in different design approaches are stated and a
possible modification is highlighted. This chapter mainly emphasizes on the practical
workable antenna design and the difficulties that need to be addressed for future
implementation.

6.2 RFID frequency bands
In Chapter 2, the RFID bands are discussed briefly. The RFID band from 26.957 –
27.283 MHz is mainly used for special application such as biomedical applications. For
biomedical purposes, the distance required for signal transmission is very less. Therefore, the
gain of those antennas need not be very high. Moreover, human body cannot be irradiated
with very high power signals. Hence, a relatively lower antenna gain suffices our cause. The
next higher RFID band is the less frequently used 433 MHz band.
For this reason, a design problem relating to the design of microstrip Peano line antenna
for 26 MHz application is considered.

6.3 Design Goal
Our design goal is to design a microstrip Peano line antenna, corresponding to a 26 MHz
frequency of operation, to be used for bio-medical applications meeting all specifications.
The design needs to be incorporated on an RFID chip, typically not exceeding 20 mm X 20
mm area dimensions. In addition, design for 433 MHz RFID band is also considered.


RFID Tag Antenna Design Using Peano Curves
32
6.4 Design Procedure
The basic idea behind the design of such an antenna at very low frequency, yet keeping
its dimensions small, is to fill the space over the substrate with a patch in the most efficient
manner. This can be achieved if the patch or the microstrip line traces the shape of a space-
filling curve, in our case, the Peano curve. As we are constrained in size, the best alternative
is to incorporate a 3
rd
order Peano curve and to try out the design. But to realize the 3
rd
order
Peano curve in reality is a tough task, especially as the fabrication difficulty creeps in. Hence,
it is wiser to stick to the more easily realizable 2
nd
order Peano curve. But, as the surface
current traverses a path of less distance over a microstrip line following a 2
nd
order Peano
curve than that following a 3
rd
order Peano curve, some alternative design method must be
thought of.
The alternative method is to follow the repetitive pattern of a 2
nd
order Peano curve and
to go on extending it keeping the symmetry of the curve intact, as long as the dimensions of
the antenna are not exceeded. Thus, the work of Fukusako et al [1] can be taken as the basic
guideline or prototype in this regard.
The design in [1] has a Peano line antenna designed for a central frequency of 1 GHz.
Extending the length of the microstrip line by following a 2
nd
order Peano curves’s repetitive
pattern, we effectively increase the path over which the surface current traverses. Hence, it
results in an addition of inductive effect and also added capacitive effect due to mutual
coupling between adjacent line segments. In this way, we reduce the resonant frequency of
the antenna to a great extent.

6.5 Design prototype
The antenna design as shown by Fukusako et al. in [1] is shown in Fig. 6.1. The substrate
used is a polymide substrate with a relative permittivity (ε
r
) of 3.5. The height of the substrate
‘h’ is 50 μm as shown in Fig. 6.2. The width of the microstrip line ‘w’ is 0.1 mm. The RFID
chip area is 20 mm X 20 mm, while the antenna footprint area is 3.2 mm X 6.1 mm. The
simulation is performed using Method of Moment based IE3D from Zeland Software™ Inc.
[2].


RFID Tag Antenna Design Using Peano Curves
33











From Fig. 6.1 it is clear that the patch or the
microstrip line does not trace a normal 2
nd
order
Peano curve. Rather, a repetitive pattern is followed,
which is an extension to the 2
nd
order Peano curve,
truncated after certain distance. This section forms
the radiator. The above design of Fig. 6.1 can be
used as the basic guideline or prototype to our
design, to fill the space above the substrate in the
most efficient manner
The total path length ( l ) that the strip of the antenna traverses is 90.75 mm.
The resonance of the antenna as inferred from the simulated results in [1] occurs at 1.02
GHz. This corresponds to a free space wavelength (λ
0
) of 294.1 mm.
The effective permittivity (ε
eff
) can be found out from the expression [3]
w
h
r r
eff
12 1
1
2
1
2
1
+
×

+
+
=
ε ε
ε
(6.1)
Thus, for ε
r
= 3.5, we get ε
eff
= 2.722

(a) (b)
Fig. 6.1 (a) Antenna Structure [1]. (b) Detailed View



Fig. 6.2 Height of the antenna substrate
from the CPW feed end

RFID Tag Antenna Design Using Peano Curves
34
So the effective wavelength of the signal that is propagating through the antenna can be
obtained from the expression
eff
eff
ε
λ
λ
0
= (6.2)
Thus, the effective wavelength in this case is 178.2 mm.
If we desire to find out, the factor by which the effective wavelength is related to the antenna
path length ( l ), we can do so by finding the factor ‘x’ from the equation
x λ
eff
= l (6.3)
For this structure, x = 0.509, thus the antenna resonates almost for a length l = λ
eff
/2.

6.6 Design approaches
This section highlights several design approaches and relevant explanations, the design
being based on the guideline stated in the last section.
Study of the design guideline/prototype:
Before proceeding
further with the design,
it is essential to go
through the impedance
characteristics of the
design guideline on
which the design is
based. Fig. 6.3 shows
the S
11
and the
impedance versus
frequency curves for the
antenna described in Fig. 6.1 over a frequency range confined around the first resonant
frequency. On the other hand, Fig. 6.4 shows the same characteristics over 0-20 GHz
frequency range for the same antenna. The simulation is performed using IE3D from Zeland
Software™ Inc. [2].

(a) (b)
Fig. 6.3 (a) S
11
in dB vs. Frequency (b) Input impedance vs. frequency

RFID Tag Antenna Design Using Peano Curves
35









Fig. 6.4 (a) S
11
vs. frequency
(b) Input impedance vs. frequency over
0-20 GHz frequency range

Discussion
The above phenomena can be attributed to the distributed nature of the antenna. The
microstrip line, which forms the antenna, is basically a transmission line, and as with all
transmission lines, the distributed effects cause the impedance characteristics to repeat itself
with frequency.
: From the above curves, it is obvious that, as the frequency range is increased the
pattern repeats itself. The S
11
matching is not as good as desired, but the impedance
characteristic reveals that the input resistance and reactance repeats itself, thus exciting
higher order modes at higher frequencies.

6.6.1 Approach 1: Increase of path length of the antenna
As discussed earlier, the method to reduce the resonant frequency of the antenna is to go
on increasing the path length l that the microstrip line of the antenna traces.
In the subsequent approaches, the basic prototype of [1] is adhered to, and the symmetric
periodic repetition of the 2
nd
order Peano curve is followed. Maintaining symmetry of
repetition, we consider a certain yardstick such that, in each step of antenna length extension,
19.85 mm path length is traced. Also, the outer dimension along the path of the curve,
increases by 1.2 mm. This process is carried out till the dimensional constraint along the path
(a)

(b)
RFID Tag Antenna Design Using Peano Curves
36
length is exceeded. The substrate is kept the same, i.e. ε
r
= 3.5. Other dimensions such as w
and h are also kept similar to that of the antenna described in the design guideline
Feed: The antenna is fed by a 50-ohm coplanar waveguide, with dimensions as given in Fig.
6.1. The simulation is done using IE3D software, taking advanced extension as the de-
embedding scheme for port definition. The signal line of the CPW is provided with a positive
voltage level, and a negative polarity to the two ground planes on the side. The total antenna
structure along-with the feed is conductor-backed.
a) Extension 1: Incrementing a path length of 19.85 mm to the previous design, the antenna
is simulated. Therefore, the total path length l now becomes 110.6 mm. Hence, the new
antenna footprint dimension is 3.2 mm X 7.3 mm, since an increment of 1.2 mm occurs
along one dimension. The structure is shown in Fig. 6.5, while the S
11
characteristic is shown
in Fig. 6.6.
Extensions of path length












The first resonance of the antenna occurs at 837.807 MHz, with an S
11
match of -15.2036 dB.
The corresponding free space wavelength (λ
0
) for this frequency is 358.07 mm.
Thus, using (6.2) we can evaluate effective wavelength (λ
eff
) which comes out to be 217.03
mm.

Fig. 6.5 Antenna structure Fig. 6.6 S
11
characteristics for antenna in Fig. 6.5
For extension 1


RFID Tag Antenna Design Using Peano Curves
37
Finally, using (6.3) we can find out the factor x as 0.5096.
As the first resonance of the antenna occurs at 837.807 MHz, therefore in order to achieve
our design goal, we require many such path length extensions.
The above process is repeated in all the next extensions, keeping the basic parameters of the
antenna unaltered.

b) Extension 2:
Antenna footprint area: 3.2 mm X 8.5
mm
Path length l = 110.6 mm + 19.85 mm =
130.45 mm.
First resonant frequency, f
r
= 718.985
MHz, with S
11
= -9.18319 dB
Free space wavelength, λ
0
= 417.25 mm.
Effective wavelength, λ
eff
= 252.9 mm.
Factor x = 0.5158.
As evident from the data above, the first
resonant frequency does not provide a
substantial S
11
match, i.e. it does not adhere
to the 2:1 VSWR acceptable limits.
The second resonance occurs at 1.45303
GHz, with S
11
= -15.1831 dB, thus
providing good matching characteristics. In
this case, we can therefore use the higher
order mode for operation, but not the
fundamental mode.
Fig. 6.7 depicts the S
11
characteristics and
the impedance characteristics of the antenna.


(a)

(b)
Fig. 6.7 (a) S
11
vs. Frequency
(b) Input impedance vs. Frequency
for antenna in extension 2

RFID Tag Antenna Design Using Peano Curves
38
c) Extension 3:
Antenna footprint area: 3.2 mm X 9.7 mm
Path length l = 130.45 mm + 19.85 mm = 150.3 mm.
First resonant frequency, f
r
= 632.242 MHz, with S
11
= -5.85262 dB.
Free space wavelength, λ
0
= 474.55 mm.
Effective wavelength, λ
eff
= 287.6 mm.
Factor x = 0.5226.
Again, we have a poor match at the first resonant frequency. The second resonance occurs at
1.27 GHz, with S
11
of -33.5 dB. This higher order mode possesses good VSWR
characteristics, and hence can be excited for operation.
Fig. 6.8 shows the S
11
characteristics and the impedance characteristics of the antenna.













Discussion: Fig. 6.8 (b) highlights a very important aspect of the design. It shows that the
first resonance actually occurs around 1.27 GHz. and not at 632.242 MHz as mentioned in the
above data.

(a)

(b)

Fig. 6.8 (a) S
11
vs. Frequency
(b) Input impedance vs. Frequency for antenna in extension 3


RFID Tag Antenna Design Using Peano Curves
39
This anomaly can be clarified by the following logic. The periodicity and the symmetry
of the impedance characteristic pattern suggests that, even though the first resonant
frequency, or the first zero reactance crossing frequency is 1.27 GHz, yet the original first
resonance of the antenna should have occurred at a much lower frequency, i.e. at 632.242
MHz. It is only due to some stray capacitance, which adds some negative reactance on the
curve near its original first resonance, and causes the curve to shift away from the zero
reactance crossing line. The cause of this stray capacitive effect shall be investigated later.
If the stray capacitances play a predominant role, the impedance pattern cannot provide
any information about the first resonance, since the reactance curve does not cross the zero
reactance line near it. Therefore, from now onwards, as we increase the path length of the
antenna in the subsequent extensions, we shall only focus on the first dip of the S
11

characteristics to find out the first resonant frequency.

d) Extension 4:
Antenna footprint area: 3.2 mm X 10.9 mm
Path length l = 150.3 mm + 19.85 mm =
170.15 mm.
First resonant frequency, f
r
= 563.6 MHz
S
11
at first resonance = -4.1565 dB
Free space wavelength, λ
0
= 532.29 mm.
Effective wavelength, λ
eff
= 322.631 mm.
Factor x = 0.527

e) Extension 5:
Antenna footprint area: 3.2 mm X 12.1 mm
Path length l = 170.15 mm + 19.85 mm = 190
mm.
First resonant frequency, f
r
= 508.32 MHz

Fig. 6.9 S
11
vs. frequency for antenna of ext.4

Fig. 6.10 S
11
vs. frequency for antenna of ext.5


RFID Tag Antenna Design Using Peano Curves
40
S
11
at first resonance = -3.07537 dB
Free space wavelength, λ
0
= 590.179 mm.
Effective wavelength, λ
eff
= 357.719 mm.
Factor x = 0.531.

f) Extension 6:
Antenna footprint area: 3.2 mm X 13.3 mm
Path length l = 190 mm + 19.85 mm = 209.85
mm.
First resonant frequency, f
r
= 462.848 MHz
S
11
at first resonance = -2.34511 dB
Free space wavelength, λ
0
= 648.1 mm.
Effective wavelength, λ
eff
= 392.8 mm.
Factor x = 0.534.

g) Extension 7:
Antenna footprint area: 3.2 mm X 14.5 mm
Path length l = 209.85 mm + 19.85 mm =
229.7 mm.
First resonant frequency, f
r
= 425.041 MHz
S
11
at first resonance = -1.8317 dB
Free space wavelength, λ
0
= 705.8 mm.
Effective wavelength, λ
eff
= 427.79 mm.
Factor x = 0.5369.




Fig. 6.11 S
11
vs. frequency for antenna of ext.6


Fig. 6.12 S
11
vs. frequency for antenna of ext.7
RFID Tag Antenna Design Using Peano Curves
41
h) Extension 8:
Antenna footprint area: 3.2 mm X 15.7 mm
Path length l = 229.7 mm + 19.85 mm = 249.55 mm.
First resonant frequency, f
r
= 392.962 MHz
S
11
at first resonance = -1.45881 dB
Free space wavelength, λ
0
= 763.43 mm.
Effective wavelength, λ
eff
= 462.72 mm.
Factor x = 0.539.

i) Extension 9:
Antenna footprint area: 3.2 mm X 16.9 mm
Path length l = 249.55 mm + 19.85 mm = 269.4
mm.
First resonant frequency, f
r
= 365.303MHz
S
11
at first resonance = -1.18201 dB
Free space wavelength, λ
0
= 821.2 mm.
Effective wavelength, λ
eff
= 497.7 mm.
Factor x = 0.541.

j) Extension 10:
Antenna footprint area: 3.2 mm X 18.1 mm
Path length l = 269.4 mm + 19.85 mm = 289.25 mm.
First resonant frequency, f
r
= 341.408 MHz
S
11
at first resonance = -0.971126 dB
Free space wavelength, λ
0
= 878.7 mm.
Effective wavelength, λ
eff
= 532.5 mm.

Fig. 6.14 S
11
vs. frequency for antenna of ext.9

Fig. 6.15 S
11
vs. frequency for antenna of
ext. 10

Fig. 6.13 S
11
vs. frequency for antenna of ext.8

RFID Tag Antenna Design Using Peano Curves
42
Factor x = 0.543.

k) Extension 11:
Antenna footprint area: 3.2 mm X 19.3 mm
Path length l = 289.25 mm + 19.85 mm = 309.1
mm.
First resonant frequency, f
r
= 320.131 MHz
S
11
at first resonance = -0.808384 dB
Free space wavelength, λ
0
= 937.1 mm.
Effective wavelength, λ
eff
= 567.99 mm.
Factor x = 0.544.

1. The first resonant frequency is still deviant from our design goal. Further reduction of the
resonant frequency is required. However, we cannot go on increasing the antenna path length
for further extensions, as the size limit of the RFID chip is fixed and prescribed.
Inferences from the extension of antenna path lengths
2. The S
11
input impedance matching characteristic for the antenna at the first resonant
frequency is very poor. Thus, the issue of impedance matching must be taken care of.

6.6.2 Approach 2: Higher permittivity substrate
The earlier approach reveals that the chip size is too small to facilitate a frequency
reduction to a frequency stated in our design goal. Hence, another alternative needs to be
employed. In this approach, we simulate the antenna, using alumina as the dielectric
substrate, instead of polymide. Alumina has a dielectric constant of 9.8.
We know from (6.2),
eff
eff
ε
λ
λ
0
=




Fig. 6.16 S
11
vs. frequency for antenna of
ext. 11

RFID Tag Antenna Design Using Peano Curves
43
Therefore, as we increase ε
r,
there is an increase in ε
eff.
So to keep the effective wavelength
λ
eff
as constant, an increase in ε
eff
must be accompanied by an increase in free space
wavelength λ
0
. Thus, using a higher dielectric medium aids in frequency reduction.
Disadvantage
In this approach, the simulation results are performed by considering alumina as the
substrate, and then following the length extension procedure of approach 1. For the sake of
clarity, the extensions are represented in a tabular form, and the S
11
curves are not shown.
Table 1 illustrates the relevant results obtained in this regard. For, ε
r
= 9.8, we get from (6.1),
ε
eff
= 7.063
: The disadvantage in this method is, as we increase the dielectric constant, the
gain of the antenna is bound to reduce. It is because; using higher permittivity substrate
confines most of the electric field lines beneath the antenna strip, leaving only a few field
lines, as ‘loosely bound’ for effective radiation.
TABLE 6.1 LENGTH EXTENSIONS FOLLOWING APPROACH 2
Extension
number
Antenna
footprint
(mm
2
)
Path
length l
(mm)
First
resonan
t freq.
(MHz.)
S
11
at first
resonance
(dB)
Free space
wavelength

0
) in mm
Effective
wavelengt
h (λ
eff
) in
mm
Factor
x
0 (design
guideline)
3.2 X 6.1 90.75 614.894 -40.9954 487.8 183.5 0.494
1 3.2 X 7.3 110.6 513.800 -11.8655 583.8 219.6 0.503
2 3.2 X 8.5 130.45 440.098 -7.44185 681.66 256.49 0.508
3 3.2 X 9.7 150.3 385.106 -5.13764 779.0 293.1 0.512
4 3.2X10.9 170.15 342.553 -3.73499 875.7 329.5 0.516
5 3.2X12.1 190.0 308.6 -2.81363 972.0 365.7 0.519
6 3.2X13.3 209.85 280.8 -2.1779 1068.0 401.0 0.523
7 3.2X14.5 229.7 257.6 -1.72367 1164.5 438.1 0.524
8 3.2X15.7 249.55 238.0 -1.389 1260.5 474.2 0.526
9 3.2X16.9 269.4 219.722 -1.13 1365.36 513.7 0.524
RFID Tag Antenna Design Using Peano Curves
44
Extension
number
Antenna
footprint
(mm
2
)
Path
length l
(mm)
First
resonan
t freq.
(MHz.)
S
11
at first
resonance
(dB)
Free space
wavelength

0
) in mm
Effective
wavelengt
h (λ
eff
) in
mm
Factor
x
10 3.2X18.1 289.25 207.938 -0.94 1442.7 542.8 0.532
11 3.2X19.3 309.1 193.863 -0.79 1547.4 582.2 0.531

Inferences from Approaches 1 & 2
1. From the above approaches of length extension, it follows that it is quite difficult to
achieve a resonance at such a low frequency, keeping the dimensions within the prescribed
limits.
:
2. Moreover, in most of the cases, the S
11
matching characteristics at the first resonant
frequency is poor.

6.6.3 Approach 3: Input impedance match using quarter wave transformer
Having discussed the concept of length extension, we now turn our attention towards the
next approach, which mainly deals with the concept of impedance matching. As revealed in
the earlier sections, the design needs to achieve the required 2:1 VSWR limit. For this
purpose, we can start with a quarter-wave transformer impedance matching scheme.
The basic idea in this approach is highlighted below:
1. First and foremost, find out the impedance at the end of the antenna. That impedance is
treated as the load impedance.
2. Design a quarter-wave transformer of length λ
eff
/4, where λ
eff
is the effective wavelength,
for the design frequency of interest.
3. The quarter-wave section will have a characteristic impedance which is the geometric
mean of the magnitude of the load impedance (in our case, it is the edge or end impedance of
the antenna) and the characteristic impedance of the section that is connected at its input. As
the quarter-wave transformer is connected to a coaxial cable at its input for initial feed, the
characteristic impedance of that section is taken as 50 ohms. Using these data, the
characteristic impedance of the transformer is calculated.
RFID Tag Antenna Design Using Peano Curves
45
4. The quarter-wave section can now be implemented using, either a microstrip line or a
coplanar waveguide (CPW).
Difficulties and bottlenecks
For our design frequency of interest, i.e. 26 MHz we can calculate the length of the
transformer for both the substrates used above.
: The main difficulty in this design is the length of the quarter-
wave transformer, which exceeds our dimensional constraints.
For 26 MHz, λ
0
= 11.538 m.
a) For polymide substrate:
ε
r
= 3.5
Therefore, using (6.1), we get ε
eff
= 2.722.
Effective wavelength, λ
eff
using (6.2) amounts to

6.993 m
Thus, the length of the quarter-wave transformer, λ
eff
/4 = 1.748 m = 1748 mm.
b) Similarly, for Alumina substrate:
ε
r
= 9.8
ε
eff
= 7.063.
Effective wavelength, λ
eff
= 4.341 m
Hence, the length of the quarter-wave transformer, λ
eff
/4 = 1.085 m = 1085 mm.

Design goal revisited
However, the difficulty persists. A frequency of 433 MHz corresponds to a free space
wavelength of 692.8 mm. The effective wavelength considering polymide substrate is 419.9
mm, while that considering alumina substrate is 260.68 mm. Therefore, the implementation
of the quarter-wave transformer on polymide substrate requires a length of 104.97 mm, while
the length of the same for alumina substrate is 65.17 mm, both of which exceed our
dimensional constraint. This modification therefore cannot overcome the dimensional
difficulties as faced earlier.
: Due to several difficulties faced, as documented above, we alter our
design goal to the next higher RFID frequency of 433 MHz. Extension number 2 from Table
6.1 specifies that, there is a first resonance at 440.098 MHz that is quite close to our revisited
design goal. For this reason, we try our design around this frequency.
RFID Tag Antenna Design Using Peano Curves
46
Inference from Approach 3: The length of the quarter-wave transformer itself (without
antenna) exceeds our prescribed dimensional constraints for an RFID chip by a good margin.

6.6.4 Approach 4: Use of different dielectric materials for feed and antenna
The problem stated above can be addressed by taking two different dielectric medium
side by side, and gluing them together, so that the quarter-wave transformer feed is placed on
the material with a higher dielectric constant, and the antenna on the lower permittivity
substrate. The antenna being placed on the low permittivity substrate ensures that its gain is
not reduced, while using higher permittivity substrate for the transformer ascertains a
miniaturized length for it. An increase in the dielectric constant brings about a decrease in the
effective wavelength at the design frequency of interest, thereby reducing the length of the
quarter-wave transformer required.
Drawback of the design
In this approach, an alternative design to address the above problem is proposed. The
schematic side view of the design is
presented in Fig. 6.17.
: The primary drawback of this design is the discontinuity arising
from the dielectric-dielectric interface. Due to this discontinuity, however small it may be, a
gap arises in the structure at the interface, thereby adding an extra amount of stray capacitive
reactance, thus deviating the antenna further away from resonance. The remedy for this
problem is to design the structure on two different dielectric media as stated above, but
avoiding the gap due to discontinuity as far as practicable. This method is illustrated in the
next approach.

6.6.5 Approach 5: Use of stacked dielectric layers
The structure is made up of two-
stacked dielectric layer with ε
r2
> ε
r1
.
On the higher permittivity substrate 2,
the quarter-wave feed is placed, while
the antenna is placed on the lower
permittivity substrate 1. The antenna
Fig. 6.17 Schematic side view of the design


RFID Tag Antenna Design Using Peano Curves
47
input end, i.e. the edge of the antenna, forms the load to the transformer. The load is
connected to the matching section by a shorting post or via.
The via or the shorting post is very small in length, as compared to the effective wavelength,
since we are operating at a very low design frequency. Therefore, it can safely be considered
as a lumped connector, due to absence of any distributed effect.
Difficulty
The modified scheme as depicted is almost similar to the scheme discussed in the previous
attempt. We now shift our design goal to the next higher RFID frequency of 433 MHz.
Considering this frequency as the design
frequency, we have a closer look at this
design approach.
: This design can be tried out, but it calls for a bit of modification. The quarter
wave transformer is actually immersed in between two different dielectric media, none of
which is air. As the quarter-wave transformer is implemented using a microstrip line,
standard close form relations are available assuming one of the dielectric medium as air. So
to avoid any loss of generality, we slightly modify the structure in the next approach.

6.6.6 Approach 6: Modification of the structure described in approach 5
For modification, we just invert the layer denoted by ‘substrate 2’ along with the quarter-
wave transformer on it. Therefore, the ground plane comes in between the two media, and
thus separates them. Both the transformer and the antenna, have a dielectric medium on one
side with a common ground plane, and air on the other side. Also, the shorting post is an
elongated one, drilling a hole on the ground plane for transformer to antenna connectivity.
Fig. 6.18 depicts the modified scheme.
As we observed from extension number
2 of Table 6.1, there is a resonance at
440.098 MHz, which happens to be the first
resonance of the antenna. Henceforth, our
design is mainly centered at this frequency.
We take that structure for further
analysis, in order to investigate the
suitability of quarter wave matching

Fig. 6.18 Modified schematic side view of the design

RFID Tag Antenna Design Using Peano Curves
48
technique on it. The first step is to remove the CPW feed from the antenna, and keep the
antenna isolated. Then, the edge impedance of the antenna is measured using “Extension for
waves” de-embedding scheme in IE3D simulator. This de-embedding scheme serves as a
wave-port for the antenna. Both the cases are simulated, and the S
11
and the impedance
curves are observed for comparative study, which reveals some important facts. Fig. 6.19
shows the S parameters and the input impedance characteristics of the antenna described in
extension 2 of Table 6.1. Fig. 6.20 shows the same antenna structure and its impedance
characteristics, with the CPW feed removed, in order to measure its edge or end impedance.














As seen from the figure above, the first dip of the S
11
curve occurs near 440.098 MHz, when
the CPW feed is connected (antenna of ext. 2, table 1). The impedance curves on the other
hand, cannot conclusively ascertain it as the first resonant frequency, since the reactance
curve does not cross the zero-reactance line. Therefore, considering the first dip of the S
11

curve, we assume 440.098 GHz as the first resonant frequency.


Fig. 6.19 (a) S
11
vs. frequency (b) Input impedance vs. frequency for antenna of ext.2, table1

(a)

(b)
RFID Tag Antenna Design Using Peano Curves
49











Fig. 6.20 (a) Antenna described in ext.2 of Table 1, with CPW feed removed
(b) Input impedance characteristics for the same antenna

The same antenna, with CPW feed removed, looks like that in Fig. 6.20 (a). Here, we have
applied a wave-port for calculating its input edge impedance. The characteristic curve is
shown in Fig. 6.20 (b).
Discussion

: If we have a closer look at the curves in Fig. 6.19 (b) and Fig. 6.20 (b), we arrive
at a very important conclusion. When the CPW is used for feeding the antenna, a stray
capacitance as discussed earlier, shifts the input reactance curve away from resonance near its
first resonant frequency. However, when the CPW feed is removed, the reactance curve near
the first resonant frequency rises, and resonance occurs, since the stray capacitive effect is
removed. We can thus infer from the observation that, the CPW structure adds the stray
capacitance.
Analysis and results
From Fig. 6.20 (b), we find the impedance at the input edge of the antenna at 440.098 MHz.
The input resistance R
in
=96.3777 ohms.
: In this section, we verify the suitability of this design for
implementing quarter wave matching technique on it.

(a)


(b)
RFID Tag Antenna Design Using Peano Curves
50
The input reactance X
in
= 97.5295 ohms.
Therefore, the magnitude of input impedance is
|Z
in
| = { (R
in
)
2
+ (X
in
)
2
}
1/2
= 137.11 ohms at 440.098 MHz.
This |Z
in
| acts as the load to the quarter-wave matching section. Let us denote it by Z
L
.
The input to the matching section is a standard 50 ohms coaxial cable, with |Z
0
| = 50 ohms.
Let the characteristic impedance of the matching section be Z
1.

Therefore, Z
1
can be calculated from the relation [4]
|Z
1
| = { |Z
0
| |Z
in
| }
1/2
(6.4)
Putting the values of |Z
in
| and |Z
0
| in (6.4), we get |Z
1
| = 82.79 ohms.
Therefore, the quarter-wave matching section has a characteristic impedance of 82.79 ohms.
We can implement it using a microstrip line.
Standard closed form synthesis formulae for microstrip design are available in the literature
[5]. They are stated below:

1
' exp 4
1
8
' exp

|
|
.
|

\
|
− =
H
H
h
w
For narrow strips ( i.e. when Z
0
> {44 – 2 ε
r
} ohms )
(6.5)
where,
|
|
.
|

\
|
+
|
|
.
|

\
|
+

+
+
=
π ε
π
ε
ε
ε
4
ln
1
2
ln
1
1
2
1
9 . 119
) 1 ( 2
'
0
r r
r
r
Z
H

( ) ( ) { } ( )
)
`
¹
¹
´
¦
− + −

+ − − − =
r
e
r
r
e e
d d d
h
w
ε πε
ε
π
517 . 0
293 . 0 1 ln
1
1 2 ln 1
2
For wide strips ( i.e. when Z
0
< {44 – 2 ε
r
} ohms )
(6.6)
where,
r
Z
d
ε
π
0
2
95 . 59
=
Case 1: Taking the substrate as polymide ( ε
r
= 3.5 ) and Z
0
= 82.79 ohms
Using (6.5), the value of microstrip width w = 0.044 mm.
Case 2: Taking the substrate as alumina ( ε
r
= 9.8 ) and Z
0
= 82.79 ohms
RFID Tag Antenna Design Using Peano Curves
51
Using the same formula, the width of the microstrip w = 0.013 mm.
Therefore, the implementation of a quarter-wave transformer for this design using a
microstrip line requires the microstrip width to be of the dimensions as stated above. The
antenna itself is placed on an alumina substrate, and as discussed in previous sections, it
cannot provide a miniaturized length for the quarter-wave section. So, the quarter-wave
transformer must be placed on a still higher permittivity substrate. However, as evident from
the formulae (6.5) and (6.6) above, and from the two case studies, the more is the dielectric
constant of the medium, the smaller is the microstrip width. For alumina, it is as small as
0.013 mm. For other higher permittivity substrates, it is even smaller. Hence, fabrication
difficulty comes in, and micro-fabrication techniques must be employed.
Difficulties from the design:
1. As discussed above, if the dielectric constant of the substrate is kept very low, the width of
the microstrip line required may be within normal fabrication limits, and the quarter-wave
section can be implemented. Nevertheless, the length of the quarter-wave section at our
design frequency of interest becomes very large.
2. Conversely, if the permittivity of the substrate used is high, it is advantageous as far as the
length of the quarter-wave section is concerned. However, the width of the microstrip line
becomes very small, and fabrication problems appear.

6.6.7 Approach 7: Single stub matching
We arrive at the last of our design approaches, i.e. impedance match using single shunt
stub. In microstrip configuration, it is difficult to realize a series stub. The basic concept
behind single shunt stub matching is very briefly presented in the next section.
Single-stub tuning: This technique uses a single open-circuited or short-circuited length of
transmission line (stub), connected in parallel with the feed line at a certain distance from the
load. The circuit is shown in Fig. 6.21.


RFID Tag Antenna Design Using Peano Curves
52
The task is to find out the distance of
the stub ‘d’ from the load end, and
the length of the open or shorted
stub, ‘l’. By selecting proper values
of ‘d’ and ‘l’, a perfect match
between the feed-line of
characteristic impedance Z
0
and the
load impedance Z
L
is accomplished.


The stub matching technique
can be implemented on the
same structure shown in Fig.
6.18. The quarter-wave
matching section in the figure
is replaced by a single shunt
stub section. The rest of the
structure remains the same.
The transmission feed line as
well as the stub is
implemented using microstrip
lines. As in earlier case, the
input edge of the antenna acts
as the load. The characteristic impedance of the microstrip transmission feed line is taken as
50 ohms. Fig. 6.22 shows the same schematic structure of Fig. 6.18 in bottom view, with
quarter-wave matching transformer replaced by a stub matching section. The antenna which
serves as the load, and connected to the matching section by a shorting post or via, is not
shown in the figure, as it is placed below the bottom layer.
Analysis [4]
Let the load be denoted as Z
L
= R
L
+ j X
L
.
:
The impedance Z, down a length d of the line from the load as shown in Fig. 6.21 is given by

Fig. 6.21 Single shunt stub tuning circuit

Fig. 6.22 Implementation of Shunt stub in structure
of Fig. 6.18 (bottom view)


Y = 1/Z
RFID Tag Antenna Design Using Peano Curves
53
t jX R j Z
t jZ jX R
Z Z
L L
L L
) (
) (
0
0
0
+ +
+ +
=
, where t = tan βd. (6.7)
The admittance at this point is
Y = G + j B = 1/Z,
where
2
0
2
2
) (
) 1 (
t Z X R
t R
G
L L
L
+ +
+
= (6.8a)

] ) ( [
) )( (
2
0
2
0
0 0
2
t Z X R Z
t Z X t X Z R
B
L L
L L L
+ +
+ − −
= (6.8b)
Now, d (which implies t) is chosen so that G = Y
0
= 1/Z
0
. From (6.8a), this results in a
quadratic equation for t. Solving for t gives
0
0
2 2
0
/ ] ) [(
Z R
Z X R Z R X
t
L
L L L L

+ − ±
= , for R
L
≠ Z
0
(6.9)
Thus, the two principal solutions for d are
¦
¦
)
¦
¦
`
¹
¦
¦
¹
¦
¦
´
¦
+
=


) tan (
2
1
tan
2
1
1
1
t
t
d
π
π
π
λ
(6.10)
The length of the stub can be found from the formulae below:
For an open-circuit stub,
|
|
.
|

\
| −
=

0
1 0
tan
2
1
Y
B l
π λ
(6.11a)
For a short-circuited stub,
|
.
|

\
|
=

B
Y l
s 0 1
tan
2
1
π λ
(6.11b)
If the length given by (6.11a) or (6.11b) is negative, λ/2 can be added to give a positive result.
Results
For our design as we found earlier, the input end of the antenna has a resistance of 96.3777
ohms and a reactance of 97.5295 ohms.
:
for t ≥ 0
for t < 0




RFID Tag Antenna Design Using Peano Curves
54
For ease of calculation, we take R
in
≈ 96 ohms, and X
in
≈ 97 ohms.
This impedance serves as the load impedance Z
L
for the matching section.
Therefore, Z
L
= R
L
+ j X
L
≈ 96 + j 97, where R
L
≈ 96 ohms and X
L
≈ 97 ohms.
We consider a 50-ohm microstrip line for transmission feed, hence, Z
0
= 50 ohms.
Putting these data in (6.9), we evaluate the value of t.
We get, t = 5.34, -1.125 as two solutions for t.
Now we have two cases, one when t = 5.34, and another when t = -1.125.
Case 1: t = 5.34
From (6.10), we get d = 0.22 λ.
Using (6.8b), B = 0.03098.
Putting B in (6.11a), we get
l
o
= -0.1587 λ
This length being negative, we add a length of λ/2 to it.
Therefore, l
o
= 0.3413 λ
Putting B in (6.11b), we get l
s
= 0.091 λ
Case 2: t = -1.125
Following the similar steps as above, we obtain,
d = 0.365 λ.
l
o
= 0.1587 λ
l
s
= -0.091 λ => -0.091 λ + 0.5 λ = 0.409 λ
Optimum solution:
From the two cases discussed, the optimum solution for minimum dimension is for t = 5.34.
The distance between the load and the stub, d = 0.22 λ
The length of the stub is least when it is a short-circuited one.
The length of the stub, l
s
= 0.091 λ

RFID Tag Antenna Design Using Peano Curves
55
Dimensions:
The wavelength λ, with respect to which all the above dimensions are provided, is basically
the effective wavelength λ
eff
for this design. We now evaluate the dimensions in two cases,
one when the substrate is polymide, and again when the substrate is alumina. The
transmission feed line in both the cases is 50-ohm microstrip line.
Z
0
= 50 ohms for microstrip feed line.
a) Polymide substrate (ε
r
= 3.5)
To implement this microstrip line, let the width required for the same is ‘w’.
Using (6.5), w = 0.1 mm.
Therefore, the width of the microstrip line of 50-ohms characteristic impedance must be 0.1
mm.
Now, for 440.098 MHz, effective wavelength λ
eff
= 413.16 mm.
The distance d of the stub from the load end is: d = 0.22 λ
eff
= 90.89 mm
Length of the short-circuited stub l
s
is: l
s
= 0.091 λ
eff
= 37.59
Similarly, for alumina substrate,
b) Alumina substrate (ε
r
= 9.8)
λ
eff
= 256.49 mm, for 440.098 MHz.
Microstrip width w = 0.04 mm
Separation distance d = 56.42 mm
Short-circuited stub length l
s
= 23.34 mm
Difficulties from the design
The results in the preceding section clearly suggest that the dimensions for d and l exceed our
overall RFID chip dimensions. In order to accommodate the stub within the prescribed limits,
we need yet higher permittivity substrates. However, it is not so convenient solution because
higher dielectric constant further reduces the width of the microstrip line, which gives rise to
many fabrication difficulties.
:


RFID Tag Antenna Design Using Peano Curves
56
6.7 Discussions
Having discussed all the design approaches, some important points worth mentioning are
discussed in this section. They are highlighted below:
1. If we look at the factor x, as calculated from (6.3) in the above length extension design
approaches, we can infer that, as the length l of the antenna increases, the value of x goes on
changing from the value of 0.5 as was initially in case of the design prototype. Hence we can
conclude, for Peano line antennas, the resonance does not occur for an antenna length l =
λ
eff
/2. The reason is explained in chapter 4, which is verified by the results provided in this
chapter.
2. The CPW feed as mentioned in some of the design approaches is a 50-ohm CPW feed
line. The length of the CPW- signal line is kept very small as compared to the effective
wavelength λ
eff
. This condition is satisfied throughout our design attempts, as our design goal
is to design the antenna for a very low RFID frequency. The antenna end, which serves as the
load to the CPW feed line, reflects a part of the signal fed. This results in standing wave
pattern in the CPW signal line. However, the loss due to those reflections can be neglected
for all practical purposes, as the signal line is very small with respect to the effective
wavelength.
Note: The dimensions for spacing between the signal line and the ground planes for realizing
a 50-ohm CPW feed in some of the above approaches, is found using ‘Line Gauge’ from
Zeland Program Manager [2]. The dimensions can also be calculated using standard CPW
formulae.

6.8 Summary
In this chapter, the approaches for an antenna design for RFID bands have been
discussed in detail. The practical constraints and limitations for each of the design schemes,
has been highlighted. Moreover, some important inferences regarding microstrip Peano line
antennas were illustrated.

References
RFID Tag Antenna Design Using Peano Curves
57
[1] T. Fukusako, T. Terada, K. Iwata, “Design and comparative study on planar small
antennas using meander and peano line structure”, International Symposium on
Antennas and Propagation Society, pp. 2451-2454, 2007
[2] Zeland Software Inc., USA, 2004, Release 10.12
[3] C.A. Balanis, Antenna Theory: Analysis and Design, 3rd Edition, John Wiley and
Sons Inc., 2005
[4] D.M. Pozar, Microwave Engineering, Third Edition, John Wiley & Sons, Reprint-
2008
[5] T. C. Edwards, Foundations for Microstrip Circuit Design, 2nd Edition, John Wiley
And Sons, 1991












Chapter 7

Probe Feed Optimization
Probe Feed Optimization
58
Chapter 7 Probe Feed Optimization
7.1 Outline of the chapter
The present chapter describes an investigative study on microstrip Peano line antenna, by
varying the probe-feed location. The changes in the impedance characteristics are highlighted
as the probe is fed by moving it inwards along the antenna length, till proper modes are
excited. For simplicity, only a few representations of the feed mechanism are depicted, while
the important observations are documented and discussed.

7.2 Antenna parameters
The patch of the antenna traces a standard second order Peano curve. The strip width is
chosen to be 1.5mm. The substrate dielectric constant is 4.3 of height 1 mm. The external
dimension of the antenna footprint is 25.5 mm X 25.5 mm. The simulation is performed using
IE3D software, considering infinite ground plane. The diameter of the inner conductor of the
probe is 1.3 mm. The probe is first fed at one end of the antenna, with inner conductor
feeding the Peano curve patch, and outer conductor completing the circuit at the ground
plane. The results are observed, and the probe position is shifted further along the length of
the antenna. This process is continued till proper modes are excited.

7.3 Variation in probe feed position
In this section, the results obtained from varying probe-feed position are highlighted. The
probe is first fed at the end of the antenna.
• Probe fed at end: The antenna structure is shown in Fig. 7.1a. The probe is located at
one end of the antenna as shown. The impedance characteristics are shown in Fig. 7.1b.
In Fig. 7.1b, the blue curve indicates the input resistance (R
in
) versus frequency plot,
while red indicates input reactance (X
in
) versus frequency plot. The frequency range from
0 to 1 GHz is studied. There are four frequencies, which are marked. From left to right,
the first resonant frequency is denoted by f
r1
, first anti-resonant frequency by f
ar1
, second
resonant frequency by f
r2
, while second anti-resonance by f
ar2
. The changes in these
frequencies with changing feed position are marked. The anti-resonant points are
identified by those frequencies, where there is a sharp transition from high inductive to a
high capacitive reactance. The anti-resonant nature is verified by the high input resistance
Probe Feed Optimization
59
at those frequencies, as observed from Fig. 7.1b. Let us define a few additional
parameters. Let the input impedances at f
r1
, f
r2
, f
ar1
and f
ar2
be denoted by R
in
(f
r1
), R
in
(f
r2
),
R
in
(f
ar1
) and R
in
(f
ar2
) respectively. Further, let us also define the steep inductive and
capacitive reactance just at both sides of the anti-resonant frequencies. Let the high
inductive reactance at the left vicinity of f
ar1
be denoted by X
in
(f
ar1
-) and the high
capacitive reactance at the right vicinity of f
ar1
be denoted by X
in
(f
ar1
+). Similarly, for f
ar2
,
the parameters are defined by X
in
(f
ar2
-) and X
in
(f
ar2
+). The VSWR characteristics are not
shown for this case, as the prescribed 2:1 VSWR limits are not achieved over the stated
frequency range.








Fig. 7.1 (a) Antenna structure when probe fed at edge (b) Impedance characteristics
The values for all the parameters defined above are stated below:
f
r1
= 210 MHz, f
r2
= 612 MHz, f
ar1
= 405 MHz, f
ar2
= 788.5 MHz,
R
in
(f
r1
) = 0.809 Ω, R
in
(f
r2
) = 0.882 Ω, R
in
(f
ar1
) = 3423 Ω, R
in
(f
ar2
) = 3047 Ω,
X
in
(f
ar1
-) = 1654 Ω, X
in
(f
ar1
+) = -1711 Ω, X
in
(f
ar2
-) = 1596 Ω, X
in
(f
ar2
+) = -1701 Ω.
The study is carried out by shifting the probe by 1.5 mm inwards along the antenna
length, in each step. This process continues till the optimized probe-feed location is
obtained for exciting the modes. The impedance pattern in all the feed positions remain
the same as shown in Fig. 7.1b, however, the parameter values changes. Therefore, those
values for changing feed positions are mentioned. Also, all the cases are not depicted for
simplicity. Very few representations are shown, though the important observations are
discussed.

(a)

(b)
Probe Feed Optimization
60
• Probe fed at 1.5 mm inwards along the antenna from the end:
f
r1
= 212 MHz, f
r2
= 616 MHz, f
ar1
= 405 MHz, f
ar2
= 791 MHz,
R
in
(f
r1
) = 0.8 Ω, R
in
(f
r2
) = 0.869 Ω, R
in
(f
ar1
) = 2944 Ω, R
in
(f
ar2
) = 3479 Ω,
X
in
(f
ar1
-) = 1749 Ω, X
in
(f
ar1
+) = -1719 Ω, X
in
(f
ar2
-) = 1593 Ω, X
in
(f
ar2
+) = -1670 Ω.
• Probe fed at 3.0 mm inwards along the antenna from the end:
f
r1
= 212 MHz, f
r2
= 620 MHz, f
ar1
= 405 MHz, f
ar2
= 791 MHz,
R
in
(f
r1
) = 0.79 Ω, R
in
(f
r2
) = 0.868 Ω, R
in
(f
ar1
) = 2921 Ω, R
in
(f
ar2
) = 3365Ω,
X
in
(f
ar1
-) = 1740 Ω, X
in
(f
ar1
+) = -1716 Ω, X
in
(f
ar2
-) = 1552 Ω, X
in
(f
ar2
+) = -1626 Ω.
• Probe fed at 6.0 mm inwards along the antenna from the end:
f
r1
= 214 MHz, f
r2
= 626 MHz, f
ar1
= 405 MHz, f
ar2
= 791 MHz,
R
in
(f
r1
) = 0.77 Ω, R
in
(f
r2
) = 0.85 Ω, R
in
(f
ar1
) = 2950 Ω, R
in
(f
ar2
) = 3149 Ω,
X
in
(f
ar1
-) = 1726 Ω, X
in
(f
ar1
+) = -1689 Ω, X
in
(f
ar2
-) = 1544 Ω, X
in
(f
ar2
+) = -1538 Ω.
• Probe fed at 15.0 mm inwards along the antenna from the end:
f
r1
= 222 MHz, f
r2
= 646 MHz, f
ar1
= 405 MHz, f
ar2
= 793 MHz,
R
in
(f
r1
) = 0.75 Ω, R
in
(f
r2
) = 0.815 Ω, R
in
(f
ar1
) = 2682 Ω, R
in
(f
ar2
) = 2520 Ω,
X
in
(f
ar1
-) = 1624 Ω, X
in
(f
ar1
+) = -1682 Ω, X
in
(f
ar2
-) = 1344 Ω, X
in
(f
ar2
+) = -1350 Ω.
• Probe fed at 21.0 mm inwards along the antenna from the end:
f
r1
= 228 MHz, f
r2
= 666 MHz, f
ar1
= 405 MHz, f
ar2
= 795 MHz,
R
in
(f
r1
) = 0.733 Ω, R
in
(f
r2
) = 0.806 Ω, R
in
(f
ar1
) = 3054 Ω, R
in
(f
ar2
) = 2154 Ω,
X
in
(f
ar1
-) = 1445 Ω, X
in
(f
ar1
+) = -1654 Ω, X
in
(f
ar2
-) = 1164 Ω, X
in
(f
ar2
+) = -1096 Ω.
• Probe fed at 30.0 mm inwards along the antenna from the end:
f
r1
= 239 MHz, f
r2
= 702 MHz, f
ar1
= 406 MHz, f
ar2
= 799 MHz,
R
in
(f
r1
) = 0.71 Ω, R
in
(f
r2
) = 0.79 Ω, R
in
(f
ar1
) = 2757 Ω, R
in
(f
ar2
) = 1369 Ω,
X
in
(f
ar1
-) = 1425 Ω, X
in
(f
ar1
+) = -1429 Ω, X
in
(f
ar2
-) = 658 Ω, X
in
(f
ar2
+) = -726 Ω.
• Probe fed at 39.0 mm inwards along the antenna from the end:
f
r1
= 247 MHz, f
r2
= 729 MHz, f
ar1
= 406 MHz, f
ar2
= 802 MHz,
Probe Feed Optimization
61
R
in
(f
r1
) = 0.67 Ω, R
in
(f
r2
) = 0.77 Ω, R
in
(f
ar1
) = 2591 Ω, R
in
(f
ar2
) = 737 Ω,
X
in
(f
ar1
-) = 1244 Ω, X
in
(f
ar1
+) = -1352 Ω, X
in
(f
ar2
-) = 374 Ω, X
in
(f
ar2
+) = -427 Ω.
• Probe fed at 48.0 mm inwards along the antenna from the end:
f
r1
= 259 MHz, f
r2
= 763 MHz, f
ar1
= 408 MHz, f
ar2
= 806 MHz,
R
in
(f
r1
) = 0.65 Ω, R
in
(f
r2
) = 0.79 Ω, R
in
(f
ar1
) = 2160 Ω, R
in
(f
ar2
) = 270 Ω,
X
in
(f
ar1
-) = 1130 Ω, X
in
(f
ar1
+) = -1123 Ω, X
in
(f
ar2
-) = 132 Ω, X
in
(f
ar2
+) = -157 Ω.
• Probe fed at 49.5 mm inwards along the antenna from the end:
f
r1
= 262 MHz, f
r2
= 769 MHz, f
ar1
= 408 MHz, f
ar2
= 807 MHz,
R
in
(f
r1
) = 0.65 Ω, R
in
(f
r2
) = 0.75 Ω, R
in
(f
ar1
) = 1951 Ω, R
in
(f
ar2
) = 173 Ω,
X
in
(f
ar1
-) = 1079 Ω, X
in
(f
ar1
+) = -1100 Ω, X
in
(f
ar2
-) = 97 Ω, X
in
(f
ar2
+) = -122 Ω.
• Probe fed at 51.0 mm inwards along the antenna from the end:
f
r1
= 265 MHz, f
r2
= 777 MHz, f
ar1
= 408 MHz, f
ar2
= 807 MHz,
R
in
(f
r1
) = 0.65 Ω, R
in
(f
r2
) = 0.79 Ω, R
in
(f
ar1
) = 1939 Ω, R
in
(f
ar2
) = 148 Ω,
X
in
(f
ar1
-) = 1018 Ω, X
in
(f
ar1
+) = -1071 Ω, X
in
(f
ar2
-) = 62 Ω, X
in
(f
ar2
+) = -84 Ω.
• Probe fed at 52.5 mm inwards along the antenna from the end:
f
r1
= 267 MHz, f
r2
= 784 MHz, f
ar1
= 408 MHz, f
ar2
= 809 MHz,
R
in
(f
r1
) = 0.65 Ω, R
in
(f
r2
) = 0.79 Ω, R
in
(f
ar1
) = 1947 Ω, R
in
(f
ar2
) = 92 Ω,
X
in
(f
ar1
-) = 976 Ω, X
in
(f
ar1
+) = -1016 Ω, X
in
(f
ar2
-) = 39 Ω, X
in
(f
ar2
+) = -51 Ω.
S
11
= -9.63dB at 809 MHz.
• Probe fed at 54.0 mm inwards along the antenna from the end:
f
r1
= 270 MHz, f
r2
= 792 MHz, f
ar1
= 408 MHz, f
ar2
= 809 MHz,
R
in
(f
r1
) = 0.63 Ω, R
in
(f
r2
) = 0.86 Ω, R
in
(f
ar1
) = 1879 Ω, R
in
(f
ar2
) = 48 Ω,
X
in
(f
ar1
-) = 897 Ω, X
in
(f
ar1
+) = -969 Ω, X
in
(f
ar2
-) = 19 Ω, X
in
(f
ar2
+) = -27 Ω.
S
11
= -30dB at 808.511 MHz.
• Optimised feed position- Probe fed at 54.1 mm inwards along the antenna from the
end:
Probe Feed Optimization
62
f
r1
= 270 MHz, f
r2
= 792 MHz, f
ar1
= 408 MHz, f
ar2
= 809 MHz,
R
in
(f
r1
) = 0.63 Ω, R
in
(f
r2
) = 0.83 Ω, R
in
(f
ar1
) = 1873 Ω, R
in
(f
ar2
) = 46 Ω,
X
in
(f
ar1
-) = 951Ω, X
in
(f
ar1
+) = -967 Ω, X
in
(f
ar2
-) = 19 Ω, X
in
(f
ar2
+) = -27 Ω.
S
11
= -36.37dB at 808.75 MHz.
Comments: At a distance of 54.1 mm along the length of the antenna going inwards from
the end, we get an input impedance match of -36.37 at 808.75 MHz. This mode is the
second order mode, and not the fundamental one. The feed location, input reactance
curve, input resistance curve, and VSWR characteristics are depicted separately in Fig.
7.2
















Continuing in a similar manner from the present location, the optimized probe-feed
location for exciting the fundamental mode is searched for.

(a)

(b)

(c)

(d)
Fig. 7.2 (a) Optimized feed location for exciting 2
nd
order mode
(b) Input reactance versus frequency plot
(c) Input resistance versus frequency plot
(d) VSWR characteristics (S
11
in dB)
Probe Feed Optimization
63
• Optimized probe-feed position for fundamental mode excitation – probe fed at 110.7
mm inwards along the antenna from the end:
f
r1
= 383 MHz, f
r2
= 443 MHz, f
ar1
= 413 MHz, f
ar2
= 809 MHz,
R
in
(f
r1
) = 0.413 Ω, R
in
(f
r2
) = 0.44 Ω, R
in
(f
ar1
) = 48 Ω, R
in
(f
ar2
) = 3072 Ω,
X
in
(f
ar1
-) = 24 Ω, X
in
(f
ar1
+) = -23 Ω, X
in
(f
ar2
-) = 1681 Ω, X
in
(f
ar2
+) = -1600 Ω.
S
11
= -23.53 at 413.75 MHz.
The feed location, input reactance curve, input resistance curve, and VSWR
characteristics for excitation of fundamental mode are shown in Fig. 7.3.




















(a)

(b)

(c)

(d)
Fig. 7.3 (a) Optimized feed location for exciting fundamental mode
(b) Input reactance versus frequency plot
(c) Input resistance versus frequency plot
(d) VSWR characteristics (S
11
in dB)
Probe Feed Optimization
64
7.4 Observations from the optimization process
The main observations are listed below:
1. As the probe position changes, the resonant frequencies are shifted. The reason for this is
explained in chapter 8.
2. The anti-resonant frequencies on the other hand, vary very slowly with changing feed
position. They can be considered as almost constant with varying feed position.
3. The observations suggest that good VSWR match occurs when the anti-resonant modes
gets excited. The S
11
curves portray a good VSWR match corresponding to the anti-resonant
frequencies. When good VSWR for a mode is obtained, the highest input resistance for that
mode is nearer to 50 ohms, which helps in impedance match with the input feed.
4. When any mode is excited and good VSWR is obtained, the reactance profile depicts that
the initially high inductive and capacitive reactance in the vicinity of the corresponding anti-
resonant frequency, almost tends towards zero.
5. There are different locations for exciting different modes. The reason is that the excitation
of a mode depends on the field configuration for that mode, and the location where the probe
is fed. If the probe is fed for the minima of a particular mode, that mode cannot be excited.
On contrary, feeding a probe at the maxima of a certain mode facilitates that mode excitation.

7.5 Radiation Pattern for the fundamental mode
The radiation pattern of the antenna for the fundamental mode is shown in Fig. 7.4. The
maximum gain of the antenna at 413.75 MHz fundamental frequency is -36.5895dBi. Fig.
7.4a shows the elevation pattern in the Φ = 0° plane, while Fig. 7.4b shows the same in Φ =
90° plane. The azimuth pattern is shown in Fig. 7.4c.
The gain for the antenna can be increased by:
• Using High Temperature Superconductor (HTS) as the material for the patch
instead of copper.
• Using superstrates over the antenna structure.


Probe Feed Optimization
65























7.6 Summary
The probe feed optimization process for exciting fundamental as well as higher order
modes has been highlighted in this chapter. The antenna is an anti-resonant antenna.

(a)

(b)


(c)
Fig. 7.4 (a) Radiation pattern in Φ = 0° elevation plane
(b) Radiation pattern in Φ =90° elevation plane
(c) Radiation pattern in azimuth plane for θ = 90°













Chapter 8

Lumped Circuit Model
Analysis
Lumped Circuit Model Analysis
66
Chapter 8 Lumped Circuit Model Analysis
8.1 Outline of the chapter
After having discussed the relevant design specific concepts of microstrip Peano line
antennas in chapters 6 and 7, the present chapter focuses on some of the analytical aspects of
the antenna. This chapter forms the most important part of the thesis. Till date, no simple
analytical model except those using complicated numerical techniques is available for the
analysis of these antennas. In this chapter, a simple analytical model is presented to calculate
the resonant frequency of the antenna employing a lumped equivalent circuit model. The
entire theory behind the modeling is presented in general, while the model is employed on
microstrip meander line and microstrip Peano line antennas in particular. Circuit response of
these antennas using our model has been compared with the simulated antenna response
using Method of Moments based simulator IE3D with good agreement, thereby validating
our approach.

8.2 Brief description of the antennas modeled
In the next few sections, the complete derivation of the lumped circuit model is discussed
elaborately. The model is implemented particularly on two types of antennas. At first, the
results pertaining to the meander line antennas are shown due to simplicity of the antenna
structure. Then the results relating to the Peano line antennas are shown. Both the antennas
are for microstrip configuration only, and this model also is entirely based on microstrip
configuration. The resonant
frequency of the antenna as
obtained from the circuit model
is validated against the results
obtained from antenna
simulation response. Fig. 8.1
shows the top view of the two
antennas on which the model is
employed. The figure shows
only the path traced by the
microstrip line, while the substrate height and the ground planes are not shown for the sake of
simplicity. The modeling procedure of these antennas is discussed in the next section.

(a)

(b)
Fig. 8.1 (a) Meander line antenna (b) Peano line antenna
Lumped Circuit Model Analysis
67
8.3 Antenna modeling procedure
Although the derived model is implemented on and verified for the meander and Peano
line antennas in particular, yet the model is expected to be valid for all other antennas based
on space filling curves in microstrip configuration, as the basic theory is the same. To model
the above-depicted antennas in terms of lumped circuit elements we first need to view the
antenna as a cascade connection of several individual line segments. It must also be noted
that, each right-angled bend at the termination of a line segment, needs to be considered
separately. Each segment is then replaced by its equivalent lumped circuit model and the
entire cascaded circuit is terminated by a suitable capacitive termination to account for the
fringing fields from the open-ended line. Each of the steps is described in detail.
8.3.1 Segmentation of the antenna
The antenna is first segmented into basic elements. The segmentation process involves
the dividing of the complicated antenna structure into simpler units comprising of individual
line segments and right-angled bends terminating the line segments. The segments are then
numbered. For ease of understanding, the line segments are represented by odd numbers,
while the right-angled bends are denoted by even numbers. Fig. 8.2 shows the segmented
representation of the antenna structure.













(a)

(b)
Fig. 8.2 (a) Segmented representation of meander line antenna
(b) Segmented representation of Peano line antenna

Lumped Circuit Model Analysis
68
8.3.2 Modeling of line segments
The odd numbered segments in both the antennas shown in Fig. 8.2 represent line
segments th0at can be considered as distributed transmission lines of particular lengths. The
lumped-parameter equivalent for a particular length of transmission line can be employed to
each of these line segments [1-2]. Fig. 8.3 shows the T-equivalent and π-equivalent network
for a two port distributed transmission line.








Expressions for the various impedances, given next, are readily obtained by equating
transmission line terminal voltages and currents with network terminal voltages and currents
[1]:
|
.
|

\
|
=
2
tanh
0
l
Z Z
A
γ
(8.1a)
l
Z
Z
B
γ sinh
0
= (8.1b)
and
l Z Z
Q
γ sinh
0
= (8.2a)
|
.
|

\
|
=
2
coth
0
l
Z Z
P
γ
(8.2b)
where, l is the length of the line segment and Z
0
is its characteristic impedance and γ being
the propagation constant.
Considering the strip to be loss-free, the expressions above are modified to:

(a)

(b)
Fig. 8.3 (a) T-equivalent network for a transmission line
(b) π- equivalent network for a transmission line
Lumped Circuit Model Analysis
69
|
.
|

\
|
=
2
tan
0
l
jZ Z
A
β
(8.3a)
l
jZ
Z
B
β sin
0

= (8.3b)
and
l jZ Z
Q
β sin
0
= (8.4a)
|
.
|

\
|
− =
2
cot
0
l
jZ Z
P
β
(8.4b)
We now consider the case of T-network only. But the model is equally valid if the line
segments are modeled by equivalent π-network.
Since the line segments are very short in length as compared to the effective wavelength,
while operating at a lower frequency range, therefore, the condition l << λ
eff
/4 is satisfied.
Under such condition, (8.3a) and (8.3b) are modified to:
2
0
l
jZ Z
A
β
= (8.5a)
l
jZ
Z
B
β
0

= (8.5b)
The approximation tan x ≈ x & sin x ≈ x, for small values of x has been used.
Z
A
being positive indicates an inductive reactance, while the negative sign in Z
B
indicates a
capacitive reactance.
Z
A
and Z
B
from (8.5a) & (8.5b) respectively can now be expressed in the form of
2
Ll
jω Z
A
= (8.6a)
Cl j
Z
B
ω
1
= (8.6b)
where, L is inductance per unit length, C is capacitance per unit length, and ω is the operating
angular frequency.
Comparing (8.5a) & (8.6a), we get
Lumped Circuit Model Analysis
70
2 2
0
Ll

l
jZ =
|
.
|

\
| β
(8.7)
Substituting,
eff
λ
π
β
2
= ,
eff
eff
ε
λ
λ
0
= ,
f
c
=
0
λ , ω = 2πf
where λ
eff
is the effective wavelength and λ
0
is the free space wavelength, f is the operating
frequency, and c is speed of light in free space, and ε
eff
is the effective permittivity given by
[3] as
w
h
r r
eff
12 1
1
2
1
2
1
+
×

+
+
=
ε ε
ε (8.8)
Where, ε
r
is the relative permittivity of the substrate, h is the height of the substrate and w is
the width of the strip.
Thus, from (8.7) we have,
c
Z
L
eff
ε
0
= (8.9a)
Similarly, comparing (8.5b) & (8.6b), we get,
0
cZ
C
eff
ε
= (8.9b)
From the inductance and capacitance per unit length (L and C respectively), the lumped
inductance and capacitance (L
A
and C
B
respectively) are calculated as follows:
Lumped inductance,
2
Ll
L
A
= (8.10a)
Lumped capacitance, Cl C
B
= (8.10b)
In the above expressions, if Z
0
is expressed in ohms, and speed of light in meters per second,
then the units of L and C are Henry per meter (H/m) and Farad per meter (F/m) respectively.
The corresponding units for lumped inductance L
A
and capacitance C
B
for a length l
expressed in meters are Henry (H) and Farad (F) respectively.

Lumped Circuit Model Analysis
71
Fig. 8.4 shows the lumped equivalent T-network for an
individual line segment. The line segments, each denoted
by odd numbers in Fig. 8.2, can be modeled as an
equivalent L-C-L lumped T-network. The values of L
A
and
C
B
may vary from segment to segment, depending only on
its length, in our case for the antennas considered. However
in general, as evident from the expressions above, the
values shall also vary with the permittivity of the substrate
in case of mixed dielectric, and on characteristic impedance
of the segments, if the segments are of varying widths.

8.3.3 Modeling of right-angled bends
In Fig. 8.2, the even numbered segments represent right-angled bends, each of which can
also be modeled as an equivalent lumped T-network, with inductances in the symmetric arm
and capacitance in the shunt arm. The capacitance arises due to additional charge
accumulation at the corners – particularly around the outer point of the bend where electric
fields concentrates. The inductances on the other hand, arise because of current flow
interruption at the sharp discontinuity caused due to the bend [1]. This is considerable,
especially bearing in mind that most of the current flows in the outer edges of microstrip. Fig.
8.5 shows the right-angled bend and its equivalent circuit.










Fig. 8.4 Lumped equivalent model
of a line segment

(b) Equivalent Circuit

Q
Q'

(a) Structure and nomenclature
w
w
Fig. 8.5 Right-angled microstrip bend
Lumped Circuit Model Analysis
72
Capacitance data has been determined theoretically by Silvester and Benedek [4] and
inductance data by Thomson and Gopinath [5].
Closed form formulae as provided by Gupta et al. [6] are listed below for the evaluation of
bend capacitance:
For w/h < 1:
h w
h w
w
C
r r b
/
) 25 . 2 83 . 1 ( / ) 5 . 12 14 ( − − +
=
ε ε
pF/m (8.11a)
For w/h > 1:
0 . 7 2 . 5 ) 25 . 1 5 . 9 ( + + + =
r r
b
h
w
w
C
ε ε pF/m (8.11b)
and bend inductance:
)
`
¹
¹
´
¦
− = 21 . 4 4 100
h
w
h
L
b
nH/m (8.12)
Here, C
b
is the bend capacitance; L
b
is the bend inductance for a bend formed by the junction
of microstrip lines of width w, on a substrate of height h.
The accuracy of equations (8.11) is quoted as within 5 percent over the ranges: 2.5 ≤ ε
r
≤ 15
and 0.1 ≤ w/h ≤ 5. The accuracy of equation (8.12) is quoted as about 3 percent for the range:
0.5 ≤ w/h ≤ 2.0.

8.3.4 Modeling of the open end
Assuming input excitation at one end of the antenna, the other end is open and needs to
be modeled by some means. It is because, at the open end, there are essentially three
phenomena associated [1]:
a) There are fringing fields extending beyond the abrupt physical end of the metallic strip.
b) Surface waves are launched at the end of the strip.
c) Energy is radiated from the open end, leading to spurious radiation in case of microstrip
circuits, though in our case of antenna, that may be a desired phenomenon if it interferes
constructively.
Lumped Circuit Model Analysis
73
Among the above phenomena, the first one plays a predominating role. Hence, it is the only
one considered for further analysis. The fringing fields from the open end may be accounted
for by assuming some equivalent capacitance connected at the open end.
The expression for the open circuit capacitance C
oc
has been formulated by Silvester and
Benedek [7], which is listed below:
¦
)
¦
`
¹
¦
¹
¦
´
¦
=
|
.
|

\
|
=


5
1
log 2036 . 2 exp
1
i
h
w
K
w
C
i
oc
ε
pF/m (8.13)
where the coefficients K
ε
are given in Table 8.1

TABLE 8.1 COEFFICIENTS K
ε
FOR EQUATION (8.13) [7]
ε
r
1.0 2.5 4.2 9.6 16.0 51.0
i
1

1.443

2.403
2

-0.2535

-0.2220
3

0.1062

0.2170
4

-0.0260

-0.0267 -0.0240
5 -0.0540 -0.0267 -0.0073 -0.0113 -0.0147 -0.0840



8.3.5 Cascading of the equivalent circuit
Having derived the lumped circuit equivalent for each segment, they are duly cascaded
in proper order and terminated by the open circuit capacitance C
oc
at one end (i.e. at the open
end of last numbered segment). The other end (open end of segment number 1) serves as the
input port of the antenna. The whole circuit looks like that as shown in fig. 8.6. This circuit
can be solved using circuit simulators to obtain the input reactance response with respect to
frequency.
Lumped Circuit Model Analysis
74










8.4 Implementation of the lumped equivalent circuit model
In this section, the model derived above is implemented on two antennas, namely
meander line and Peano line antennas. The results relating to both are shown. The dimensions
and the basic parameters for both the antennas are stated. The results obtained from the
analysis of the circuit model are compared with that obtained from the antenna simulation.

8.4.1 Meander line antenna
Dimensions: The first antenna taken for
analysis is the microstrip meander line
antenna, as shown in Fig. 8.7. The footprint
area of the antenna is 25.5 X 25.5 mm
2
. The
width of the microstrip line ‘w’ is chosen as
1.5 mm. The relative permittivity (ε
r
) of the
substrate is taken to be 4.3, with a height ‘h’
as 1 mm. As evident from Fig. 8.2, this
antenna structure when segmented yields 33
segments. Segments 1 & 33 are each 24 mm
long. Other line segments parallel to
segments 1 & 33, viz. 5, 9 etc. are each 22.5

Fig 8.6 Circuit representation of the antenna using lumped models






Fig. 8.7 Simulated antenna structure
(Meander line antenna)

25.5
mm
25.5
mm
1.5
mm
1.5 mm
Lumped Circuit Model Analysis
75
mm long. The remaining odd numbered smaller line segments viz. 3, 7, 11, 15 etc. are each
1.5 mm long, and they constitute the effective radiating elements of the antenna. Apart from
the line segments, there are 16 right-angled bends denoted by even numbers, each of
dimensions 1.5mm X 1.5mm.
Standard analysis formulae in closed form for the calculation of Z
0
are readily available in
literature [1]. They are stated below:
Analysis:
For ‘narrow’ strips (w/h < 3.3):

|
|
.
|

\
|
+
|
|
.
|

\
|
+


¦
)
¦
`
¹
¦
¹
¦
´
¦
+
|
.
|

\
|
+
+
=
π ε
π
ε
ε
ε
4
ln
1
2
ln
1
1
2
1
2 16 4 ln
) 1 ( 2
9 . 119
2
0
r r
r
r
w
h
w
h
Z (8.14)
For ‘wide’ strips (w/h > 3.3):
1
2
2
0
94 . 0
2
ln
2
ln
2
1 1
2
) 16 / ln( 4 ln
2
2
9 . 119

)
`
¹
¹
´
¦
|
.
|

\
|
+ +
+
+
|
|
.
|

\
| −
+ + =
h
w e e
h
w
Z
r
r
r
r
r
π
πε
ε
ε
ε
π
π
π
ε
π
(8.15)
where ‘e’ is the exponential base: e = 2.7182818…
For our case, as all the line segments are of equal width w = 1.5 mm and placed on a single
uniform substrate, therefore the characteristic impedance Z
0
is same for all of them.
W = 1.5 mm and h = 1 mm implies, w/h = 1.5. Therefore, (8.14) is used for analysis.
We obtain from (8.14), Z
0
= 58.14 ohms.
Also, using (8.8), we obtain ε
eff
= 3.2.
Now, for 24 mm line segments:
• L
A
= 4.16 nH, and C
B
= 2.46 pF [using (8.9) & (8.10)]
For 22.5 mm line segments:
• L
A
= 3.9 nH, and C
B
= 2.307 pF [using (8.9) & (8.10)]
For 1.5 mm line segments:
• L
A
= 0.26 nH, and C
B
= 0.1538 pF [using (8.9) & (8.10)]
For right-angled bends:
• L
b
= 0.0688 nH, and C
b
= 0.1387 pF [using (8.11) & (8.12)]
Lumped Circuit Model Analysis
76
For the open-circuited end:
• C
oc
= 0.0329 pF [using (8.13)]
Now having obtained all the circuit parameters, the lumped circuit for each segment is duly
cascaded, and the circuit looks like that in Fig. 8.8. For clarity, the entire circuit is not shown
in detail. In addition, the inductance values are in nH, and the capacitances are in pF, the
units being avoided in the representation.









Results:
The resonant frequency points as observed from IE3D simulations are in good agreement
with the results obtained from MODUA and MDSPICE circuit simulations. The frequency
points of zero reactance, where there is a sharp transition from inductive to capacitive
reactance; correspond to the anti-resonant frequencies. That is evident, if we consider input
resistance versus frequency plots from IE3D, where we observe that at those anti-resonant
points, the input resistance shoots up to a maximum value. The resistance plots have been
The antenna is simulated using IE3D from Zealand Software Inc. As shown in Fig.
8.7, a wave-port at the input end is applied for simulation. Fig. 8.9(a) shows the input
reactance versus frequency plots of the antenna using IE3D, in the frequency range of 0 to 1
GHz. The circuit elements on the other hand are simulated using MDSPICE and MODUA
from Zeland Software Inc. Fig. 8.9(b) shows the input reactance versus frequency plots of the
equivalent circuit using MODUA simulator. Since both simulators MDSPICE and MODUA
show similar resonant and anti resonant properties only the MODUA results are presented
below.

Fig. 8.8 Equivalent circuit of the meander line antenna shown in Fig. 8.7
0.138
Lumped Circuit Model Analysis
77
avoided though, as our aim lies mainly in finding out the resonant frequencies. The anti-
resonant points from both simulations also match closely.























In the above figures, f
r1
, f
r2
and f
r3
represent the 1
st
, 2
nd
and 3
rd
resonant frequencies
respectively, while f
ar1
and f
ar2
represent the anti-resonant frequencies. A comparative study

(a) Antenna simulation results using IE3D

(b) Circuit simulation results using MODUA

Fig. 8.9 Input reactance versus frequency plots (meander line antenna)
Lumped Circuit Model Analysis
78
between the resonant and the anti-resonant frequencies in both figures are presented in Table
8.2.

TABLE 8.2 COMPARISON OF IE3D & MODUA RESULTS FOR MEANDER LINE ANT.
IE3D results
(Antenna)
MODUA results
(Circuit)
% error
1
st
resonance
(MHz)
200.685 181.507 9.55
1
st
anti-
resonance (MHz)
400.685 362.329 9.57
2
nd
resonance
(MHz)
591.096 539.041 8.80
2
nd
anti-
resonance (MHz)
785.6165 715.753 8.89
3
rd
resonance
(MHz)
960.959 886.9855 7.69


8.4.2 Peano line antenna
Having verified the validity
of the model on meander line
antennas, we turn our attention
towards the next class of
antennas- the Peano line antennas.
Dimensions: The dimensions for
the microstrip Peano line antenna
are shown in Fig. 8.10. The
antenna footprint area in this case
also is taken as 25.5 X 25.5 mm
2
.


Fig. 8.10 Simulated antenna structure
(Peano line antenna)

25.5 mm
25.5 mm
1.5mm
1.5 mm
Lumped Circuit Model Analysis
79
As in case of meander line antenna, width of the microstrip line ‘w’ is chosen as 1.5 mm. The
relative permittivity (ε
r
) of the substrate is taken to be 4.3, with a height ‘h’ as 1 mm. Fig. 8.2
reveals that a 2
nd
order Peano line antenna when segmented, yields 81 segments. For our
example, segments 1 & 81 are each 6 mm long. Other smaller segments, parallel to 1 & 81
viz. 5, 13, 21, 25 etc are 4.5 mm long. The longer line segments, parallel to 1 & 81, viz. 9, 17,
37, 45 etc. are 13.5 mm long. The remaining odd numbered line segments of smallest length,
viz. 3, 15, 19 etc are each 1.5 mm long. Each of the right-angled bends, denoted by an even
number, is of dimension 1.5mm X 1.5mm.
For the present case also, as all the line segments are of equal width w = 1.5 mm and placed
on a single uniform substrate, therefore the characteristic impedance Z
0
is same for all of
them.
Analysis:
W = 1.5 mm and h = 1 mm implies, w/h = 1.5. Therefore, (8.14) is used again for analysis.
We obtain from (8.14), Z
0
= 58.14 ohms.
Also, using (8.8), we obtain ε
eff
= 3.2.
Now, for 13.5 mm line segments:
• L
A
= 2.34 nH, and C
B
= 1.384 pF [using (8.9) & (8.10)]
For 6.0 mm line segments:
• L
A
= 1.04 nH, and C
B
= 0.6153 pF [using (8.9) & (8.10)]
For 4.5 mm line segments:
• L
A
= 0.78 nH, and C
B
= 0.4615 pF [using (8.9) & (8.10)]
For 1.5 mm line segments:
• L
A
= 0.26 nH, and C
B
= 0.1538 pF [using (8.9) & (8.10)]
For right-angled bends:
• L
b
= 0.0688 nH, and C
b
= 0.1387 pF [using (8.11) & (8.12)]
For the open-circuited end:
• C
oc
= 0.0329 pF [using (8.13)]
Lumped Circuit Model Analysis
80
The final step is to cascade the individual lumped equivalents of the segments and to
terminate it by C
oc
. The final circuit is shown in Fig. 8.11, where the inductances are in nH
and the capacitances are in pF.








Results:

In this section, IE3D and MODUA results for the Peano line antenna are shown. As
described earlier, IE3D results are the antenna simulation results, while MODUA results deal
with the circuit simulation results. The input reactance versus frequency plots are depicted in
Fig. 8.12.












Fig. 8.11 Equivalent circuit of the Peano line antenna shown in Fig. 8.10


(a) Antenna simulation results using IE3D
Lumped Circuit Model Analysis
81















TABLE 8.3 COMPARISON OF IE3D & MODUA RESULTS FOR PEANO LINE ANT.
IE3D results
(Antenna)
MODUA results
(Circuit)
% error
1
st
resonance
(MHz)
207.534 193.836 6.60
1
st
anti-resonance
(MHz)
413.014 392.466 4.97
2
nd
resonance
(MHz)
606.164 581.507 4.06
2
nd
anti-
resonance(MHz)
806.164 780.137 3.22
3
rd
resonance(MHz) 996.575 967.808 2.88


(b) Circuit simulation results using MODUA


Fig. 8.12 Input reactance versus frequency plots (Peano line antenna)
Lumped Circuit Model Analysis
82
Table 8.3 illustrates a comparative study between IE3D and MODUA results for the Peano
line antenna. The comparative study and the reactance plots reveal that the resonant and the
anti-resonant frequency points from IE3D and MODUA simulations are in close proximity to
each other.

8.5 Discussions
• As listed in Table 8.2 and Table 8.3, a marginal percentage of error (< 10 %) in the
resonant and the anti-resonant frequencies can be attributed to the electromagnetic
coupling between the parallel and adjacently placed line segments, which has been
ignored in our analysis.
• However, the percentage of error is less for a Peano line antenna than for a meander
line antenna. It is because, within the same footprint area and with the same path
length traced, incorporating a Peano line antenna calls for more numbers of smaller
line segments, than relatively longer and lesser number of line segments for a
meander line antenna. As the line segments are smaller in case of Peano line antenna,
lumped approximation is more accurate.
• For accurate evaluation of reflection coefficient and VSWR at the feed, radiation
resistance needs to be calculated. The input resistance of the antenna is the radiation
resistance R
rad
, which essentially depends on the radiated power P
rad
, and hence on the
field configuration
.
This model considers the antenna as a reactive circuit, and thus
provides us with the reactance plots. To obtain proper resistance plots, the field
configuration of the antenna needs to be known, which shall facilitate the accurate
measurement of VSWR.
• The derived circuit model, by virtue of its cascaded structure, can now explain the
change in resonant frequency, each time the probe is fed at a different position, as
observed in chapter 7. Each time the probe feed position along the antenna is
changed; it essentially adds a shunt inductance at a different position of the equivalent
cascaded circuit, thereby changing the entire circuit itself. Thus, the resonant
frequency is bound to change in each case.
• Earlier, the phenomena of multiple resonance or higher order modes were explained
using the concept of distributed transmission lines. It can also be explained using the
Lumped Circuit Model Analysis
83
concept of lumped equivalent circuit model. As the circuit itself is a cascaded
connection of many T-networks, each of the T-network segments has its own resonant
frequency and reactance properties. Moreover, there are several permutations and
combinations of the T-segments grouped together, each having their own resonance
and reactance profile. Finally, the circuit as a whole has its own resonance. All these
reactance profiles taken into account together, explains the reason for multiple
resonance and higher order harmonics.
• This circuit model is valid as long as the dimensions of the individual segments are
small enough as compared to the effective wavelength, for a particular frequency
range of interest.

8.6 Summary
This chapter presented a simple circuit model analysis of a meander line antenna and a
Peano line antenna to compute its resonant frequency. The lumped circuit equivalent of the
antenna has been derived and validated against a standard electromagnetic simulator. The
computed results reveal a close match with an error percentage of less than 10%. Several
inferences from the circuit model were also discussed.

References
[1] T. C. Edwards, Foundations for Microstrip Circuit Design, 2nd Edition, John Wiley
And Sons, 1991
[2] J. D. Ryder, Networks, Lines and Fields, 2nd Edition, Prentice Hall India, 1959
[3] C.A. Balanis, Antenna Theory: Analysis and Design, 3rd Edition, John Wiley and
Sons Inc., 2005
[4] P. Silvester, P. Benedek, “Microstrip discontinuity capacitances for right-angle bends,
T-junction and crossings”, IEEE Trans., MTT- 21, No. 5, pp. 341-346, May 1973
[5] A.F. Thomson, A. Gopinath, “Calculation of microstrip discontinuity inductances”,
IEEE Trans., MTT- 23, No. 8, pp.648-655, August 1975
[6] K.C. Gupta, R. Garg, I. Bahl, P. Bhartia, Microstrip Lines and Slot Lines, 2nd Edition,
Artech House Publishers, 1996
[7] P. Silvester, P. Benedek, “ Equivalent capacitances of micostrip open circuits”, IEEE
Microwave Theory and Techniques, Vol. 20, No. 8, pp. 511 – 516, 1972











Chapter 9

Conclusion and Scope for
Future Works
Conclusion and Scope for Future Works
84
Chapter 9 Conclusion and Scope for Future Works
9.1 Conclusion
The entire thesis deals with an investigative study on microstrip Peano line antennas. In
the initial chapters, the concept of RFID technology has been discussed and the theoretical
background of the antenna has been elucidated. A chapter on review of the antenna has been
included to portray an update of the latest works done in this field.
In Chapter 6, a design problem has been dealt with. Interestingly, more attention has
been given towards the difficulties faced during the design, their causes as well as their
remedies. Chapter 7 provided a brief study of the change in antenna characteristics with
changing feed position. Some important observations regarding changing resonant frequency
with feed position, excitation of different modes etc. has been presented in that chapter.
Chapter 8 dealt with analytical aspect of the antenna. A simple lumped circuit equivalent
model has been presented in the chapter, and its reliability has been validated.

9.2 Scope for future research works
There is a good scope for future research work in this field of study, as most of the
avenues are still unexplored. Some of the major areas are highlighted below, that may be
taken up as a way forward for works in future.
• The design related aspects of the antenna for very low RFID frequencies, viz. 26
MHz for biomedical applications, can be a worthy and challenging aspect of
research.
• Impedance matching related issues might also be dealt with.
• This thesis gives some insight into design and circuit analysis studies on the
antenna. However, the analyses of radiation characteristics are yet to be explored.
This provides an interesting area of research. The structure being a highly
convoluted one, the analysis of its radiation pattern is not so straightforward. One
suggestive method is to segment the antenna into several rectangular cavities,
apply the cavity model [1] of microstrip patch antenna on each of them, and
finally integrate the total system by using Multiport Network Model [2]. Another
method is the use of numerical techniques to evaluate the current distribution on
Conclusion and Scope for Future Works
85
each of the individual segments, and finally the vector addition of radiation in
free space, due to each of them.
• Once the radiation pattern analysis is possible, the radiation resistance can be
evaluated from the radiated power. This completes the circuit model, and the
VSWR characteristics at the feed can then be accurately determined.

References
[1] C.A. Balanis, Antenna Theory: Analysis and Design, 3rd Edition, John Wiley and
Sons Inc., 2005
[2] R. Garg, P. Bhartia, I. Bahl, A. Ittipiboon, Microstrip Antenna Design Handbook,
Arctec House Inc., 2001


FACULTY OF ENGINEERING AND TECHNOLOGY JADAVPUR UNIVERSITY CERTIFICATE
This is to certify that the dissertation titled “Investigations on Microstrip Peano Line Antenna” has been carried out by Arkaprovo Das (University Registration No.:108453 of 2009-10) under my guidance and supervision and be accepted in partial fulfillment of the requirement for the degree of Master of Electronics & Telecommunication Engineering. The research results presented in the thesis have not been included in any other paper submitted for the award of any degree to any other University or Institute.

________________________ Prof. Bhaskar Gupta
Head of the Department Electronics and Telecommunication Engineering, Jadavpur University, Kolkata- 700032. ___________________________________

Prof. Niladri Chakraborty
Dean Faculty of Engineering and Technology, Jadavpur University, Kolkata- 700032.
___________________________________

Prof. Bhaskar Gupta
Project Supervisor Department of Electronics and Telecommunication Engineering, Jadavpur University, Kolkata-700032

FACULTY OF ENGINEERING AND TECHNOLOGY JADAVPUR UNIVERSITY

CERTIFICATE OF APPROVAL*

The foregoing thesis is hereby approved as a creditable study of an engineering subject carried out and presented in a manner satisfactory to warrant its acceptance as a pre-requisite to the degree for which it has been submitted. It is understood that by this approval the undersigned do not endorse or approve any statement made, option expressed or conclusion drawn therein but approve the thesis only for the purpose for which it is submitted. Committee on final examination for the evaluation of the thesis ________________________ (Signature of the Examiner)

_________________________ (Signature of the Supervisor)

*

Only in case the thesis is approved

I have fully cited and referenced all material and results that are not original to this work. I also declare that. as a part of his degree of Master of Electronics and Telecommunication Engineering studies. All information in this document have been obtained and presented in accordance with academic rules and ethical conduct. Name Examination Roll Number Thesis Title : ARKAPROVO DAS : M4ETC11-01 : INVESTIGATIONS ON MICROSTRIP PEANO LINE ANTENNA Signature with date : . as required by these rules and conduct.FACULTY OF ENGINEERING AND TECHNOLOGY JADAVPUR UNIVERSITY DECLARATION OF ORIGINALITY AND COMPLIANCE OF ACADEMIC ETHICS I hereby declare that this thesis titled “INVESTIGATIONS ON MICROSTRIP PEANO LINE ANTENNA” contains literature survey and original research work done by the undersigned candidate.

Sanghamitra Dasgupta. Working with him during the entire tenure was indeed an absolute pleasure. Mr. Arkaprovo Das . Avra Kundu and to my friends Santanu. directly or indirectly involved with me. In addition. Prof. Dr. I would certainly like to thank my relatives and well-wishers. Sudhabindu Roy. D. Lastly. Amrita. I would certainly like to mention that all the discussions I had with him on technical issues. Anupam Das. I dedicate this thesis to my grandfather Late Prafulla Kumar Sarkar. I am also grateful and thankful to my teachers at Jadavpur University. Mina Das towards my proper upbringing and for all the encouragement and support they rendered me for pursuing higher studies and research. I would like to thank him for the faith he has shown in me to take up this work and proceed further. Sudipta Maity.. Mr. Mr.ACKNOWLEDGEMENT I would like to take this opportunity to thank a few individuals. I would like to convey my sincere gratitude towards my project supervisor Prof. Rajendra Prasad Ghosh. Poddar. First and foremost. helped me grow my analytical thinking capacity to a great extent. Most importantly.R. I thank the Almighty and pray for everyone’s wellbeing. to name a few. Jadavpur University for his constant guidance. I also thank all my other seniors and colleagues whom I could not mention in this section. S. without which this work would not have been initiated. Mr. I must convey special thanks to my friend and senior. Mr. Their lectures were immensely beneficial which helped me strengthen my knowledge base for pursuing research work. Sankaralingam. Sayantan Dhar. Above all. Asim Kar for their full support and cooperation during my stay at the institute. who certainly would have been very happy on seeing the successful completion of this dissertation. who has helped me out from all sorts of difficulties that I faced while working on my project. Bhaskar Gupta. and my mother Mrs. Head of the Dept. of Electronics & Telecommunication Engg. Rahul and Ankita for providing their support and unforgettable moments which I shall certainly cherish throughout my life. for the successful completion of this project. My whole-hearted thanks to my senior colleagues Ms. Prof. support and best wishes. Sonali. I cannot ever forget the contribution of my father Mr.

5 Outline of the chapter Fractals Space-filling curves Fractal Dimension Hilbert curves 10-16 10 10 10 12 13 .7 Design criteria for RFID tag antennas Summary References 3.2 2.5.3 Preface Objective of the thesis Organization of the thesis 1-2 1 1 2 2.1 2.1 2.1 3.5 RFID tag or transponder RFID reader 3-9 3 3 4 4 4 5 5 6 7 8 9 9 RFID frequency bands and applications RFID antennas 2.3.6 2.4 3.3 3. What is RFID and what are the Antennas Used 2. An Introduction to Space-filling Curves 3. INTRODUCTION 1.5.4 2.2 1.2 2.1 2.3 Outline of the chapter RFID Components of RFID systems 2.3.1 1.2 Near-field systems Far-field systems 2.2 3.CONTENTS 1.

3.1 5.2 4. A Brief Review of Peano Line Antenna 5.4 Outline of the chapter Advancements in miniaturization of antennas Review of Peano line antennas Summary References 25-30 25 25 26 29 29 6.2 6. RFID Tag Antenna Design Using Peano Curves 6.6 3. Peano Line Antennas 4.3 5.8 Outline of the chapter Concept of miniaturization Electrically small antennas Miniaturization of wire antennas Peano line wire antennas Miniaturization of microstrip antennas Microstrip Peano line antennas Summary References 17-24 17 17 17 18 18 19 20 23 23 5.5 4.1 4.2 5.6 4.3 4.1 6.8 Peano curves Process of generating 2nd order Peano curves Summary References 14 15 16 16 4.7 3.7 4.3 Outline of the chapter RFID frequency bands Design Goal 31-57 31 31 31 .4 4.

1 7.3 Approach 1: Increase of path length of the antenna Approach 2: Higher permittivity substrate Approach 3: Input impedance match using quarter wave transformer 6.6.2 7.7 6.5 6.7 Approach 7: Single stub matching 32 32 34 35 42 44 46 46 47 51 56 56 56 6.6.3 7.4 Approach 4: Use of different dielectric materials for feed and antenna 6.6.6 Approach 5: Use of stacked dielectric layers Approach 6: Modification of the structure described in approach 5 6.5 6.6.5 Outline of the chapter Antenna parameters Variation in probe feed position Observations from the optimization process Radiation Pattern for the fundamental mode 58-65 58 58 58 64 64 . Probe Feed Optimization 7.6.6.8 Discussions Summary References 7.2 6.4 6.4 7.1 6.6.6.6 Design Procedure Design prototype Design approaches 6.

2 Conclusion Scope for future research works References 84-85 84 84 85 .1 8.5 8.2 8.4 8.2 Meander line antenna Peano line antenna 8.4 Segmentation of the antenna Modeling of line segments Modeling of right-angled bends Modeling of the open end Cascading of the equivalent circuit 66-83 66 66 67 67 68 71 72 73 74 74 78 82 83 83 Implementation of the lumped equivalent circuit model 8.2 8.7.3.3 8.1 8.3.3.5 8.3 Outline of the chapter Brief description of the antennas modeled Antenna modeling procedure 8. Lumped Circuit Model Analysis 8.4.1 9.4.3.1 8.6 Summary 65 8.6 Discussions Summary References 9.3. Conclusion and Scope for Future Works 9.

Chapter 1 Introduction .

Introduction

Chapter 1
1.1

Introduction

Preface In mathematical analysis, a space-filling curve is a curve, which helps in filling up a

two-dimensional space. Giuseppe Peano (1858-1932) was the first to discover one, which is commonly known as Peano curve. A few space-filling curves, other than the Peano curves are Hilbert curve, Moore curve, Sierpinski curve etc. These space-filling curves find immense application in the field of antenna engineering. When the wire itself takes the shape of a space-filling curve, and radiates in free space, they are referred to as space-filling curve wire antennas. Again, in case of microstrip antennas, the patch can be meandered by tracing the shape of space-filling curves, which are then referred to as space-filling curve microstrip antennas. While extensive research work in terms of antenna performance has been performed till date in the former category, the latter category is a relatively newer field of study, and requires greater attention. This thesis deals with some investigations and studies on microstrip Peano line antennas, where the patch of the microstrip antenna traces the shape of a Peano curve. This configuration is still newer among the latter category stated above, and hence calls for extensive research work.

1.2

Objective of the thesis The past decade has seen phenomenal advances in portable electronics technology like

mobile phones, RFID tags and MP3 players. This has led to the development of System in Package (SiP) which combines all the necessary components into a single package. Miniaturization of RF circuit technology has resulted in the need for miniaturized antennas. Antennas based on space filling curves are mainly used where compactness and miniaturization are the key objectives. Considering that purpose, microstrip Peano line antennas have been explored in the following chapters. This thesis highlights the applications, where these antennas are used. It also provides a brief theory, both from graphical point of view, as well as from the viewpoint of antennas. It enumerates in detail the approaches towards a particular design problem, and the variation in input impedance characteristics with varying probe feed position. Most importantly, a simple lumped circuit equivalent model of the antenna is derived and presented, which forms the most important part of the thesis.
1

Introduction

1.3

Organization of the thesis This thesis is divided into nine chapters. Chapter 2 deals with an introduction to RFID,

and the antennas used for miniaturization. Here the contextual application of Peano line antennas is also presented. Chapter 3 gives some insight on the graphical theory of Peano curves, while the basic theory of Peano line antenna is briefly presented in Chapter 4. Chapter 5 highlights the historical background of Peano line antennas, and the research work that has been carried out in this field in the recent years. The next three chapters describe mainly the research work done during the M.E. project. Chapter 6 highlights a particular design problem, and the problems faced therein. In addition, the problems faced are modified with better alternatives, and are studied. To get further idea on microstrip Peano line antennas, Chapter 7 deals with a few studies on variation of probe feed location along the antenna, where the patch traces a second iteration Peano curve. In Chapter 8, a simple lumped equivalent circuit model of the antenna is derived and presented. In addition to that, the same for microstrip meander line antenna is also presented, and a comparison for both the cases is described. Finally, Chapter 9 summarizes the entire project, with a concluding note on scope for future works, and a suggestion for a way forward.

2

Chapter 2

What is RFID and what are the Antennas Used

and detailed mathematical formulations for this chapter are beyond the scope of this thesis. 2.normally digital. a rapidly growing technology. which uses RF signals for automatic identification of objects.1. smart cards and shop security etc [1]. 2. The entire discussion is qualitative only.1 A simplified RFID system [1] 3 . RFID systems are short ranged. electronic passports. Fig. A simplified diagram of RFID system is shown in Fig.1 What is RFID and what are the Antennas Used Outline of the chapter This chapter presents a very brief introduction to RFID technology and the antennas used for it. asset tracking. The usefulness of miniaturized antennas is also mentioned in this regard.2 RFID RFID is an abbreviation for Radio Frequency Identification. wireless systems. 2. Examples of applications include animal tagging.What is RFID and what are the Antennas Used Chapter 2 2.

2. However. Types of RFID tags: a) Passive tags: These are the tags. at one end of the system is the RFID tag or transponder. which is analogous to the Universal Product Code (UPC) format used by Bar codes. Types of RFID readers: a) Mono-static readers: In this case.3 Components of RFID systems There are two main components of an RFID system.e. the fields from the reader serves as energy excitation for these tags to be operational.3.What is RFID and what are the Antennas Used 2.1.e.3. they contain some user memory. a system that is placed on the objects to be identified. They are comparatively cheaper in cost. 4 . The reader is usually connected to a host network or a computer. The tag contains an integrated circuit (IC). 2. the transmission and reception of EM signal uses the same antenna. i. The tags have an Electronic Product Code (EPC) label. They are used in high data rate communications. b) Semi-active tags: Semi-active tags have limited power supply. a reader can also write or transfer data to the tag. The tags are also programmable and can contain userspecific information. i. contrary to its name. c) Active tags: They have inbuilt battery power supply which makes them functional. They are: • • Tag or transponder Reader 2. which needs to be powered by the RFID reader. without relying on external aid. prompts communication with the tags.2 RFID reader An RFID reader or interrogator on the other hand. and an antenna to receive EM waves from the reader for its identification.1 RFID tag or transponder As shown in Fig. and hence allows faster operation than the passive tags.

often called 13. used for electronic ticketing.56 MHz band.957-27. • • 2. and thus eliminates any likelihood of complete polarization mismatch at the tag.5 KHz (LF) – can be used globally without a license. not so widely used.454 GHz (SHF) – Backscatter coupling. It is because. contactless payment etc. However. 2. work in progress tracking etc. used for special applications. a circularly polarized wave can be decomposed into two mutually orthogonal. only used in USA/Canada.4 • RFID frequency bands and applications [2] 125-134.567 MHz (HF) . • • 433 MHz (UHF) – Backscatter coupling. garment tracking. AVI (Automatic Vehicle Identification). 2. This is important in tagging applications where high reliability is required. 13. The scheme is shown in Fig. 2. used for remote car keys. used for asset management.4-2. container tracking. and may be either balanced or unbalanced.5 RFID antennas In general. • • 6.553-13.765-6. The reader antenna is generally electrically larger than the tag antenna.446-2. often used in vehicle identification. to suit the differential inputs of the tag IC. Reader antennas are often circularly polarized when the orientation of the tag is unknown. 2.483 GHz (SHF) – Backscatter coupling. vertical and horizontal linearly polarized components. baggage tracking. 858-930 MHz (UHF) – Restricted usage.2 KHz and 140-148. It is a linearly polarized electrically small antenna.725-5. used for long-range tracking and with active tags. the antenna for the RFID tag is a balanced one.795 MHz (MF) – inductive coupling.1. • 5. there is a 3-dB loss between 5 . • 26.What is RFID and what are the Antennas Used b) Bi-static readers: In bi-static readers. where the second optional antenna is also used.283 MHz (HF) – Inductive coupling. access control. biomedical applications.875 GHz (SHF) – Backscatter coupling. the antennas and the associated circuitry are different.

Fig. multi-turn coils are used as tag and reader antennas. The coupling between the antennas is dependent on their separation. with the coil in the reader serving as the primary winding. Generally. however. and always occurs in the near-field region. it may also be capacitive.1 Near-field systems Near-field systems use coil/loop antennas for communication between the reader and the tag. LF and HF systems use near field coupling between the reader and the tag antennas. where the tag can be placed on a lossy dielectric such as an animal or a solid object etc.What is RFID and what are the Antennas Used circularly and linearly polarized antennas and hence maximum power transfer between the reader and tag cannot be achieved. When the coupling is inductive. 2. than that of an antenna. while that in the tag serving as the secondary.5. The effect is more like that of a transformer. Far-field coupling is used generally for UHF systems. where the transfer of energy between the reader and the tag is determined by Friis equation [3]. 2. the coupling is inductive. The reader and tag loop antennas are shown in Fig.2 Simplified Reader and Tag loop antennas [1] 6 . 2. The reader and tag antennas with radii a & b respectively are separated by a distance d.2. This near-field coupling method is used in applications.

2. 2. along the z-axis. Patches are often used to provide circular polarization (shown in Fig.3 Rectangular Tag coil [1] Fig. 2. PIFAs and patches are among the options at the reader. credit card dimensions are often used for smart cards. 2. 2.5). The IC is connected to the strap prior to being connected to the coil.3. though coil antennas are very rarely used. 2.2 Far-field systems Systems with carrier frequencies greater than 100 MHz generally operate by transferring power in the far field. Multi-turn loops are also often employed in order to increase the mutual inductance. A wide variety of antennas is possible in the far field systems. 2. also uses this method of RFID communication.5.What is RFID and what are the Antennas Used The magnetic field produced due to the coil antennas is greatest along the axis of the loop. For instance.4 Coil antenna used within passport [1] The antenna in Fig.4 has five turns using thin wires. The tag antenna size and geometry is governed by the application. 2. i.e.4 depicts the photograph of an example of a commercial coil antenna. Dipoles.45 GHz. A typical far field system is shown in Fig.5. ISM band centered around 2. Fig. A typical rectangular tag antenna is shown in Fig. while Fig. Modified dipoles are commonly used as tag antennas.used in a European passport. Fig.5 Typical far field reader and tag antennas [1] 7 . 2.

PR is the power received by the tag. Size and form: The size and form of the tags depend on the applications.1) where. Those RFID systems.6 Design criteria for RFID tag antennas The requirements that largely determine the design of an RFID tag antenna are enumerated below [4]: 1. The read range depends on: • • • EIRP: Effective Isotropic Radiated Power is determined by local country regulations. 4. where the tag is to be used. 2. in order to facilitate miniaturization. Read range: The maximum distance at which the RFID reader can detect the backscattered signal from the tag is defined as the read range. Doppler shift must be accounted for. Frequency band: the desired frequency band depends on the regulations formulated by the country. PT is the power transmitted by the reader.What is RFID and what are the Antennas Used The transfer of energy between the reader and the tag is governed by Friis transmission equation as stated below [3] PR  λ  =  GT GR PT  4πd  2 (2. and hence affirm the need for miniaturized antennas. Orientation: Read range depends on antenna orientation. require miniaturized RFID circuitry. 3. GT & GR are the gains for reader (Transmitter) and tag antennas (Receiver) respectively. or other antennas based on space-filling curves are being used lately. λ being the wavelength corresponding to the frequency of signal transmitted. which are used for biomedical applications. 2. such as omnidirectional or hemispherical coverage. Some applications require tags to have a specific directive pattern. where it is to be used. in the computations. meander line antennas. Objects: Tag performance changes when placed on different objects. separated d distance apart. Among the wire and microstrip antenna configurations. Applications with mobility: In case of mobile tag applications. 8 .

1946 K. so as to reduce the total cost of the RFID tag. December 2005 9 . and their characteristics were discussed. 2008 http://www. John Wiley & Sons Ltd. No. References [1] [2] [3] [4] Y.Seshagiri Rao. As this technology does not require Line of Sight (LOS) propagation. P.V.Boyle. Proceedings IRE. The antennas used for this technology were also illustrated briefly. “A note on a simple transmission formula”.Huang and K. This puts restriction on the choice of antenna structure. 2. As presented in earlier sections. 53.F.Lam.radio-electronics. “Antenna Design for UHF RFID Tags: A Review and a Practical Application”.Friis. Vol.Nikitin. IEEE Transactions of Antenna and Propagation. S.V. due to the applications on which it is used.What is RFID and what are the Antennas Used 5. 41.T. it is evident that.com/info/wireless/radio-frequency-identificationrfid/low-high-frequency-bands-frequencies. this technology is a growing technology. it is replacing the optical Bar code technology very rapidly. pp. Cost: The antennas should be of low cost.php H.7 Summary A brief account on RFID technology has been portrayed in this chapter.254256. 12. Antennas from Theory to Practice. and the choice of materials used for its construction.

Chapter 3 An Introduction to Spacefilling curves .

that is repeated at even-smaller scales to produce irregular shapes and surfaces that cannot be represented by classical geometry. the property of a space-filling curve may be explained as: if a particular geometrical curve occupies a definite area in 2-dimensional space in its 1st iteration. Peano curves constitutes the most important part of this chapter. Thus.2 Fractals A fractal is defined as “a rough or fragmented geometric shape that can be split into parts. with reduced magnification. 3." [1] a property called self-similarity. often in inter-connected fashion to obtain the final structure. are patterned on this curve. Process of generating a standard 2nd iteration Peano space-filling curve is also explained. orientations and positions. Few of the space-filling curves are also explained in the text. Fractals consist of identical and similar elements repeated in different magnifications. and many repetitive patterns. Two important properties of fractal shapes are self-similarity and scale-invariance. as the iteration order increases. The term fractal was coined by Benoit Mandelbrot in 1975 and was derived from the Latin word fractus meaning "broken" or "fractured" [2]. since the antennas described in the later chapters. 3. these curves effectively aid in filling up a particular space (in 2-D). or a particular volume (in 3-D). One important feature of fractal geometry in 2-dimension is. yet the curve occupies a finite area. the curve 10 . This point is explained later. However. each of which is (at least approximately) a reduced-size copy of the whole.An Introduction to Space-filling Curves Chapter 3 3. the length of the curve traced becomes infinite. In simpler words. it occupies the same area. then in higher iterations. as the iteration order increases. As the name suggests. after infinite numbers of recursive iterations. fractal forms a part of non-Euclidean geometrical shapes. A fractal shape is generated using recursive algorithms.1 An Introduction to Space-filling Curves Outline of the chapter This chapter presents a very brief description of the space-filling curves in general.3 Space-filling curves Space-filling curves are subset of a broader class of fractals. In other words.

1. The first six iterations for Hilbert curve are shown in Fig. Hilbert curves. Peano curves.An Introduction to Space-filling Curves traces more path.rotated. the repetition of which generates the higher iterations. yet confined in a definite area. this second iteration Hilbert curve is the unit cell for the third iteration. or keeping it intact. 2. Now.intact) to fill the same space in second iteration.1 Six iterations of the Hilbert curve construction [3] Construction: The first iteration is the unit cell. The repetitive pattern brings in the concept of self-similarity as described earlier. 1st iteration 2nd iteration 3rd iteration 4th iteration 5th iteration 6th iteration Fig 3. the first iteration curve is used in the second iteration with reduced size. 3. and helping in compression. Examples: Some of the common types of space-filling curves are Moore curves. As seen from the figure above. The unit cell of first iteration is repeated 4 times (2. Additional line segments are added in order to ensure continuity of the curve. and goes on with increasing iteration orders. Wunderlich curves etc. 11 . This process is a recursive one. either rotated by 90 degrees clockwise/anticlockwise.

then 4 such smaller squares are required to fill in the space by the original bigger square. or. then it would require 9 such squares to fill the bigger square of length l. N = n. and D is the dimension. Therefore. it would produce the bigger square. some examples are considered. where for this case. each of those smaller line segments when magnified by a factor 2.4 Fractal Dimension Fractal dimension is an important parameter to quantify the space-filling ability of a space-filling curve. Let it be n. Similarly. each one of them when magnified by a factor ‘n’ yields the original line segment. each of the segments is of length l/2. As seen above. 9. if the sides of any of the 9 smaller squares be trebled. with sides of length l. each of them may be magnified by a factor of 3 to get the original segment. 16 etc. n = n1. the same parameter must approach a value of 3 so as to effectively fill a 3-D volume. and D = 1. if a line is divided equally into ‘n’ self-similar smaller line segments. we can say each of the smaller 9 squares have a magnification factor 3. i. the bigger square can be divided equally into n number of smaller self-similar squares. if a line segment is divided into three parts. In the above expressions.An Introduction to Space-filling Curves 3. in general. The second case of a square suggests that. i. if l/3 be the length of each side of the smaller squares. each with magnification 2. number of smaller squares. 12 .e. the fractal dimension must approach 2. Let us consider a line segment of length l. N is the magnification factor. Similar examples may be taken for the case of cubes. D = 1 indicates. If we now assume a smaller square with each side of length l/2. n is the number of smaller line segments. To illustrate the above point. a straight line is one-dimensional. yields the original line segment. Let us now consider the case of a square. When the line is equally halved. Hence. For the first case of a straight line. Alternatively. more generally. it can be expressed as unit power of n. it can be divided into 4. it can be divided equally into any numbers of smaller selfsimilar segments. the bigger square can be divided into 4 smaller self-similar squares. n = ND. Thus.e. n being a perfect square number. Similarly. Similarly. For this case. For a space-filling curve to effectively fill a 2-D space.

D = 2 Which clearly suggests. D is the dimension.1) For regular geometries. Taking logarithm on both sides of (3.1). while for fractal geometries.An Introduction to Space-filling Curves So. then 9 = 32. D = log n / log N (3. 3.2 Fig. maximum value of D = 2. 3. the following relation holds true: n = ND (3. the maximum attainable fractal dimension D = 1. where n = 9. or. log n = D log N.5 Hilbert curves The basic pictorial representation of Hilbert curves is depicted in Fig.1 (b) Iteration .1.e. we get. dimension or fractal dimension is defined as the ratio of the logarithm of number of self-similar pieces to the logarithm of the magnification factor. In this section.2) Therefore. for any general D-dimensional space.26. N = 3.2). 3. 3. i. For a Koch fractal curve. This clearly suggests that. the values of the dimensions of its line segments are provided (Fig. it occupies a definite area in 2dimensional plane. while for a Hilbert curve. D is defined as ‘Fractal Dimension’. Therefore. it fills the 2-D space more effectively than a similar order Koch curve. as the iteration number of a Hilbert curve increases.2 Hilbert curve 13 . the dimension of a square is 2. if a square is divided into 9 smaller squares. Dimensions: l d=l/3 d=l l (a) Iteration .

An Introduction to Space-filling Curves Let l be the side dimension of the Hilbert curve. we find that for a definite footprint area and fixed iteration order. Then. 3. and n be the iteration order. There are different types of Peano curves – the standard first Peano curve. the dimensions of the curves are given as follows: d= l . d= l .3.6 Peano curves The first space-filling curve was discovered by Giuseppe Peano in the year 1890. minimal N-shaped curve etc. and it is shown in Fig. and 3 −1 n S = ( 32n – 1 ) d = ( 3n + 1 ) l Comments: If we observe the length dimensions for the Hilbert and Peano curves. with d being the length of each smallest unit of line segment. A Peano curve has a fractal dimension of the value 2. Let the total length of the curve be S.3 Peano curves [3] (c) Third iteration Dimensions: For a Peano curve. (a) First iteration (b) Second iteration Fig. is greater than that covered by a Hilbert curve. a Peano curve offers more compression and space-filling property than a similar Hilbert curve. and 2 −1 n S = ( 22n – 1 ) d = ( 2n + 1 ) l 3. It was named after him as Peano curve. The first Peano curve is the most commonly used one for antenna design purposes. This suggests that. the path length S covered by a Peano curve. 3. This 14 . for a Hilbert curve.

space containing the first iteration curve) into 9 equal squares. which asserts Peano line antennas as better candidates for miniaturization.e. till we arrive at square number 9 15 . especially in case of antenna design. and so on. and its mirror image. 3. alternatively with the reduced Peano 1st iteration curve and its mirror image. The square number 1 contains reduced 1st iteration curve. square number 2 contains its mirror image. They are depicted below: Original Curve Mirror image Step 2: Reduce the scale of the size of first iteration curve and its mirror image Step 3: Divide the square space to be filled (i. 3 2 1 4 5 6 9 8 7 Step 4: Go on filling the smaller squares in the numbered order stated above.An Introduction to Space-filling Curves contrasting property is beneficial. consider the first iteration.7 Process of generating 2nd order Peano curves Step 1: First.

taking the second order Peano curve and its mirror image for a unit cell.org/wiki/Space-filling_curve 16 . 1983 http://en. Freeman & Co. that can be applied to generate any higher order of the same curve. This process is a recursive one. The same process.B.H. Also the process of generating a second order Peano curve has been discussed. 3. and can be applied to generate any higher iteration. The generated curve is the second iteration Peano curve.8 Summary This chapter presented an overview of the space-filling curves in general and its concepts. The Fractal Geometry of Nature. Mandelbrot. References [1] [2] [3] B. can be repeated to produce a third order Peano curve..wikipedia. New York: W.wikipedia. while emphasizing mainly on the Hilbert and Peano curves.An Introduction to Space-filling Curves Step 5: Add extra line segments to ensure the continuity of the curve.org/wiki/Fractal http://en.

Chapter 4 Peano Line Antennas .

In the context of SAL. It is defined as the upper frequency boundary. then it certainly aids in size reduction. This chapter presents the application of the same curve in antenna engineering. at which an antenna may be considered as electrically small [1].Peano Line Antennas Chapter 4 4. In that context. Wheeler defined an electrically small antenna as one whose maximum dimension is less than λ/2π [2]. • Size miniaturization: The above concept can be rephrased in a different manner. if the antenna can realize a lower resonant frequency. This relation is often expressed as: ka < 1 (4. a brief idea on Peano space-filling curves has been presented. yet not exceeding the total physical size. 4. where in terms of effective wavelength. as compared to the conventional design at lower frequency. the size is much larger. Although they mean the same thing. 4.1 Peano Line Antennas Outline of the chapter In chapter 3. Since the design of a resonant antenna depends on the effective wavelength. as explained in last point. yet they may be considered separately: • Frequency Reduction: Compared to a particular conventional antenna design. if any other design can reduce the resonant frequency of the antenna.2 Concept of miniaturization Miniaturization of an antenna can be thought of from two different viewpoints. the basic theory regarding miniaturization of antennas is also discussed.3 Electrically small antennas An antenna is said to be small when its size is much smaller than the operating wavelength. then the new design can be considered to be as a miniaturized one. Small Antenna Limit (SAL) determines the criteria for an antenna to be electrically small. hence a lower frequency antenna design requires the physical size of the antenna to be much larger. without exceeding the antenna size. However.1) 17 .

5 Peano line wire antennas In Fig. and hence it results in more frequency reduction for the antenna. the wire takes meandering turns to Fig. called meander line antenna. or by following any other space-filling curve. λ is the free space wavelength. A sphere of radius a = 1/k = λ/2π is defined as the radiansphere. 4. Thus. yet they suffer from certain drawbacks. when the wire traces a space-filling curve. a is the radius of a sphere circumscribing the maximum physical dimension of the antenna.Peano Line Antennas where.1. when an antenna can be enclosed into a radiansphere of radius a. Therefore. if the meander pattern of the wire is replaced by a Peano curve pattern.1 Meander line wire antenna reduce the resonant frequency. an antenna is electrically small for a particular frequency of interest. From (4. As shown in chapter 3. It follows directly that. the space-filling property of the curve increases. The compression property of spacefilling curves has been discussed in chapter 3. and the whole antenna is placed over a ground plane in a monopole like fashion. In the figure. and hence causes a reduction in the resonant frequency of the antenna. 4. leading to miniaturization. a frequency reduction. 4. without exceeding Ground Plane the physical size limits. k = 2π/λ. just like that 18 . Though these space-filling wire antennas result in miniaturization. One particular simple configuration of a meander line wire antenna is shown in Fig.4 Miniaturization of wire antennas Wire antennas can be miniaturized if the antenna length of a conventional dipole or monopole design is extended following a particular meandering pattern. 4. As the iteration order increases. the current path over it follows an extended path. When used as a vertical monopole antenna. it can be inferred that. causes an antenna to be electrically small. thus helping in miniaturization.1. it is said to be an electrically small antenna. 4.1). the Peano curve may be of any iteration order. then it forms a Peano line wire antenna.

• Use of shorting pins: Use of edge shorted patch. Therefore. or shorting pins at the vicinity of a coaxial feed are well known techniques for size reduction of a microstrip antenna [4-7]. since the length is of the order of half of the effective wavelength.2) implies that the resonant length L is proportional to 1 εr at a fixed resonant frequency. This reduces the resonant frequency of the antenna. In such cases.2) where c is the speed of light in free space. 4. Hence. 19 .Peano Line Antennas shown in Fig.1. the problem of impedance matching is a serious concern. and hence. different methods are to be employed to reduce the size of the antennas. 4. For these reasons. They are: • Use of high permittivity substrate: Rectangular patch antennas are operated at a resonant frequency fr. L is the length of the patch and εr is the permittivity of the substrate used. The microstrip patch antennas have a significant number of advantages over conventional antennas. Secondly. which eventually increases the total size of the package. and thus hampers our original goal for miniaturization. Nevertheless. these antennas require a large horizontal ground plane. where the space-filling property of the curves is utilized for miniaturization. the length of the patch required for an operation at lower frequency range at its fundamental mode is too high. smaller and cheaper antennas are required to be integrated into electronic devices. a high permittivity substrate results in smaller antenna length. which effectively increases the inductive effect. we switch over to microstrip configuration antennas. This calls for the introduction of miniaturized planer antennas in most of the applications. the antenna is fed at the end as a monopole. where fr is given by [3]: fr = c 2L ε r (4. thus increasing the magnetic field associated with it. The shorting post or pin adds an extra path for the current to flow.6 Miniaturization of microstrip antennas With growing demands for miniaturized circuitry in today’s world. Thus (4. the input impedance cannot be changed by changing the feed position.

the closely placed 20 . and hence miniaturization. In addition. which tends to reduce the resonant frequency of the antenna. • Use of space-filling curve: Miniaturization can also be effected by the introduction of space-filling curves [10-12]. 4. The fabricated photograph shows a Peano line antenna fabricated on a Duroid substrate and separated from the ground plane using foam. and extended path for surface current. Frequency Reduction: The convoluting nature of the patch as shown in figures above suggests that the surface current that flows over it follows a greatly meandered path. as compared to its path on a rectangular patch with similar footprint area. 4.Peano Line Antennas • Insertion of slots: Inserting suitable slots in the radiating patch is also an important technique for reducing the dimensions of the patch antenna [8-9].3. 4. Due to the extended path over which the surface current flows.7 Microstrip Peano line antennas Fig.2 shows the geometry of a microstrip Peano line antenna.2 Geometry of Microstrip Peano line antenna [13] © IEEE 2007 Fig. Fig. This technique is used in the thesis. 4. extra inductive effect comes into picture. and the space-filling curve that is implemented in microstrip form is the 2nd order Peano curve. 4. The slots introduce parasitic capacitance. then it greatly helps in frequency reduction. If the patch of the microstrip antenna traces a space-filling curve without increasing the original physical footprint area.3 A fabricated 2nd order Peano line antenna [13] © IEEE 2007 The patch of the microstrip antenna traces a second order Peano curve. while the photograph of a fabricated Peano line antenna is shown in Fig.

This fact is verified in Chapter 6 by the simulation results. following the relation: fr = 1 2π LC (4. the effective inductance L and capacitance C of the antenna equivalent circuit can be represented as a series resonant circuit.Peano Line Antennas adjacent lines facilitates electromagnetic coupling between them. it is noteworthy that the fringing fields from both the open ends contribute very little to the effective radiation. and thus provides very small radiating edges for the fringing fields to occur.3) provided. Thus when the patch follows a 2nd order Peano curve. the electrons or the charges encounter sharp discontinuities. leaving only the horizontal components as effective radiating slots separated by a distance of half the effective wavelength. the width of the microstrip line forming a Peano line antenna is very small. In this context. Radiation Mechanism: The radiation from the antenna mainly occurs due to the discontinuities inherent within the structure itself. resulting in maximum radiation in the broadside direction. which accounts for extra capacitive effect as well. radiation occurs from the discontinuities and not due to the fringing fields. and therefore the design rule is not followed. The vertical components of those fields cancel out each other. it has been mentioned that since the path length of the curve as occupied by a 2nd order Peano curve is greater than that covered by a 2nd order Hilbert curve. This is the main reason for these antennas not complying with the conventional design rule of l = λeff/2. But in case of Peano line antennas. It is because. where l is the length of the antenna. This causes radiation to occur from the discontinuities. As the inductive and capacitive effects increases. and λeff the effective wavelength. the surface current on it traverses more path length than that caused due to a 2nd 21 . When the length of the patch follows the above rule. where the resonance occurs for an antenna length equal to half the effective wavelength. where they are accelerated or decelerated. the fringing fields from both the open ends is exactly oppositely directed. the resonant frequency (fr) of the antenna falls. therefore Peano curves provide better compression properties. The radiation pattern of this antenna is depicted in Chapter 7. Comparison with Hilbert antennas: In Chapter 3. The above rule is valid only for microstrip rectangular patch antenna. As the current over the antenna traverses a convoluted path.

Peano Line Antennas order Hilbert curve within the same footprint area. than in (a) (b) meander line antennas. Peano line antenna can provide more radiation gain than the corresponding meander line antenna of same antenna length and same footprint area. However. the currents in the parallel adjacently placed line segments being oppositely directed helps in cancellation of radiation due to them at the far field. there are more numbers of discontinuities in Peano line antennas. Both the antennas cover similar path lengths. However. Another important point in this regard is that in case of meander line antennas. which helps in storage of a lot of energy in the near field of the antenna. 4. [11] reported such an observation. by virtue of its structure. Comparison with Meander line antennas: As far as miniaturization is concerned. The 3-dimensional figures are not shown for simplicity. Quality factor.4.4 (a) Microstrip meander line antenna (b) Microstrip Peano line antenna constructively. and microstrip Peano line antenna has been shown in Fig. As evident from the figures. 4. microstrip Peano line antennas provide better miniaturization and more frequency reduction properties than a similar Hilbert antenna. and occupy equal footprint area. both the antennas almost provide similar miniaturization characteristics. Bandwidth & Radiation efficiency: • Due to closely placed line segments. complete cancellation of radiation due to parallel adjacent line segments is impossible in case of Peano line antennas. The structure of only the patches for microstrip meander line antenna. there is a huge amount electromagnetic coupling between them. while the smaller line segments comprises of the effective radiating elements. This enhances the quality factor (Q) of the antenna 22 . Therefore. The work of Fukusako et al. a comparison may be drawn considering the structure of the two. So if the radiation from the discontinuities add up Fig.

which results in reduction of antenna radiation efficiency. Waterhouse. References [1] S.Peano Line Antennas • As bandwidth is inversely proportional to the quality factor. Kosiavas.794-797. “Fundamental Limitations of Small Antennas”. Wong and K. • Application of space-filling curves. L. Microwave Opt. therefore microstrip Peano line antenna suffers from lower bandwidth. effectively increases the antenna path length. Nov 1997 23 [2] [3] [4] [5] [6] [7] [8] [9] . Bhartia. as compared to meander line antennas. Fayyaz. I. 6.1479-1484. “A novel dual-band patch antenna for GSM band”. Safavi-Naeini.A. This enhances the loss resistance of the antenna. pp. A. pp. Technol. As described in the chapter. Wheeler. 156–159. No. 46. Ittipiboon. ‘‘The C-Patch: A Small Microstrip Element’’. December 1947 R. Vol. Electronics Letters. and in that context the basic principles of Peano line antennas has been discussed. The focus has been put mainly on Peano line antennas for miniaturization. This is the case with all antennas based on space-filling curves. IEEE International Symposium on Antennas and Propagation. Vol. 253–254. Peano line antennas are advantageous in terms of frequency reduction as compared to Hilbert antennas. 1994 G. Lett. J. However. Microstrip Antenna Design Handbook. Sanad. Electron. Sauvan. hence increases the conductor loss of the antenna. “Effect of the Shorting Posts on Short Circuit Microstrip Antennas”. pp. 1998 R.P. Vol. pp. 33. 1989 K.8 Summary This chapter explained the basic principles of miniaturization.1916–1917. Arctec House Inc. with an emphasis on electrically small antennas. “A Discussion on the Properties of Electrically Small Self-Resonant Wire Antennas”. Antennas and Propagation for Wireless Communications.. A. Dey and R. 2001 N. Proceedings IRE. Lett. “Small microstrip patch antenna”. they suffer from drawbacks. 12–14. pp. Shin. and in terms of radiation gain. E. 13. 2. Boisset. 1996 M. 4. “Compact microstrip patch antenna”. IEEE Antennas and Propagation Magazine. Yang. 1995 S. “Small dual-frequency microstrip antenna with cross slot”. April 13. Papiernik. P. Best. Lett. S. which is common to all antennas based on space-filling curves. 604–605. P. Electron. 25. 31. Mittra. Sept. M. Vol.R. December 2004 H. Garg. Bahl.

“Design and comparative study on planar small antennas using meander and peano line structure”.5. “Miniaturization of top-loaded monopole antennas using Peano-curves”.581-584. K.Peano Line Antennas [10] [11] [12] [13] H. “A Down-Sized Printed Hilbert Antenna for UHF Band”. 2003 J. T. Radio and Wireless Symposium. Mc Vay and A. S.Safavi-Naeini. 253-256. May 1999 T. No. Vol. pp. 2007 X. 2. IEEE Antennas and Propagation Society International Symposium. 2007 24 . Wang and M. pp. pp. Vol. Lancaster. “Aperture-Coupled Thin-Film Superconducting Meander Antennas”. Chen.Y. Hoorfar. Iwata.47. IEEE Transactions on Antennas and Propagation. Terada. Liu. IEEE. 2451-2454.J. International Symposium on Antennas and Propagation Society. Y. Fukusako.

Chapter 5 A Brief Review of Peano Line Antennas .

In the very next year. it calls for more and more studies on the other space filling curve antennas. The electrical size of an antenna has been an important field for research work. fractal geometry can also be applied to microstrip configuration. the most popularly used antenna for these purposes are the meander line antennas [6-7]. were thus studied in greater detail. Also. In addition. The advancements in the field of Peano line antennas is described next. popularly known as Chu limit [2]. when H. while the review of other space filling curve antennas are beyond the scope of this thesis.Wheeler published a paper on small antennas [1]. or in microstrip form. The wire may follow different fractal shapes like Koch. to result in miniaturization. In addition. The chronological advancements in the development of Peano line antennas have been elucidated in the following sections. [5] to reduce the resonant frequency of the antenna. L. the concept of fractals is also introduced in the field of antennas. 5. Although. As mentioned in Chapter 1. where the edges of the patch are protruded in the form of the fractal shapes. right from 1947.2 Advancements in miniaturization of antennas The methods for miniaturization of antennas have been discussed in the previous chapter. For miniaturization purposes. Hilbert curve antennas [8-9] both in wire configuration and in microstrip form have attracted several researchers due to its simplicity.J. space-filling curves like Hilbert curves. They may be in wire configuration. Hence. Nevertheless. Wunderlich curves. Peano curves etc.A Brief Review of Peano Line Antenna Chapter 5 5. this field is a relatively newer field of study in the branch of antenna engineering. This concept paved the way for the introduction of graph theory in antenna engineering. and the review of those works is not the focus of attention in this thesis. there has been several other researchers in this field over the years [3-4]. a lot of thought process on miniaturization techniques has taken place. Some of them have been discussed in Chapter 4. 25 . Chu in his journal formulated a fundamental limit. Moore curves. Minkowski etc. the others are a bit less explored. Therefore.A.1 A Brief Review of Peano Line Antenna Outline of the chapter This chapter briefly highlights the research works that has been carried out in the field of Peano line antennas over the years.

investigated the effectiveness of space-filling curves as small antennas [10]. though they were suitable for reaching higher miniaturization ratios than conventional quarter-wave monopoles. The input impedance and current distribution of all the antennas 26 . it was possible to locate an off-centre feed. yet the standard first Peano curve variant was not yet implemented. Other than frequency reduction. Gonzalez-Arbesu et al. This study mainly focused on the feed point effects. X. yet the impedance bandwidth was quiet poor as compared to similar order Hilbert antenna. yet it was a late introduction in the field of antenna engineering.Chen et al. Lastly. However. It was in the year 2003 that Jose M. Though the size reduction was substantial. This work was a generalized one on a few space-filling printed antennas. the main advantage of this antenna compared to Hilbert antenna was lower cross polarization level. The importance of space-filling geometries. as optimal or efficient curves for small antenna design was assessed in this work using bidimensional wire monopoles. which included three different variants of Peano monopoles in first two iterations. The study revealed that when the feed point was placed at the point of symmetry. patterned after Peano space-filling curve. current distribution pattern.A Brief Review of Peano Line Antenna 5. J. impedance bandwidth and other radiation characteristics. The magnitude of the current distribution was reported to be almost symmetric. cross polarization. The important conclusion was. the radiation efficiency was reported to be decreasing with increasing iteration order. Several variants of space-filling curves were taken in the form of monopole wires in his comparative study. the Peano antenna being one of them. In 2004. the real part of input impedance was very low. In the same year. where there was a perfect 50-ohm or 75-ohm input impedance match. Though there were a few papers on broadband arrays using Peano-Gosper curve. [11] investigated the characteristics of a single antenna made of thin wire. they store a lot of energy in the near field of the antenna and have higher ohmic losses.3 Review of Peano line antennas Although Giuseppe Peano discovered the Peano curve a long time earlier (1890). resulting in high quality factor and lower radiation efficiencies. reported the first work on printed Peano line antenna [12].Zhu et al. The radiation pattern of the antenna at its fundamental resonant frequency in the planes of Φ = 0° and Φ = 90° almost resembled the patterns of a linear dipole. while the phase varied very slowly along the wire length. The same team had earlier worked on similar investigation for Hilbert curve wire antennas.

The magnitude and phase of the current distribution along the same antenna is depicted in Fig. The resonant frequencies were obtained by searching for the roots of the imaginary part of input impedance. Among the antennas modeled. 5. (c) (a) (b) Fig.2 (a) Geometry of the antenna (b) VSWR characteristics (c) Gain patterns [13] © IEEE 2006 27 . 5. (a) (b) Fig. is shown in Fig. The feed structure was modeled as a delta-gap voltage source.2 shows the schematic diagram of the antenna. 5. the schematic structure of the patch of Peano line antenna.1b.1a. 5.1 (a) Printed 2nd order Peano antenna (b) Current distribution along the antenna [12] © IEEE 2004 J. and utilized probe feed and shorting posts so that it could be matched to a 50-ohm source over a wide bandwidth. Fig. VSWR characteristics and radiation pattern. Hoorfar in their paper [13] in 2006 investigated the characteristics of utilizing planer metallic path in the form of Peano curve and Hilbert curve to provide top loading properties to electrically short monopole antenna elements. 5. Mc Vay and A. This antenna focused on planer space-filling curve configuration with a much lower profile. which was taken for analysis.A Brief Review of Peano Line Antenna were calculated using Method of Moment (MOM) with Mixed Potential Integral Equation (MPIE) formulation.

5. gains of 15. 4. over same direction. This was the first instance of a broadband.4 (a) VSWR characteristics (b) Radiation pattern [15] ©IEEE 2007 28 . Fig. In 2007.2 and Fig.A Brief Review of Peano Line Antenna The role of shorting posts was to reconfigure the structure into radiating like a matched toploaded monopole with vertical polarization. 5.5dB were obtained. this time the work being based on microstrip configuration with dielectric substrates. [14] proposed a high directivity antenna using Peano space-filling curve. The antenna was proposed in microstrip configuration.4. unlike their previous work in 2006. high gain Peano line antenna. the directivity of the antenna increased substantially. The VSWR characteristics and the radiation pattern for the Peano line antenna are shown in Fig. with three different designs. The antenna used six 180°-phase shifters to orient the current distribution throughout the antenna.6dBi was obtained using this design.Mc Vay and A. As all the line segments had current distribution oriented over same direction.3. J. 5.El-Khouly et al.2dB and 17. E. The proposed design is shown in Fig.3 Proposed antenna design [14] ©IEEE 2007 (a) (b) Fig. 17. but at the cost of greater size dimensions. A gain of about 4. In the three designs. In the same year. Hoorfar again published a paper on Peano line antenna and Hilbert curve antenna [15].7dB. 5. 4. The Peano antenna geometry and the fabricated photograph of the antenna are shown in Fig.3 respectively in chapter 4.

5 (a) Antenna structure (b) VSWR characteristics [16] ©IEEE 2007 5. Normal second order Peano curve was not used. Japan [7]. by a team of Fukusako et al.9dBi was reported. but in a modified form. 5. this branch is quiet open for the researchers willing to make a contribution in this field.A Brief Review of Peano Line Antenna However. and the polarization was investigated to be elliptical. for RFID applications. The most interesting aspect in this design was the application of a second order Peano curve. The work reported a design of meander line as well (a) (b) as Peano line antenna for a centre frequency of 1GHz.Terada et al. instead. an extended version to the second order curve was implemented. The details of this antenna are considered in Chapter 6. Fig. [16] from Kumamoto University. very few papers have been reported in the literature. This enhanced the miniaturization of the antenna largely. a brief historical background of Peano line antennas has been presented.5. Both from design as well as analysis point of view. A similar design was proposed by a team of T. Japan in the year 2008. the first RFID design specific work in the field of microstrip Peano line antenna was reported in the year 2007. The antenna structure and its measured VSWR characteristics are shown in Fig. References 29 . from Kumamoto University. where it is used as the design prototype for a particular design problem. which occupied a footprint area of as small as λ/50 X λ/100. which are devoted to Peano line antennas. Although Hilbert curve antenna has been explored by many researchers.4 Summary In this chapter. An antenna gain of -7. The team fabricated and measured the impedance and radiation characteristics of a small and low profile printed antenna using Peano line. 5.

Y. May 1999 T.2. December 1947 L.3. “Radiation characteristics of small and low profile print antenna using peano line”. Terada. IEEE Transactions on Antennas and Propagation. Vol. 2006 E. pp. 4. “A Down-Sized Printed Hilbert Antenna for UHF Band”. Proceedings IRE. 2451-2454.581-584. Vol.R-Samii. Iwata. 3425-3428. “Fundamental limitations in Antennas”.1-4. “Miniaturization of top-loaded monopole antennas using Peano-curves”. Blanch. S. 4.P. IEEE Antennas Wireless Propagat. Liu.A Brief Review of Peano Line Antenna [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] H.3723-3726. Hoorfar. No. Terada.A. Ide. Chen.. Gozalez-Arbesu. 1948 H. Vol. 19. Vol. Safavi-Naeini. Asia Pacific Microwave Conference. 2003 J. “Fundamental Limitations of Small Antennas”. Vol. 1. Feb. Cross-Polarization.. “High Directivity Antenna Using a Modified Peano Space-Filling Curve”.A. Chu. Wheeler.J. 1975 R. T. “Design and comparative study on planar small antennas using meander and peano line structure”. IEEE. Dec. Chen.Safavi-Naeini.Y. IEEE International Symposium on Antennas and Propagation Society. pp.1981 J. H.462-469. IEEE Antennas and Wireless Propagation Letters. 2003 J. IEEE Antennas and Wireless Propagation Letters. pp. Romeu. “Bandwidth. 2008 30 . Vol. K. 147–150. Iwata. S. Zhu and A. IEEE Antennas and Wireless Propagation Letters. Lancaster. J. Fukusako.47. vol. Wheeler. “Fractal Antennas: A Novel Antenna Miniaturization Technique. and Feed-Point Characteristics of Matched Hilbert Antennas”. vol. Hoorfar. T.2. 44. Wang and M. Hoorfar. No. February 2002 H.5. “Small Antennas”. pp. 2004 J. Phys. and Applications”. IEEE Antennas and Propagation Society International Symposium.El-Khouly. Fukusako. 2004 X. Vol. vol. 1163–1175.A. IEEE Antennas and Propagation Society International Symposium. IEEE Antenna’s and Propagation Magazine.1479-1484. Appl. Liu. Gianvittorio. Y. Lett. and J. “Are space-filling curves efficient small antennas?”. K. S. pp. “Printed Plane-Filling Fractal Antennas for UHF Band”. pp. Zhu and A. AP-23. pp. pp. Vol. “Peano Antennas”. McVay and A. 2007 J. 2.6. IEEE Transactions on Antennas and Propagation. pp. 253-256. International Symposium on Antennas and Propagation Society. “Physical limitations on omni-directional antennas.” J. Hoorfar. No. Proceedings IEEE. Hansen. 2. K. No. pp.C. 2003 J. “A Miniaturized Planar Space-filling Curve Antenna with Wideband Monopole-like Radiation Characteristics”. 2007 X. Mc Vay and A. M. Ghali.69. “Aperture-Coupled Thin-Film Superconducting Meander Antennas”. Radio and Wireless Symposium. S. Khamis. 2007 T. Y.

Chapter 6 RFID Tag Antenna Design Using Peano Curves .

957 – 27. Moreover. Moreover. For biomedical purposes. Peano line antennas in microstrip configuration find immense applications in RFID chips. Hence.283 MHz is mainly used for special application such as biomedical applications. human body cannot be irradiated with very high power signals. the distance required for signal transmission is very less. the RFID bands are discussed briefly. The RFID band from 26. the relevant theory and the concepts regarding Peano line antennas are discussed in detail.1 RFID Tag Antenna Design Using Peano Curves Outline of the chapter In the previous chapters. a design problem relating to the design of microstrip Peano line antenna for 26 MHz application is considered. a relatively lower antenna gain suffices our cause. design for 433 MHz RFID band is also considered. typically not exceeding 20 mm X 20 mm area dimensions. 6. This chapter mainly emphasizes on the practical workable antenna design and the difficulties that need to be addressed for future implementation. As discussed earlier. In addition.2 RFID frequency bands In Chapter 2. The design needs to be incorporated on an RFID chip. For this reason. 31 .3 Design Goal Our design goal is to design a microstrip Peano line antenna. the focus is shifted towards a particular design problem using microstrip Peano line antenna. 6. to be used for bio-medical applications meeting all specifications. In this chapter. the difficulties faced in different design approaches are stated and a possible modification is highlighted.RFID Tag Antenna Design Using Peano Curves Chapter 6 6. The next higher RFID band is the less frequently used 433 MHz band. corresponding to a 26 MHz frequency of operation. Therefore. the gain of those antennas need not be very high. Design of such a chip antenna is taken into consideration and the necessary design steps are discussed in detail.

But. as long as the dimensions of the antenna are not exceeded.RFID Tag Antenna Design Using Peano Curves 6. the Peano curve. in our case. Hence. Thus. some alternative design method must be thought of. The width of the microstrip line ‘w’ is 0. while the antenna footprint area is 3. it is wiser to stick to the more easily realizable 2nd order Peano curve. 32 . But to realize the 3rd order Peano curve in reality is a tough task. is to fill the space over the substrate with a patch in the most efficient manner. The design in [1] has a Peano line antenna designed for a central frequency of 1 GHz. 6. The substrate used is a polymide substrate with a relative permittivity (εr) of 3.5 Design prototype The antenna design as shown by Fukusako et al. yet keeping its dimensions small. 6. the work of Fukusako et al [1] can be taken as the basic guideline or prototype in this regard.2.2 mm X 6. it results in an addition of inductive effect and also added capacitive effect due to mutual coupling between adjacent line segments. In this way. [2].1 mm. we reduce the resonant frequency of the antenna to a great extent. Extending the length of the microstrip line by following a 2nd order Peano curves’s repetitive pattern. in [1] is shown in Fig. As we are constrained in size. The alternative method is to follow the repetitive pattern of a 2nd order Peano curve and to go on extending it keeping the symmetry of the curve intact. The height of the substrate ‘h’ is 50 μm as shown in Fig. Hence.4 Design Procedure The basic idea behind the design of such an antenna at very low frequency. 6.5.1 mm. especially as the fabrication difficulty creeps in. as the surface current traverses a path of less distance over a microstrip line following a 2nd order Peano curve than that following a 3rd order Peano curve.1. The RFID chip area is 20 mm X 20 mm. The simulation is performed using Method of Moment based IE3D from Zeland Software™ Inc. the best alternative is to incorporate a 3rd order Peano curve and to try out the design. This can be achieved if the patch or the microstrip line traces the shape of a spacefilling curve. we effectively increase the path over which the surface current traverses.

6. for εr = 3. Rather. which is an extension to the 2nd order Peano curve. (b) Detailed View From Fig. we get εeff = 2. The above design of Fig. This corresponds to a free space wavelength (λ0) of 294.1 can be used as the basic guideline or prototype to our design.RFID Tag Antenna Design Using Peano Curves (a) (b) Fig.1) Thus.5.1 (a) Antenna Structure [1]. 6. This section forms Fig.02 GHz. 6.722 33 .1 it is clear that the patch or the microstrip line does not trace a normal 2nd order Peano curve.75 mm. to fill the space above the substrate in the most efficient manner The total path length ( l ) that the strip of the antenna traverses is 90. The resonance of the antenna as inferred from the simulated results in [1] occurs at 1. truncated after certain distance. The effective permittivity (εeff) can be found out from the expression [3] ε eff = εr +1 εr −1 2 + 2 × 1 1 + 12 h w (6. a repetitive pattern is followed. 6.2 Height of the antenna substrate from the CPW feed end the radiator.1 mm.

If we desire to find out. Frequency (b) Input impedance vs. [2]. 6. the effective wavelength in this case is 178.RFID Tag Antenna Design Using Peano Curves So the effective wavelength of the signal that is propagating through the antenna can be obtained from the expression λeff = λ0 εeff (6. On the other hand.509. 34 .2 mm. frequency S11 and the versus impedance frequency curves for the antenna described in Fig.3 (a) S11 in dB vs. Fig. 6.6 Design approaches This section highlights several design approaches and relevant explanations. 6. (6. the factor by which the effective wavelength is related to the antenna path length ( l ).3) 6. we can do so by finding the factor ‘x’ from the equation x λeff = l For this structure. it is essential to go through the impedance characteristics of the design guideline on which the design is based.2) Thus. thus the antenna resonates almost for a length l = λeff /2.4 shows the same characteristics over 0-20 GHz frequency range for the same antenna. Fig. Study of the design guideline/prototype: Before proceeding further with the design. the design being based on the guideline stated in the last section. 6. x = 0.1 over a frequency range confined around the first resonant frequency.3 shows the (a) (b) Fig. The simulation is performed using IE3D from Zeland Software™ Inc.

6. and as with all transmission lines.4 (a) S11 vs. In the subsequent approaches. the method to reduce the resonant frequency of the antenna is to go on increasing the path length l that the microstrip line of the antenna traces. Also. 6. The S11 matching is not as good as desired. in each step of antenna length extension. 6.85 mm path length is traced. The above phenomena can be attributed to the distributed nature of the antenna. 19. it is obvious that. and the symmetric periodic repetition of the 2nd order Peano curve is followed. increases by 1. frequency over 0-20 GHz frequency range (b) Discussion: From the above curves. frequency (b) Input impedance vs. the outer dimension along the path of the curve.1 Approach 1: Increase of path length of the antenna As discussed earlier. is basically a transmission line. we consider a certain yardstick such that. The microstrip line.2 mm. the basic prototype of [1] is adhered to. the distributed effects cause the impedance characteristics to repeat itself with frequency. which forms the antenna. thus exciting higher order modes at higher frequencies. This process is carried out till the dimensional constraint along the path 35 . but the impedance characteristic reveals that the input resistance and reactance repeats itself. Maintaining symmetry of repetition. as the frequency range is increased the pattern repeats itself.RFID Tag Antenna Design Using Peano Curves (a) Fig.

3 mm. Hence.5. 6. εr = 3. the new antenna footprint dimension is 3. with an S11 match of -15. 6.807 MHz. 36 . using (6.6.6 S11 characteristics for antenna in Fig.e.2 mm occurs along one dimension. while the S11 characteristic is shown in Fig. Thus. since an increment of 1.6 mm. Extensions of path length a) Extension 1: Incrementing a path length of 19.2036 dB.03 mm. Fig. 6. 6.5 The first resonance of the antenna occurs at 837. The total antenna structure along-with the feed is conductor-backed.5.85 mm to the previous design. The substrate is kept the same.2 mm X 7. The corresponding free space wavelength (λ0) for this frequency is 358. The simulation is done using IE3D software.1. The signal line of the CPW is provided with a positive voltage level.5 Antenna structure For extension 1 Fig. The structure is shown in Fig. 6. 6. with dimensions as given in Fig. i.RFID Tag Antenna Design Using Peano Curves length is exceeded. the antenna is simulated. Other dimensions such as w and h are also kept similar to that of the antenna described in the design guideline Feed: The antenna is fed by a 50-ohm coplanar waveguide.2) we can evaluate effective wavelength (λeff) which comes out to be 217. the total path length l now becomes 110. Therefore. and a negative polarity to the two ground planes on the side. taking advanced extension as the deembedding scheme for port definition.07 mm.

807 MHz. Frequency for antenna in extension 2 37 .7 (a) S11 vs. b) Extension 2: Antenna footprint area: 3. keeping the basic parameters of the antenna unaltered.e.5 mm Path length l = 110.5096. λeff = 252. In this case. with S11 = -15. λ0 = 417. Fig. fr = 718. As evident from the data above. 6. we require many such path length extensions.9 mm. using (6. with S11 = -9.45 mm.985 MHz. (a) Factor x = 0.2 mm X 8.25 mm. 6. thus providing good matching characteristics.RFID Tag Antenna Design Using Peano Curves Finally. The second resonance occurs at 1. Frequency (b) Input impedance vs. but not the fundamental mode. (b) Fig. therefore in order to achieve our design goal. the first resonant frequency does not provide a substantial S11 match.18319 dB Free space wavelength.7 depicts the S11 characteristics and the impedance characteristics of the antenna. The above process is repeated in all the next extensions. i. Effective wavelength. we can therefore use the higher order mode for operation.6 mm + 19.85 mm = 130.3) we can find out the factor x as 0.1831 dB. As the first resonance of the antenna occurs at 837.5158. First resonant frequency. it does not adhere to the 2:1 VSWR acceptable limits.45303 GHz.

7 mm Path length l = 130.85 mm = 150. λ0 = 474. 6. (a) (b) Fig. Again.5226.85262 dB. Free space wavelength. λeff = 287. and not at 632. with S11 = -5. fr = 632. It shows that the first resonance actually occurs around 1.5 dB. Factor x = 0.242 MHz as mentioned in the above data.55 mm. with S11 of -33.RFID Tag Antenna Design Using Peano Curves c) Extension 3: Antenna footprint area: 3. 6. and hence can be excited for operation. Fig. The second resonance occurs at 1.8 (a) S11 vs.8 (b) highlights a very important aspect of the design. Effective wavelength.27 GHz. This higher order mode possesses good VSWR characteristics.2 mm X 9. 38 .45 mm + 19.8 shows the S11 characteristics and the impedance characteristics of the antenna. First resonant frequency. we have a poor match at the first resonant frequency.3 mm.27 GHz.6 mm. Frequency (b) Input impedance vs. 6. Frequency for antenna in extension 3 Discussion: Fig.242 MHz.

even though the first resonant frequency. from now onwards.4 e) Extension 5: Antenna footprint area: 3. First resonant frequency. If the stray capacitances play a predominant role.RFID Tag Antenna Design Using Peano Curves This anomaly can be clarified by the following logic.10 S11 vs.6 MHz S11 at first resonance = -4. It is only due to some stray capacitance.1 mm Path length l = 170.9 S11 vs.29 mm. we shall only focus on the first dip of the S11 characteristics to find out the first resonant frequency.27 GHz.15 mm. First resonant frequency. since the reactance curve does not cross the zero reactance line near it.e.15 mm + 19. or the first zero reactance crossing frequency is 1.2 mm X 10.3 mm + 19. frequency for antenna of ext. Therefore.631 mm. i.5 39 .242 MHz. 6. which adds some negative reactance on the curve near its original first resonance.527 Fig. Factor x = 0. The periodicity and the symmetry of the impedance characteristic pattern suggests that. d) Extension 4: Antenna footprint area: 3. 6. the impedance pattern cannot provide any information about the first resonance. The cause of this stray capacitive effect shall be investigated later. λeff = 322.32 MHz Fig. Effective wavelength.85 mm = 190 mm.85 mm = 170.2 mm X 12.9 mm Path length l = 150.1565 dB Free space wavelength. and causes the curve to shift away from the zero reactance crossing line. λ0 = 532. fr = 563. as we increase the path length of the antenna in the subsequent extensions. at 632. frequency for antenna of ext. yet the original first resonance of the antenna should have occurred at a much lower frequency. fr = 508.

Effective wavelength. Factor x = 0. λeff = 357.8 mm.848 MHz S11 at first resonance = -2.11 S11 vs.6 g) Extension 7: Antenna footprint area: 3. λ0 = 705. Factor x = 0. frequency for antenna of ext. First resonant frequency. λeff = 427.85 mm = 229. Fig.RFID Tag Antenna Design Using Peano Curves S11 at first resonance = -3.12 S11 vs.1 mm.07537 dB Free space wavelength. f) Extension 6: Antenna footprint area: 3. λ0 = 590.5 mm Path length l = 209.34511 dB Free space wavelength.8 mm. Fig. fr = 462. λeff = 392.3 mm Path length l = 190 mm + 19.531. 6.8317 dB Free space wavelength.179 mm.7 40 .79 mm.5369.7 mm. Effective wavelength. First resonant frequency. frequency for antenna of ext.2 mm X 13. fr = 425.2 mm X 14. Effective wavelength. Factor x = 0.534. 6. λ0 = 648.85 mm.85 mm + 19.719 mm.85 mm = 209.041 MHz S11 at first resonance = -1.

55 mm + 19.85 mm = 289.1 mm Path length l = 269. frequency for antenna of ext. λ0 = 763. λeff = 532. First resonant frequency.43 mm.539. Factor x = 0. Fig.541. Effective wavelength. frequency for antenna of ext. First resonant frequency.2 mm X 16.RFID Tag Antenna Design Using Peano Curves h) Extension 8: Antenna footprint area: 3. Factor x = 0.7 mm Path length l = 229.2 mm X 15. Effective wavelength. fr = 365.8 i) Extension 9: Antenna footprint area: 3.55 mm.15 S11 vs.18201 dB Free space wavelength. λ0 = 821. 6. 6.72 mm.7 mm. 6.971126 dB Free space wavelength. λeff = 462.25 mm. Fig.7 mm. Fig.9 mm Path length l = 249. First resonant frequency. λeff = 497.4 mm. frequency for antenna of ext. fr = 341.7 mm + 19.14 S11 vs.303MHz S11 at first resonance = -1.2 mm X 18.13 S11 vs.408 MHz S11 at first resonance = -0. 10 41 . fr = 392.85 mm = 249.85 mm = 269.45881 dB Free space wavelength.2 mm. λ0 = 878.4 mm + 19.5 mm.9 j) Extension 10: Antenna footprint area: 3. Effective wavelength.962 MHz S11 at first resonance = -1.

We know from (6.1 mm. The S11 input impedance matching characteristic for the antenna at the first resonant frequency is very poor. λeff = 567.85 mm = 309.16 S11 vs. 6. The first resonant frequency is still deviant from our design goal. Hence. Fig. using alumina as the dielectric substrate.25 mm + 19.2 mm X 19. k) Extension 11: Antenna footprint area: 3. frequency for antenna of ext.6. λeff = λ0 εeff 42 . 6. In this approach. the issue of impedance matching must be taken care of. First resonant frequency. fr = 320.131 MHz S11 at first resonance = -0. Alumina has a dielectric constant of 9.544. Effective wavelength. 2. we cannot go on increasing the antenna path length for further extensions.8. Thus. λ0 = 937.808384 dB Free space wavelength.99 mm.3 mm Path length l = 289. another alternative needs to be employed.2). However. instead of polymide.543. 11 Inferences from the extension of antenna path lengths 1. Further reduction of the resonant frequency is required.2 Approach 2: Higher permittivity substrate The earlier approach reveals that the chip size is too small to facilitate a frequency reduction to a frequency stated in our design goal. Factor x = 0.RFID Tag Antenna Design Using Peano Curves Factor x = 0. we simulate the antenna. as the size limit of the RFID chip is fixed and prescribed.1 mm.

5 365.2X13.72367 -1.894 -40.0 209.5 3.0 1068.8. an increase in εeff must be accompanied by an increase in free space wavelength λ0.2X12. there is an increase in εeff.512 0. εr = 9. and then following the length extension procedure of approach 1.9 3.6 280.8 183. the simulation results are performed by considering alumina as the substrate.RFID Tag Antenna Design Using Peano Curves Therefore.106 342.494 3.508 0. the extensions are represented in a tabular form.526 0. It is because.2X14.2 X 6.9954 487.13764 -3.1).7 3.4 513.389 -1.2 513. as ‘loosely bound’ for effective radiation.3 3. For.0 1164.8 257.0 219.1779 -1. In this approach.85 229.44185 -5.524 43 .5 3.519 0.9 110.55 269.098 385.15 190.7 3. the gain of the antenna is bound to reduce.7 0.0 875. we get from (6.524 0.2X10. For the sake of clarity.6 130.503 0.7 972.75 614.66 779.2 X 7.73499 -2. εeff = 7.49 293.1 LENGTH EXTENSIONS FOLLOWING APPROACH 2 Extension number Antenna footprint (mm2) Path First S11 at first resonance (dB) Free space wavelength (λ0) in mm Effective wavelengt h (λeff) in mm Factor x length l resonan (mm) t freq. as we increase the dielectric constant.8655 -7. and the S11 curves are not shown.1 90.5 1365.553 308. Table 1 illustrates the relevant results obtained in this regard.3 3.063 TABLE 6. as we increase εr. So to keep the effective wavelength λeff as constant.13 583.2 X 8.45 150. using a higher dielectric medium aids in frequency reduction.800 440.3 170.1 474. leaving only a few field lines. Disadvantage: The disadvantage in this method is.523 0.36 219.1 329.0 438.2 X 9.) 0 (design guideline) 1 2 3 4 5 6 7 8 9 3.2X16.8 681.2X15.722 -11. using higher permittivity substrate confines most of the electric field lines beneath the antenna strip.81363 -2.5 0.1 3.7 249.6 256. (MHz.516 0.5 1260.6 238.7 401. Thus.

the characteristic impedance of that section is taken as 50 ohms. in most of the cases. find out the impedance at the end of the antenna.25 309. That impedance is treated as the load impedance. As revealed in the earlier sections.2X18.7 1547. it is the edge or end impedance of the antenna) and the characteristic impedance of the section that is connected at its input.863 -0.938 193. 3. 2. Design a quarter-wave transformer of length λeff/4. it follows that it is quite difficult to achieve a resonance at such a low frequency. Using these data.2 0. The basic idea in this approach is highlighted below: 1.1 3. we now turn our attention towards the next approach. As the quarter-wave transformer is connected to a coaxial cable at its input for initial feed.3 Approach 3: Input impedance match using quarter wave transformer Having discussed the concept of length extension. From the above approaches of length extension. 44 . the design needs to achieve the required 2:1 VSWR limit. 2. 6. keeping the dimensions within the prescribed limits. which mainly deals with the concept of impedance matching. we can start with a quarter-wave transformer impedance matching scheme. where λeff is the effective wavelength. the S11 matching characteristics at the first resonant frequency is poor.532 0. The quarter-wave section will have a characteristic impedance which is the geometric mean of the magnitude of the load impedance (in our case.1 207.94 -0. For this purpose.RFID Tag Antenna Design Using Peano Curves Extension number Antenna footprint (mm ) 2 Path length l (mm) First resonan t freq.8 582.2X19.79 1442. Moreover. the characteristic impedance of the transformer is calculated.6.531 Inferences from Approaches 1 & 2: 1. First and foremost.3 289.4 542.) S11 at first resonance (dB) Free space wavelength (λ0) in mm Effective wavelengt h (λeff) in mm Factor x 10 11 3. (MHz. for the design frequency of interest.

68 mm. the length of the quarter-wave transformer.722. which exceeds our dimensional constraints. the difficulty persists.17 mm. the length of the quarter-wave transformer.1 specifies that. The quarter-wave section can now be implemented using. λeff/4 = 1. Therefore. Difficulties and bottlenecks: The main difficulty in this design is the length of the quarterwave transformer. λeff/4 = 1. Effective wavelength. However. The effective wavelength considering polymide substrate is 419. for Alumina substrate: εr = 9. as documented above. λeff using (6.9 mm. 45 .538 m. either a microstrip line or a coplanar waveguide (CPW).8 mm.085 m = 1085 mm.8 εeff = 7. For this reason. For our design frequency of interest.97 mm. A frequency of 433 MHz corresponds to a free space wavelength of 692.063.5 Therefore. both of which exceed our dimensional constraint.e. using (6. we try our design around this frequency. 26 MHz we can calculate the length of the transformer for both the substrates used above. there is a first resonance at 440. while that considering alumina substrate is 260. we alter our design goal to the next higher RFID frequency of 433 MHz.748 m = 1748 mm.2) amounts to 6. λeff = 4. For 26 MHz. the implementation of the quarter-wave transformer on polymide substrate requires a length of 104.RFID Tag Antenna Design Using Peano Curves 4. b) Similarly. Effective wavelength.341 m Hence. Extension number 2 from Table 6. This modification therefore cannot overcome the dimensional difficulties as faced earlier.098 MHz that is quite close to our revisited design goal. λ0 = 11. Design goal revisited: Due to several difficulties faced.993 m Thus. while the length of the same for alumina substrate is 65. we get εeff = 2. i. a) For polymide substrate: εr = 3.1).

6. 6. thereby adding an extra amount of stray capacitive reactance. The structure is made up of twostacked dielectric layer with εr2 > εr1. This method is illustrated in the next approach. On the higher permittivity substrate 2. the quarter-wave feed is placed. and the antenna on the lower permittivity substrate. while the antenna is placed on the lower permittivity substrate 1. Due to this discontinuity. An increase in the dielectric constant brings about a decrease in the effective wavelength at the design frequency of interest.RFID Tag Antenna Design Using Peano Curves Inference from Approach 3: The length of the quarter-wave transformer itself (without antenna) exceeds our prescribed dimensional constraints for an RFID chip by a good margin. thereby reducing the length of the quarter-wave transformer required. but avoiding the gap due to discontinuity as far as practicable. The remedy for this problem is to design the structure on two different dielectric media as stated above.6.5 Approach 5: Use of stacked dielectric layers In this approach. and gluing them together. however small it may be.4 Approach 4: Use of different dielectric materials for feed and antenna The problem stated above can be addressed by taking two different dielectric medium side by side. The antenna Fig. The schematic side view of the design is presented in Fig. thus deviating the antenna further away from resonance. an alternative design to address the above problem is proposed. The antenna being placed on the low permittivity substrate ensures that its gain is not reduced. 6.6. 6.17. a gap arises in the structure at the interface.17 Schematic side view of the design 46 . Drawback of the design: The primary drawback of this design is the discontinuity arising from the dielectric-dielectric interface. so that the quarter-wave transformer feed is placed on the material with a higher dielectric constant. while using higher permittivity substrate for the transformer ascertains a miniaturized length for it.

Fig. it can safely be considered as a lumped connector. i. the ground plane comes in between the two media.RFID Tag Antenna Design Using Peano Curves input end.6.6 Approach 6: Modification of the structure described in approach 5 For modification.098 MHz. Considering this frequency as the design frequency. As we observed from extension number 2 of Table 6.18 Modified schematic side view of the design We take that structure for further analysis. 6. Therefore. As the quarter-wave transformer is implemented using a microstrip line. Fig. have a dielectric medium on one side with a common ground plane. drilling a hole on the ground plane for transformer to antenna connectivity. The modified scheme as depicted is almost similar to the scheme discussed in the previous attempt. Both the transformer and the antenna. there is a resonance at 440. the edge of the antenna. and air on the other side. 6. Henceforth. since we are operating at a very low design frequency. we have a closer look at this design approach. The load is connected to the matching section by a shorting post or via.18 depicts the modified scheme. which happens to be the first resonance of the antenna. forms the load to the transformer. as compared to the effective wavelength.e. 6. Therefore. the shorting post is an elongated one. but it calls for a bit of modification. So to avoid any loss of generality. Also. standard close form relations are available assuming one of the dielectric medium as air. The quarter wave transformer is actually immersed in between two different dielectric media. our design is mainly centered at this frequency. none of which is air. The via or the shorting post is very small in length. we slightly modify the structure in the next approach. suitability in of order to investigate wave the quarter matching 47 . and thus separates them. we just invert the layer denoted by ‘substrate 2’ along with the quarterwave transformer on it. Difficulty: This design can be tried out. We now shift our design goal to the next higher RFID frequency of 433 MHz. due to absence of any distributed effect.1.

Fig. 6. 48 . 2. frequency for antenna of ext.19 shows the S parameters and the input impedance characteristics of the antenna described in extension 2 of Table 6. The first step is to remove the CPW feed from the antenna. Fig. which reveals some important facts. Then. cannot conclusively ascertain it as the first resonant frequency. with the CPW feed removed. This de-embedding scheme serves as a wave-port for the antenna.20 shows the same antenna structure and its impedance characteristics. (a) (b) Fig.2. we assume 440. considering the first dip of the S11 curve. table 1). 6. the edge impedance of the antenna is measured using “Extension for waves” de-embedding scheme in IE3D simulator.RFID Tag Antenna Design Using Peano Curves technique on it. and keep the antenna isolated.1. when the CPW feed is connected (antenna of ext. in order to measure its edge or end impedance. The impedance curves on the other hand. and the S11 and the impedance curves are observed for comparative study. since the reactance curve does not cross the zero-reactance line.098 MHz. Therefore. Both the cases are simulated.19 (a) S11 vs. table1 As seen from the figure above. the first dip of the S11 curve occurs near 440. 6.098 GHz as the first resonant frequency. frequency (b) Input impedance vs.

098 MHz. However. when the CPW feed is removed.20 (b). we find the impedance at the input edge of the antenna at 440. 6. 6. Here. 6. we arrive at a very important conclusion. since the stray capacitive effect is removed. 49 .19 (b) and Fig. we have applied a wave-port for calculating its input edge impedance. Analysis and results: In this section.20 (a). We can thus infer from the observation that. The input resistance Rin =96. the reactance curve near the first resonant frequency rises. From Fig. and resonance occurs. the CPW structure adds the stray capacitance. 6. 6. with CPW feed removed (b) Input impedance characteristics for the same antenna The same antenna. Discussion: If we have a closer look at the curves in Fig. The characteristic curve is shown in Fig. a stray capacitance as discussed earlier.2 of Table 1. we verify the suitability of this design for implementing quarter wave matching technique on it.RFID Tag Antenna Design Using Peano Curves (a) (b) Fig. shifts the input reactance curve away from resonance near its first resonant frequency. with CPW feed removed.3777 ohms.20 (a) Antenna described in ext. looks like that in Fig.20 (b). 6.20 (b). When the CPW is used for feeding the antenna.

Therefore. Standard closed form synthesis formulae for microstrip design are available in the literature [5].79 ohms.11 ohms at 440.5295 ohms. Let us denote it by ZL.e.95π 2 Z0 ε r Case 1: Taking the substrate as polymide ( εr = 3.79 ohms Using (6. d = (6. the magnitude of input impedance is |Zin| = { (Rin)2 + (Xin)2 }1/2 = 137.4).293 − h π πε r  εr  where.4) (6. Therefore. Case 2: Taking the substrate as alumina ( εr = 9. when Z0 < {44 – 2 εr} ohms ) ε −1  0. we get |Z1| = 82.517  w 2 = {(d e − 1) − ln (2d e − 1)} + r  ln (d e − 1) + 0. They are stated below: • For narrow strips ( i. Z1 can be calculated from the relation [4] |Z1| = { |Z0| |Zin| }1/2 Putting the values of |Zin| and |Z0| in (6.098 MHz. This |Zin| acts as the load to the quarter-wave matching section. the quarter-wave matching section has a characteristic impedance of 82.5). Therefore.5) For wide strips ( i. We can implement it using a microstrip line.044 mm. with |Z0| = 50 ohms.9 4 1  ε − 1  π 1  +  r   ε + 1  ln 2 + ε ln π  2 r r   −1 (6.5 ) and Z0 = 82. the value of microstrip width w = 0. Let the characteristic impedance of the matching section be Z1.8 ) and Z0 = 82. H '= • Z 0 2(ε r + 1) 119.RFID Tag Antenna Design Using Peano Curves The input reactance Xin = 97.e.79 ohms 50 .6) 59. when Z0 > {44 – 2 εr} ohms )  w  exp H ' 1 =  8 − 4 exp H '   h   where. The input to the matching section is a standard 50 ohms coaxial cable.79 ohms.

In microstrip configuration.5) and (6. However. the width of the microstrip line becomes very small.013 mm. it is even smaller.21. As discussed above. Therefore. fabrication difficulty comes in.e. The basic concept behind single shunt stub matching is very briefly presented in the next section. Single-stub tuning: This technique uses a single open-circuited or short-circuited length of transmission line (stub). Hence. Conversely. and from the two case studies. if the dielectric constant of the substrate is kept very low. it is difficult to realize a series stub.RFID Tag Antenna Design Using Peano Curves Using the same formula. and micro-fabrication techniques must be employed. it is advantageous as far as the length of the quarter-wave section is concerned. it cannot provide a miniaturized length for the quarter-wave section.6) above. 6. the width of the microstrip w = 0. The antenna itself is placed on an alumina substrate. and as discussed in previous sections. as evident from the formulae (6. So. the more is the dielectric constant of the medium. and fabrication problems appear.013 mm. i. The circuit is shown in Fig. it is as small as 0. the smaller is the microstrip width. For other higher permittivity substrates. if the permittivity of the substrate used is high. impedance match using single shunt stub. the width of the microstrip line required may be within normal fabrication limits. 6.6. and the quarter-wave section can be implemented. the quarter-wave transformer must be placed on a still higher permittivity substrate. 51 . However. Nevertheless. connected in parallel with the feed line at a certain distance from the load.7 Approach 7: Single stub matching We arrive at the last of our design approaches. Difficulties from the design: 1. For alumina. the length of the quarter-wave section at our design frequency of interest becomes very large. the implementation of a quarter-wave transformer for this design using a microstrip line requires the microstrip width to be of the dimensions as stated above. 2.

Analysis [4]: Let the load be denoted as ZL = RL + j XL. and the length of the open or shorted stub. 6. is not shown in the figure. The quarter-wave matching section in the figure is replaced by a single shunt stub section. The transmission feed line as well Fig. as it is placed below the bottom layer.21 is given by 52 . 6. with quarter-wave matching transformer replaced by a stub matching section. The impedance Z.18 (bottom view) as the stub is implemented using microstrip lines.18.22 Implementation of Shunt stub in structure of Fig. The rest of the structure remains the same.22 shows the same schematic structure of Fig.21 Single shunt stub tuning circuit The stub matching technique can be implemented on the same structure shown in Fig. ‘l’. a perfect match Y = 1/Z between the feed-line of characteristic impedance Z0 and the load impedance ZL is accomplished. 6. and connected to the matching section by a shorting post or via. 6. By selecting proper values of ‘d’ and ‘l’.18 in bottom view. the input edge of the antenna acts as the load. The antenna which serves as the load. down a length d of the line from the load as shown in Fig. As in earlier case.RFID Tag Antenna Design Using Peano Curves The task is to find out the distance of the stub ‘d’ from the load end. The characteristic impedance of the microstrip transmission feed line is taken as 50 ohms. Fig. Fig. 6. 6. 6.

λ/2 can be added to give a positive result.7) The admittance at this point is Y = G + j B = 1/Z.RFID Tag Antenna Design Using Peano Curves Z = Z0 ( RL + jX L ) + jZ 0 t .11b) λ If the length given by (6.8a).11a) λ For a short-circuited stub. the input end of the antenna has a resistance of 96. for RL ≠ Z0 (6.5295 ohms. d (which implies t) is chosen so that G = Y0 = 1/Z0. where t = tan βd. ls = 1 Y  tan −1  0  2π B (6.8a) (6. R L (1 + t 2 ) where G = 2 RL + ( X L + Z 0 t ) 2 B= 2 R L − ( Z 0 − X L t )( X L + Z 0 t ) 2 Z 0 [ RL + ( X L + Z 0 t ) 2 ] (6. From (6.11a) or (6. this results in a quadratic equation for t. 53 .11b) is negative. Z 0 + j ( RL + jX L )t (6.10) The length of the stub can be found from the formulae below: For an open-circuit stub. the two principal solutions for d are  1  −1  2π tan t  d   =  λ 1 −1 (π + tan t )  2π    for t ≥ 0 for t < 0 (6. Results: For our design as we found earlier.9) Thus.3777 ohms and a reactance of 97. Solving for t gives t= 2 X L ± R L [( Z 0 − R L ) 2 + X L ] / Z 0 RL − Z 0 .8b) Now. l0 = B −1 tan −1  Y 2π  0     (6.

11a). Using (6. we get lo = -0. Putting B in (6. we add a length of λ/2 to it.34 From (6. The length of the stub. lo = 0. d = 0. d = 0.1587 λ ls = -0.22 λ. Putting these data in (6.8b). B = 0.34. Now we have two cases. Therefore. we get d = 0. where RL ≈ 96 ohms and XL ≈ 97 ohms.3413 λ Putting B in (6.10). lo = 0. one when t = 5. ls = 0.5 λ = 0. -1. This impedance serves as the load impedance ZL for the matching section. the optimum solution for minimum dimension is for t = 5.091 λ 54 . The distance between the load and the stub.125 as two solutions for t. Case 1: t = 5. Therefore. ZL = RL + j XL ≈ 96 + j 97.11b). hence. Z0 = 50 ohms.125.22 λ The length of the stub is least when it is a short-circuited one.9). and another when t = -1.091 λ + 0.34. we evaluate the value of t. we get ls = 0. t = 5. we obtain.365 λ. and Xin ≈ 97 ohms.091 λ Case 2: t = -1.409 λ Optimum solution: From the two cases discussed.34.091 λ => -0.RFID Tag Antenna Design Using Peano Curves For ease of calculation. We consider a 50-ohm microstrip line for transmission feed.125 Following the similar steps as above.1587 λ This length being negative.03098. We get. we take Rin ≈ 96 ohms.

098 MHz. Therefore. We now evaluate the dimensions in two cases.22 λeff = 90.091 λeff = 37.49 mm.16 mm.1 mm.5) Z0 = 50 ohms for microstrip feed line.04 mm Separation distance d = 56. is basically the effective wavelength λeff for this design.42 mm Short-circuited stub length ls = 23. we need yet higher permittivity substrates. 55 . The distance d of the stub from the load end is: d = 0.5). for alumina substrate.8) Similarly. with respect to which all the above dimensions are provided.RFID Tag Antenna Design Using Peano Curves Dimensions: The wavelength λ. one when the substrate is polymide. a) Polymide substrate (εr = 3. λeff = 256. it is not so convenient solution because higher dielectric constant further reduces the width of the microstrip line. and again when the substrate is alumina. for 440. Using (6. let the width required for the same is ‘w’.89 mm Length of the short-circuited stub ls is: ls = 0.59 b) Alumina substrate (εr = 9.098 MHz.1 mm. the width of the microstrip line of 50-ohms characteristic impedance must be 0. which gives rise to many fabrication difficulties. Now. To implement this microstrip line. effective wavelength λeff = 413. for 440. Microstrip width w = 0.34 mm Difficulties from the design: The results in the preceding section clearly suggest that the dimensions for d and l exceed our overall RFID chip dimensions. However. The transmission feed line in both the cases is 50-ohm microstrip line. w = 0. In order to accommodate the stub within the prescribed limits.

has been highlighted. If we look at the factor x. some important inferences regarding microstrip Peano line antennas were illustrated. the resonance does not occur for an antenna length l = λeff/2. The antenna end.7 Discussions Having discussed all the design approaches. References 56 . 6. 2. as calculated from (6.5 as was initially in case of the design prototype. This condition is satisfied throughout our design attempts. reflects a part of the signal fed.signal line is kept very small as compared to the effective wavelength λeff. Note: The dimensions for spacing between the signal line and the ground planes for realizing a 50-ohm CPW feed in some of the above approaches. as our design goal is to design the antenna for a very low RFID frequency. the loss due to those reflections can be neglected for all practical purposes. the value of x goes on changing from the value of 0. However. is found using ‘Line Gauge’ from Zeland Program Manager [2].RFID Tag Antenna Design Using Peano Curves 6. we can infer that. as the length l of the antenna increases. which serves as the load to the CPW feed line. The length of the CPW. The reason is explained in chapter 4. Hence we can conclude.8 Summary In this chapter. Moreover. The practical constraints and limitations for each of the design schemes. They are highlighted below: 1. for Peano line antennas. The CPW feed as mentioned in some of the design approaches is a 50-ohm CPW feed line. This results in standing wave pattern in the CPW signal line. some important points worth mentioning are discussed in this section. the approaches for an antenna design for RFID bands have been discussed in detail. The dimensions can also be calculated using standard CPW formulae. as the signal line is very small with respect to the effective wavelength. which is verified by the results provided in this chapter.3) in the above length extension design approaches.

2005 D. Terada.. “Design and comparative study on planar small antennas using meander and peano line structure”. 1991 57 .. USA.M. International Symposium on Antennas and Propagation Society. Fukusako. T. 3rd Edition.12 C. Foundations for Microstrip Circuit Design. Antenna Theory: Analysis and Design. Edwards. Release 10. K. pp. Balanis.A. John Wiley & Sons. Reprint2008 T. Third Edition. C. 2004. 2007 Zeland Software Inc. John Wiley And Sons. Iwata. 2451-2454. 2nd Edition. Pozar.RFID Tag Antenna Design Using Peano Curves [1] [2] [3] [4] [5] T. John Wiley and Sons Inc. Microwave Engineering.

Chapter 7 Probe Feed Optimization .

The changes in these frequencies with changing feed position are marked. the results obtained from varying probe-feed position are highlighted. first anti-resonant frequency by far1.5 mm X 25. The results are observed. In Fig. 7. The substrate dielectric constant is 4. The strip width is chosen to be 1. The impedance characteristics are shown in Fig. The probe is first fed at the end of the antenna. This process is continued till proper modes are excited. considering infinite ground plane. with inner conductor feeding the Peano curve patch. The diameter of the inner conductor of the probe is 1.3 of height 1 mm. while red indicates input reactance (Xin) versus frequency plot. There are four frequencies.1b.5mm.3 Variation in probe feed position In this section. 7. The frequency range from 0 to 1 GHz is studied. The simulation is performed using IE3D software. The anti-resonant nature is verified by the high input resistance 58 . 7.3 mm.Probe Feed Optimization Chapter 7 7. The probe is located at one end of the antenna as shown. and outer conductor completing the circuit at the ground plane. and the probe position is shifted further along the length of the antenna. 7.5 mm. which are marked. The probe is first fed at one end of the antenna. For simplicity.1a. second resonant frequency by fr2. only a few representations of the feed mechanism are depicted. The anti-resonant points are identified by those frequencies. while second anti-resonance by far2.1b.1 Probe Feed Optimization Outline of the chapter The present chapter describes an investigative study on microstrip Peano line antenna. 7. the blue curve indicates the input resistance (Rin) versus frequency plot. till proper modes are excited. by varying the probe-feed location. The external dimension of the antenna footprint is 25. The changes in the impedance characteristics are highlighted as the probe is fed by moving it inwards along the antenna length.2 Antenna parameters The patch of the antenna traces a standard second order Peano curve. From left to right. the first resonant frequency is denoted by fr1. while the important observations are documented and discussed. • Probe fed at end: The antenna structure is shown in Fig. where there is a sharp transition from high inductive to a high capacitive reactance.

however. (b) (a) Fig. The VSWR characteristics are not shown for this case. in each step. Rin(fr2). Xin(far2-) = 1596 Ω. far1 = 405 MHz. 59 . those values for changing feed positions are mentioned. Similarly. Let the high inductive reactance at the left vicinity of far1 be denoted by Xin(far1-) and the high capacitive reactance at the right vicinity of far1 be denoted by Xin(far1+). This process continues till the optimized probe-feed location is obtained for exciting the modes.1b. far2 = 788. The impedance pattern in all the feed positions remain the same as shown in Fig. Xin(far2+) = -1701 Ω. 7. Rin(far1) and Rin(far2) respectively. the parameters are defined by Xin(far2-) and Xin(far2+).5 mm inwards along the antenna length. Further. as observed from Fig. all the cases are not depicted for simplicity. 7. Let us define a few additional parameters. 7. Therefore.809 Ω. as the prescribed 2:1 VSWR limits are not achieved over the stated frequency range. Let the input impedances at fr1. Rin(far2) = 3047 Ω.1 (a) Antenna structure when probe fed at edge (b) Impedance characteristics The values for all the parameters defined above are stated below: fr1 = 210 MHz. though the important observations are discussed. Xin(far1+) = -1711 Ω. Rin(fr1) = 0. Also. Very few representations are shown.5 MHz.1b. far1 and far2 be denoted by Rin(fr1). for far2.882 Ω. fr2 = 612 MHz. fr2. Rin(fr2) = 0. Xin(far1-) = 1654 Ω. The study is carried out by shifting the probe by 1. Rin(far1) = 3423 Ω. the parameter values changes. let us also define the steep inductive and capacitive reactance just at both sides of the anti-resonant frequencies.Probe Feed Optimization at those frequencies.

0 mm inwards along the antenna from the end: fr1 = 212 MHz. Xin(far1+) = -1716 Ω. Rin(fr1) = 0. Xin(far2-) = 1544 Ω. Rin(far2) = 1369 Ω.0 mm inwards along the antenna from the end: fr1 = 239 MHz. Rin(fr2) = 0. Xin(far1+) = -1719 Ω. Xin(far1-) = 1624 Ω.79 Ω. far2 = 791 MHz. Xin(far2-) = 658 Ω. far1 = 405 MHz. • Probe fed at 30. Rin(fr1) = 0. fr2 = 729 MHz.71 Ω. Xin(far2+) = -1626 Ω. Xin(far1+) = -1682 Ω. Rin(fr2) = 0.0 mm inwards along the antenna from the end: fr1 = 228 MHz.85 Ω.5 mm inwards along the antenna from the end: fr1 = 212 MHz. fr2 = 620 MHz. Rin(fr2) = 0.Probe Feed Optimization • Probe fed at 1. Xin(far2-) = 1164 Ω. • Probe fed at 6. fr2 = 626 MHz. far1 = 405 MHz. Rin(fr1) = 0. Xin(far2+) = -1350 Ω. fr2 = 666 MHz. fr2 = 702 MHz.806 Ω. Xin(far1+) = -1654 Ω. far1 = 405 MHz.0 mm inwards along the antenna from the end: fr1 = 247 MHz. fr2 = 616 MHz. Rin(fr2) = 0. Rin(far1) = 2950 Ω. Rin(far1) = 2757 Ω. far1 = 405 MHz.868 Ω. Xin(far1-) = 1749 Ω. Rin(far2) = 3365Ω. far1 = 406 MHz. Xin(far1+) = -1689 Ω.869 Ω. far2 = 795 MHz. far2 = 791 MHz. Rin(fr1) = 0. • Probe fed at 21. • Probe fed at 15.77 Ω. far1 = 406 MHz. far2 = 802 MHz. Rin(fr1) = 0. Xin(far2+) = -726 Ω. Xin(far1-) = 1445 Ω. Xin(far2+) = -1670 Ω. Xin(far1-) = 1425 Ω.0 mm inwards along the antenna from the end: fr1 = 222 MHz. Xin(far2-) = 1552 Ω. Xin(far1+) = -1429 Ω. Rin(far1) = 3054 Ω.815 Ω. Xin(far2+) = -1096 Ω. Rin(far2) = 2520 Ω. far2 = 799 MHz. 60 . fr2 = 646 MHz.75 Ω. Rin(far2) = 3479 Ω. • Probe fed at 3. Xin(far2-) = 1344 Ω. Rin(fr2) = 0.733 Ω. Xin(far1-) = 1740 Ω. Rin(far1) = 2682 Ω. Rin(far1) = 2944 Ω. far2 = 791 MHz. Rin(far2) = 3149 Ω. Xin(far2+) = -1538 Ω. • Probe fed at 39. far1 = 405 MHz. far2 = 793 MHz.8 Ω. Xin(far1-) = 1726 Ω.79 Ω. Xin(far2-) = 1593 Ω. Rin(fr2) = 0. Rin(fr1) = 0. Rin(far1) = 2921 Ω. Rin(far2) = 2154 Ω.0 mm inwards along the antenna from the end: fr1 = 214 MHz.

Probe Feed Optimization

Rin(fr1) = 0.67 Ω, Rin(fr2) = 0.77 Ω, Rin(far1) = 2591 Ω, Rin(far2) = 737 Ω, Xin(far1-) = 1244 Ω, Xin(far1+) = -1352 Ω, Xin(far2-) = 374 Ω, Xin(far2+) = -427 Ω. • Probe fed at 48.0 mm inwards along the antenna from the end: fr1 = 259 MHz, fr2 = 763 MHz, far1 = 408 MHz, far2 = 806 MHz, Rin(fr1) = 0.65 Ω, Rin(fr2) = 0.79 Ω, Rin(far1) = 2160 Ω, Rin(far2) = 270 Ω, Xin(far1-) = 1130 Ω, Xin(far1+) = -1123 Ω, Xin(far2-) = 132 Ω, Xin(far2+) = -157 Ω. • Probe fed at 49.5 mm inwards along the antenna from the end: fr1 = 262 MHz, fr2 = 769 MHz, far1 = 408 MHz, far2 = 807 MHz, Rin(fr1) = 0.65 Ω, Rin(fr2) = 0.75 Ω, Rin(far1) = 1951 Ω, Rin(far2) = 173 Ω, Xin(far1-) = 1079 Ω, Xin(far1+) = -1100 Ω, Xin(far2-) = 97 Ω, Xin(far2+) = -122 Ω. • Probe fed at 51.0 mm inwards along the antenna from the end: fr1 = 265 MHz, fr2 = 777 MHz, far1 = 408 MHz, far2 = 807 MHz, Rin(fr1) = 0.65 Ω, Rin(fr2) = 0.79 Ω, Rin(far1) = 1939 Ω, Rin(far2) = 148 Ω, Xin(far1-) = 1018 Ω, Xin(far1+) = -1071 Ω, Xin(far2-) = 62 Ω, Xin(far2+) = -84 Ω. • Probe fed at 52.5 mm inwards along the antenna from the end: fr1 = 267 MHz, fr2 = 784 MHz, far1 = 408 MHz, far2 = 809 MHz, Rin(fr1) = 0.65 Ω, Rin(fr2) = 0.79 Ω, Rin(far1) = 1947 Ω, Rin(far2) = 92 Ω, Xin(far1-) = 976 Ω, Xin(far1+) = -1016 Ω, Xin(far2-) = 39 Ω, Xin(far2+) = -51 Ω. S11 = -9.63dB at 809 MHz. • Probe fed at 54.0 mm inwards along the antenna from the end: fr1 = 270 MHz, fr2 = 792 MHz, far1 = 408 MHz, far2 = 809 MHz, Rin(fr1) = 0.63 Ω, Rin(fr2) = 0.86 Ω, Rin(far1) = 1879 Ω, Rin(far2) = 48 Ω, Xin(far1-) = 897 Ω, Xin(far1+) = -969 Ω, Xin(far2-) = 19 Ω, Xin(far2+) = -27 Ω. S11 = -30dB at 808.511 MHz. • Optimised feed position- Probe fed at 54.1 mm inwards along the antenna from the end:
61

Probe Feed Optimization

fr1 = 270 MHz, fr2 = 792 MHz, far1 = 408 MHz, far2 = 809 MHz, Rin(fr1) = 0.63 Ω, Rin(fr2) = 0.83 Ω, Rin(far1) = 1873 Ω, Rin(far2) = 46 Ω, Xin(far1-) = 951Ω, Xin(far1+) = -967 Ω, Xin(far2-) = 19 Ω, Xin(far2+) = -27 Ω. S11 = -36.37dB at 808.75 MHz. Comments: At a distance of 54.1 mm along the length of the antenna going inwards from the end, we get an input impedance match of -36.37 at 808.75 MHz. This mode is the second order mode, and not the fundamental one. The feed location, input reactance curve, input resistance curve, and VSWR characteristics are depicted separately in Fig. 7.2

(a)

(b)

(d) (c) Fig. 7.2 (a) Optimized feed location for exciting 2nd order mode (b) Input reactance versus frequency plot (c) Input resistance versus frequency plot (d) VSWR characteristics (S11 in dB)

Continuing in a similar manner from the present location, the optimized probe-feed location for exciting the fundamental mode is searched for.

62

Probe Feed Optimization

Optimized probe-feed position for fundamental mode excitation – probe fed at 110.7 mm inwards along the antenna from the end: fr1 = 383 MHz, fr2 = 443 MHz, far1 = 413 MHz, far2 = 809 MHz, Rin(fr1) = 0.413 Ω, Rin(fr2) = 0.44 Ω, Rin(far1) = 48 Ω, Rin(far2) = 3072 Ω, Xin(far1-) = 24 Ω, Xin(far1+) = -23 Ω, Xin(far2-) = 1681 Ω, Xin(far2+) = -1600 Ω. S11 = -23.53 at 413.75 MHz. The feed location, input reactance curve, input resistance curve, and VSWR characteristics for excitation of fundamental mode are shown in Fig. 7.3.

(a)

(b)

(d) (c) Fig. 7.3 (a) Optimized feed location for exciting fundamental mode (b) Input reactance versus frequency plot (c) Input resistance versus frequency plot (d) VSWR characteristics (S11 in dB)

63

64 .5 Radiation Pattern for the fundamental mode The radiation pattern of the antenna for the fundamental mode is shown in Fig. the reactance profile depicts that the initially high inductive and capacitive reactance in the vicinity of the corresponding antiresonant frequency. 7. the highest input resistance for that mode is nearer to 50 ohms. while Fig.4 Observations from the optimization process The main observations are listed below: 1. The reason is that the excitation of a mode depends on the field configuration for that mode. 7. There are different locations for exciting different modes. The S11 curves portray a good VSWR match corresponding to the anti-resonant frequencies. On contrary. When any mode is excited and good VSWR is obtained. Fig. The observations suggest that good VSWR match occurs when the anti-resonant modes gets excited. 7.4a shows the elevation pattern in the Φ = 0° plane. 5. feeding a probe at the maxima of a certain mode facilitates that mode excitation. • Using superstrates over the antenna structure. The reason for this is explained in chapter 8. 3. 7. 4. almost tends towards zero.5895dBi. The gain for the antenna can be increased by: • Using High Temperature Superconductor (HTS) as the material for the patch instead of copper. If the probe is fed for the minima of a particular mode.4b shows the same in Φ = 90° plane.4c. The azimuth pattern is shown in Fig. which helps in impedance match with the input feed. They can be considered as almost constant with varying feed position. As the probe position changes. and the location where the probe is fed. 7. The anti-resonant frequencies on the other hand. the resonant frequencies are shifted. that mode cannot be excited. vary very slowly with changing feed position. 2. When good VSWR for a mode is obtained.4.Probe Feed Optimization 7.75 MHz fundamental frequency is -36. The maximum gain of the antenna at 413.

Probe Feed Optimization (b) (a) (c) Fig. 65 .6 Summary The probe feed optimization process for exciting fundamental as well as higher order modes has been highlighted in this chapter.4 (a) Radiation pattern in Φ = 0° elevation plane (b) Radiation pattern in Φ =90° elevation plane (c) Radiation pattern in azimuth plane for θ = 90° 7. 7. The antenna is an anti-resonant antenna.

Chapter 8 Lumped Circuit Model Analysis .

8. Both the antennas are for microstrip configuration only. 8. The modeling procedure of these antennas is discussed in the next section. 8.1 (a) Meander line antenna (b) Peano line antenna microstrip line. At first. Then the results relating to the Peano line antennas are shown. the complete derivation of the lumped circuit model is discussed elaborately. The resonant frequency of the antenna as obtained from the circuit model is validated against the results obtained from antenna simulation response. thereby validating our approach.2 Brief description of the antennas modeled In the next few sections. In this chapter. 66 . Circuit response of these antennas using our model has been compared with the simulated antenna response using Method of Moments based simulator IE3D with good agreement. Till date. This chapter forms the most important part of the thesis. The entire theory behind the modeling is presented in general.1 Lumped Circuit Model Analysis Outline of the chapter After having discussed the relevant design specific concepts of microstrip Peano line antennas in chapters 6 and 7. while the substrate height and the ground planes are not shown for the sake of simplicity. The model is implemented particularly on two types of antennas. no simple analytical model except those using complicated numerical techniques is available for the analysis of these antennas. The figure shows only the path traced by the Fig. Fig. the present chapter focuses on some of the analytical aspects of the antenna. while the model is employed on microstrip meander line and microstrip Peano line antennas in particular.Lumped Circuit Model Analysis Chapter 8 8. and this model also is entirely based on microstrip configuration. a simple analytical model is presented to calculate the resonant frequency of the antenna employing a lumped equivalent circuit model.1 shows the top view of the two antennas on which the model is (a) (b) employed. the results pertaining to the meander line antennas are shown due to simplicity of the antenna structure.

8. while the right-angled bends are denoted by even numbers. It must also be noted that. each right-angled bend at the termination of a line segment. the line segments are represented by odd numbers. 8. Fig.3. 8.2 shows the segmented representation of the antenna structure.Lumped Circuit Model Analysis 8.3 Antenna modeling procedure Although the derived model is implemented on and verified for the meander and Peano line antennas in particular. To model the above-depicted antennas in terms of lumped circuit elements we first need to view the antenna as a cascade connection of several individual line segments.2 (a) Segmented representation of meander line antenna (b) Segmented representation of Peano line antenna 67 . needs to be considered separately. The segments are then numbered.1 Segmentation of the antenna The antenna is first segmented into basic elements. (a) (b) Fig. as the basic theory is the same. The segmentation process involves the dividing of the complicated antenna structure into simpler units comprising of individual line segments and right-angled bends terminating the line segments. For ease of understanding. Each of the steps is described in detail. yet the model is expected to be valid for all other antennas based on space filling curves in microstrip configuration. Each segment is then replaced by its equivalent lumped circuit model and the entire cascaded circuit is terminated by a suitable capacitive termination to account for the fringing fields from the open-ended line.

8.2b) where. The lumped-parameter equivalent for a particular length of transmission line can be employed to each of these line segments [1-2]. the expressions above are modified to: 68 . (a) Fig. 8. are readily obtained by equating transmission line terminal voltages and currents with network terminal voltages and currents [1]:  γl  Z A = Z 0 tanh  2 ZB = Z0 sinh γl (8. 8.2 Modeling of line segments The odd numbered segments in both the antennas shown in Fig.1b) and Z Q = Z 0 sinh γl  γl  Z P = Z 0 coth  2 (8.2 represent line segments th0at can be considered as distributed transmission lines of particular lengths.equivalent network for a transmission line (b) Expressions for the various impedances. given next. Fig. Considering the strip to be loss-free.3.3 (a) T-equivalent network for a transmission line (b) π.1a) (8. l is the length of the line segment and Z0 is its characteristic impedance and γ being the propagation constant.Lumped Circuit Model Analysis 8.2a) (8.3 shows the T-equivalent and π-equivalent network for a two port distributed transmission line.

5b) The approximation tan x ≈ x & sin x ≈ x.5a) & (8. Comparing (8. ZA being positive indicates an inductive reactance. ZA and ZB from (8. and ω is the operating angular frequency. we get 69 . while the negative sign in ZB indicates a capacitive reactance. the condition l << λeff/4 is satisfied.5a) ZB = − jZ 0 βl (8.6a).3a) ZB = and − jZ 0 sin βl (8.Lumped Circuit Model Analysis  βl  Z A = jZ 0 tan   2 (8. Under such condition.3a) and (8.5b) respectively can now be expressed in the form of Z A = jω Ll 2 (8.6a) ZB = 1 jωCl (8.5a) & (8.4b)  βl  Z P = − jZ 0 cot    2 We now consider the case of T-network only. But the model is equally valid if the line segments are modeled by equivalent π-network. Since the line segments are very short in length as compared to the effective wavelength.3b) are modified to: Z A = jZ 0 βl 2 (8. for small values of x has been used.3b) Z Q = jZ 0 sin βl (8. (8. C is capacitance per unit length.4a) (8. while operating at a lower frequency range. L is inductance per unit length.6b) where. therefore.

and εeff is the effective permittivity given by [3] as ε eff = ε r +1 ε r −1 2 + 2 × 1 h 1 + 12 w (8.10b) Lumped capacitance.9a) Similarly. from (8. if Z0 is expressed in ohms. we get.6b). then the units of L and C are Henry per meter (H/m) and Farad per meter (F/m) respectively. L A = Ll 2 (8. and speed of light in meters per second. λeff = λ0 c .8) Where. C= ε eff cZ 0 (8.7) Substituting. The corresponding units for lumped inductance LA and capacitance CB for a length l expressed in meters are Henry (H) and Farad (F) respectively. λ0 = .9b) From the inductance and capacitance per unit length (L and C respectively). ω = 2πf f ε eff where λeff is the effective wavelength and λ0 is the free space wavelength. and c is speed of light in free space. C B = Cl In the above expressions.10a) (8. εr is the relative permittivity of the substrate. h is the height of the substrate and w is the width of the strip.5b) & (8. L= Z 0 ε eff c (8. Thus. comparing (8. the lumped inductance and capacitance (LA and CB respectively) are calculated as follows: Lumped inductance. 70 . f is the operating frequency. β= 2π λeff .7) we have.Lumped Circuit Model Analysis Ll  βl  jZ 0   = jω 2  2 (8.

especially bearing in mind that most of the current flows in the outer edges of microstrip. each of which can also be modeled as an equivalent lumped T-network.3 Modeling of right-angled bends In Fig. 8. 8. 8. The inductances on the other hand. 8. This is considerable.4 shows the lumped equivalent T-network for an individual line segment. 8.4 Lumped equivalent model of a line segment values shall also vary with the permittivity of the substrate in case of mixed dielectric. depending only on its length. the Fig.5 Right-angled microstrip bend 71 . as evident from the expressions above.5 shows the right-angled bend and its equivalent circuit. 8. can be modeled as an equivalent L-C-L lumped T-network. Fig. with inductances in the symmetric arm and capacitance in the shunt arm. The capacitance arises due to additional charge accumulation at the corners – particularly around the outer point of the bend where electric fields concentrates. each denoted by odd numbers in Fig. if the segments are of varying widths. in our case for the antennas considered. The line segments. the even numbered segments represent right-angled bends. The values of LA and CB may vary from segment to segment. Q Q' w w (a) Structure and nomenclature (b) Equivalent Circuit Fig. However in general. and on characteristic impedance of the segments. 8.2.Lumped Circuit Model Analysis Fig.2.3. arise because of current flow interruption at the sharp discontinuity caused due to the bend [1].

4 Modeling of the open end Assuming input excitation at one end of the antenna.11) is quoted as within 5 percent over the ranges: 2. Cb is the bend capacitance. [6] are listed below for the evaluation of bend capacitance: For w/h < 1: C b (14ε r + 12. there are essentially three phenomena associated [1]: a) There are fringing fields extending beyond the abrupt physical end of the metallic strip. The accuracy of equation (8.Lumped Circuit Model Analysis Capacitance data has been determined theoretically by Silvester and Benedek [4] and inductance data by Thomson and Gopinath [5].83ε r − 2.1 ≤ w/h ≤ 5. Closed form formulae as provided by Gupta et al. that may be a desired phenomenon if it interferes constructively. It is because. though in our case of antenna. at the open end.5) w / h − (1.5 ≤ εr ≤ 15 and 0. 72 . b) Surface waves are launched at the end of the strip.0 w h pF/m (8.5 ≤ w/h ≤ 2. The accuracy of equations (8.2ε r + 7.5ε r + 1. the other end is open and needs to be modeled by some means.3.0. 8.12) Here.25) = w w/ h pF/m (8. c) Energy is radiated from the open end. on a substrate of height h. leading to spurious radiation in case of microstrip circuits.21 h  h  nH/m (8.11b) and bend inductance:  w  Lb = 1004 − 4.11a) For w/h > 1: Cb w = (9.25) + 5.12) is quoted as about 3 percent for the range: 0. Lb is the bend inductance for a bend formed by the junction of microstrip lines of width w.

0240 -0. The whole circuit looks like that as shown in fig.5 4.443 -0.1 COEFFICIENTS Kε FOR EQUATION (8.2170 -0. at the open end of last numbered segment). 73 .5 Cascading of the equivalent circuit Having derived the lumped circuit equivalent for each segment. they are duly cascaded in proper order and terminated by the open circuit capacitance Coc at one end (i.2220 0. which is listed below: i −1 5  C oc w     = exp2.0267 -0.403 -0.0260 -0. This circuit can be solved using circuit simulators to obtain the input reactance response with respect to frequency. 8.2 9. The fringing fields from the open end may be accounted for by assuming some equivalent capacitance connected at the open end.0540 -0.0 1. the first one plays a predominating role.Lumped Circuit Model Analysis Among the above phenomena. The expression for the open circuit capacitance Coc has been formulated by Silvester and Benedek [7].0113 -0.13) where the coefficients Kε are given in Table 8.6.0147 2.0073 -0. it is the only one considered for further analysis.1062 -0.6 16. The other end (open end of segment number 1) serves as the input port of the antenna.2535 0.2036 ∑ K  log   ε w h   i =1    pF/m (8.e. Hence.0267 -0.0840 8.3.0 51.1 TABLE 8.0 2.13) [7] εr i 1 2 3 4 5 1.

4 Implementation of the lumped equivalent circuit model In this section.5 mm. The relative permittivity (εr) of the substrate is taken to be 4. 9 etc. 8.4.5 mm2.5 mm 1. with a height ‘h’ as 1 mm. are each 22.5 X 25.5 mm Fig. viz. 8. the model derived above is implemented on two antennas.1 Meander line antenna Dimensions: The first antenna taken for analysis is the microstrip meander line antenna. Segments 1 & 33 are each 24 mm long.3. The footprint area of the antenna is 25. 5.6 Circuit representation of the antenna using lumped models 8.7.2. as shown in Fig.7 Simulated antenna structure (Meander line antenna) 74 . 8. 8. The results obtained from the analysis of the circuit model are compared with that obtained from the antenna simulation.5 mm 25.5 mm 25. The dimensions and the basic parameters for both the antennas are stated. The width of the microstrip line ‘w’ is chosen as 1. namely meander line and Peano line antennas. As evident from Fig.Lumped Circuit Model Analysis Fig 8. this antenna structure when segmented yields 33 segments. Other line segments parallel to segments 1 & 33.5 1. The results relating to both are shown.

7182818… For our case. and CB = 2. W = 1. for 24 mm line segments: • LA = 4.5 mm line segments: • LA = 0. 11. 3. and CB = 0.14) For ‘wide’ strips (w/h > 3. and they constitute the effective radiating elements of the antenna. Apart from the line segments. 15 etc.15) For 22.14 ohms.9) & (8. (8.11) & (8. The remaining odd numbered smaller line segments viz.2.9 (8. Z0 = 58.5mm X 1.5mm. Therefore.1387 pF 75 [using (8.307 pF [using (8.14) is used for analysis.3): 119.10)] −1 (8. Now.10)] For 1. and Cb = 0.94  2π 2 ε r  2h π    r   r    where ‘e’ is the exponential base: e = 2.46 pF [using (8.3): Z0 = 2       ln 4 h + 16 h  + 2  − 1  ε r − 1  ln π + 1 ln 4         2(ε r + 1)   w  w  2  ε r + 1  2 ε r π      119. are each 1.16 nH.0688 nH. They are stated below: For ‘narrow’ strips (w/h < 3.5.9) & (8. Also.9π  w ln 4 ln(eπ 2 / 16)  ε r − 1  ε r + 1  πe   w  2 + Z0 = +  +  ε  2πε ln 2 + ln 2h + 0. as all the line segments are of equal width w = 1.Lumped Circuit Model Analysis mm long. therefore the characteristic impedance Z0 is same for all of them.5 mm and placed on a single uniform substrate.12)] .9) & (8.26 nH.9 nH.14). each of dimensions 1.5 mm long. 7.8). using (8. and CB = 2. w/h = 1. We obtain from (8. we obtain εeff = 3.5 mm line segments: • LA = 3. Analysis: Standard analysis formulae in closed form for the calculation of Z0 are readily available in literature [1].1538 pF [using (8.10)] For right-angled bends: • Lb = 0.5 mm and h = 1 mm implies. there are 16 right-angled bends denoted by even numbers.

Fig. 8. and the circuit looks like that in Fig. Fig. As shown in Fig. Since both simulators MDSPICE and MODUA show similar resonant and anti resonant properties only the MODUA results are presented below. the entire circuit is not shown in detail. a wave-port at the input end is applied for simulation.Lumped Circuit Model Analysis For the open-circuited end: • Coc = 0. the inductance values are in nH.8.9(a) shows the input reactance versus frequency plots of the antenna using IE3D.7.0329 pF [using (8. 8. 8.7 Results: The antenna is simulated using IE3D from Zealand Software Inc.13)] Now having obtained all the circuit parameters. That is evident. the input resistance shoots up to a maximum value. in the frequency range of 0 to 1 GHz. For clarity.8 Equivalent circuit of the meander line antenna shown in Fig. 8. The resonant frequency points as observed from IE3D simulations are in good agreement with the results obtained from MODUA and MDSPICE circuit simulations. the units being avoided in the representation. where there is a sharp transition from inductive to capacitive reactance. 8. correspond to the anti-resonant frequencies. and the capacitances are in pF.138 Fig. The circuit elements on the other hand are simulated using MDSPICE and MODUA from Zeland Software Inc. if we consider input resistance versus frequency plots from IE3D. 0. The frequency points of zero reactance. 8.9(b) shows the input reactance versus frequency plots of the equivalent circuit using MODUA simulator. where we observe that at those anti-resonant points. the lumped circuit for each segment is duly cascaded. The resistance plots have been 76 . In addition.

8. The antiresonant points from both simulations also match closely. fr1. A comparative study 77 .9 Input reactance versus frequency plots (meander line antenna) In the above figures. 2nd and 3rd resonant frequencies respectively. as our aim lies mainly in finding out the resonant frequencies. (a) Antenna simulation results using IE3D (b) Circuit simulation results using MODUA Fig. fr2 and fr3 represent the 1st. while far1 and far2 represent the anti-resonant frequencies.Lumped Circuit Model Analysis avoided though.

9855 7.041 8.2 Peano line antenna 25. 1.10.80 715.55 362.507 % error 9.69 8.89 886. TABLE 8. 8.5 mm Fig.6165 591. 8.096 400.753 8.4.5 mm 1.57 539. we turn our attention towards the next class of 25.10 Simulated antenna structure (Peano line antenna) 78 .5 X 25.the Peano line antennas.2 COMPARISON OF IE3D & MODUA RESULTS FOR MEANDER LINE ANT.959 785. Dimensions: The dimensions for the microstrip Peano line antenna are shown in Fig.5mm antennas.685 200.329 9.685 MODUA results (Circuit) 181. IE3D results (Antenna) 1st resonance (MHz) 1st antiresonance (MHz) 2nd resonance (MHz) 2nd antiresonance (MHz) 3rd resonance (MHz) 960.Lumped Circuit Model Analysis between the resonant and the anti-resonant frequencies in both figures are presented in Table 8. The antenna footprint area in this case also is taken as 25.5 mm Having verified the validity of the model on meander line antennas.2.5 mm2.

6153 pF [using (8. 17.5 mm long. w/h = 1. yields 81 segments. and CB = 0.5 mm.9) & (8.384 pF [using (8.13)] 79 . 37. 8.2.5. Therefore.5 mm line segments: • LA = 2. 15. viz. The remaining odd numbered line segments of smallest length. are 13. W = 1.5 mm line segments: • LA = 0. 13.4615 pF [using (8. with a height ‘h’ as 1 mm.12)] For the open-circuited end: • Coc = 0.9) & (8.5 mm and h = 1 mm implies. and CB = 1.8).0688 nH. segments 1 & 81 are each 6 mm long.04 nH. Z0 = 58. denoted by an even number. Other smaller segments.14) is used again for analysis.5 mm line segments: • LA = 0.11) & (8.26 nH.2 reveals that a 2nd order Peano line antenna when segmented. for 13. we obtain εeff = 3.78 nH.14). 25 etc are 4.5 mm and placed on a single uniform substrate. and CB = 0.5 mm long. Now.0329 pF [using (8.Lumped Circuit Model Analysis As in case of meander line antenna.3. We obtain from (8.10)] For right-angled bends: • Lb = 0. 5. (8. The relative permittivity (εr) of the substrate is taken to be 4. parallel to 1 & 81 viz. and CB = 0. 45 etc. viz.5mm X 1. Analysis: For the present case also. The longer line segments. as all the line segments are of equal width w = 1.9) & (8. Each of the right-angled bends.14 ohms.10)] For 6.10)] For 4.5 mm long. 9.34 nH. 19 etc are each 1. using (8. For our example.0 mm line segments: • LA = 1. 21. is of dimension 1. Fig.10)] For 1.1538 pF [using (8.5mm.9) & (8.1387 pF [using (8. width of the microstrip line ‘w’ is chosen as 1. parallel to 1 & 81. Also. 3. therefore the characteristic impedance Z0 is same for all of them. and Cb = 0.

The final circuit is shown in Fig. (a) Antenna simulation results using IE3D 80 . where the inductances are in nH and the capacitances are in pF. 8. Fig. IE3D and MODUA results for the Peano line antenna are shown.11. 8. 8. As described earlier. The input reactance versus frequency plots are depicted in Fig.10 Results: In this section.11 Equivalent circuit of the Peano line antenna shown in Fig. IE3D results are the antenna simulation results. 8.Lumped Circuit Model Analysis The final step is to cascade the individual lumped equivalents of the segments and to terminate it by Coc.12. while MODUA results deal with the circuit simulation results.

466 4.836 % error 6.88 81 .22 967.534 MODUA results (Circuit) 193.06 780. IE3D results (Antenna) 1st resonance (MHz) 1st anti-resonance (MHz) 2nd resonance (MHz) 2ndantiresonance(MHz) 3rdresonance(MHz) 996.808 2.014 207.Lumped Circuit Model Analysis (b) Circuit simulation results using MODUA Fig.575 806.3 COMPARISON OF IE3D & MODUA RESULTS FOR PEANO LINE ANT.60 392. 8.164 413.164 606.137 3.12 Input reactance versus frequency plots (Peano line antenna) TABLE 8.97 581.507 4.

the phenomena of multiple resonance or higher order modes were explained using the concept of distributed transmission lines. than relatively longer and lesser number of line segments for a meander line antenna. This model considers the antenna as a reactive circuit. within the same footprint area and with the same path length traced. can now explain the change in resonant frequency. Thus. To obtain proper resistance plots. Each time the probe feed position along the antenna is changed. • Earlier. • However. • For accurate evaluation of reflection coefficient and VSWR at the feed. a marginal percentage of error (< 10 %) in the resonant and the anti-resonant frequencies can be attributed to the electromagnetic coupling between the parallel and adjacently placed line segments. The input resistance of the antenna is the radiation resistance Rrad. The comparative study and the reactance plots reveal that the resonant and the anti-resonant frequency points from IE3D and MODUA simulations are in close proximity to each other. which has been ignored in our analysis. it essentially adds a shunt inductance at a different position of the equivalent cascaded circuit. It is because. It can also be explained using the 82 . 8. incorporating a Peano line antenna calls for more numbers of smaller line segments.2 and Table 8.5 • Discussions As listed in Table 8. which essentially depends on the radiated power Prad. each time the probe is fed at a different position. which shall facilitate the accurate measurement of VSWR. the percentage of error is less for a Peano line antenna than for a meander line antenna. the resonant frequency is bound to change in each case. the field configuration of the antenna needs to be known. thereby changing the entire circuit itself. as observed in chapter 7. and thus provides us with the reactance plots. and hence on the field configuration. • The derived circuit model.Lumped Circuit Model Analysis Table 8.3. lumped approximation is more accurate. by virtue of its cascaded structure. radiation resistance needs to be calculated.3 illustrates a comparative study between IE3D and MODUA results for the Peano line antenna. As the line segments are smaller in case of Peano line antenna.

T-junction and crossings”. John Wiley and Sons Inc. Prentice Hall India. As the circuit itself is a cascaded connection of many T-networks. C.6 Summary This chapter presented a simple circuit model analysis of a meander line antenna and a Peano line antenna to compute its resonant frequency. Bhartia. 2nd Edition. IEEE Trans. 1991 J. Vol.23.C.. “Calculation of microstrip discontinuity inductances”.21. Silvester. Benedek. P. Moreover. Gopinath. I. 2005 P. 8. The lumped circuit equivalent of the antenna has been derived and validated against a standard electromagnetic simulator. John Wiley And Sons.A.648-655. No. 20. No. • This circuit model is valid as long as the dimensions of the individual segments are small enough as compared to the effective wavelength. The computed results reveal a close match with an error percentage of less than 10%. Silvester. MTT. Gupta. MTT. “Microstrip discontinuity capacitances for right-angle bends. explains the reason for multiple resonance and higher order harmonics.. each having their own resonance and reactance profile. Finally. Lines and Fields. Foundations for Microstrip Circuit Design. Balanis. 1972 83 . All these reactance profiles taken into account together. Bahl. there are several permutations and combinations of the T-segments grouped together. R. the circuit as a whole has its own resonance. No.F. 8. IEEE Microwave Theory and Techniques. Garg. Benedek. P. August 1975 K.. May 1973 A. Ryder. pp. Edwards. 341-346. 2nd Edition. A. Thomson. Antenna Theory: Analysis and Design. 511 – 516. pp. 3rd Edition. pp. “ Equivalent capacitances of micostrip open circuits”. for a particular frequency range of interest. 5. References [1] [2] [3] [4] [5] [6] [7] T. Artech House Publishers. Several inferences from the circuit model were also discussed. 8. 2nd Edition. Networks. D. each of the T-network segments has its own resonant frequency and reactance properties. Microstrip Lines and Slot Lines. 1996 P. IEEE Trans. 1959 C.Lumped Circuit Model Analysis concept of lumped equivalent circuit model. P.

Chapter 9 Conclusion and Scope for Future Works .

A simple lumped circuit equivalent model has been presented in the chapter. a design problem has been dealt with.Conclusion and Scope for Future Works Chapter 9 9. Interestingly. Chapter 7 provided a brief study of the change in antenna characteristics with changing feed position. 9. This thesis gives some insight into design and circuit analysis studies on the antenna.2 Scope for future research works There is a good scope for future research work in this field of study. and finally integrate the total system by using Multiport Network Model [2]. In Chapter 6. the concept of RFID technology has been discussed and the theoretical background of the antenna has been elucidated. • The design related aspects of the antenna for very low RFID frequencies. Some important observations regarding changing resonant frequency with feed position. However. more attention has been given towards the difficulties faced during the design. Some of the major areas are highlighted below. • • Impedance matching related issues might also be dealt with. Another method is the use of numerical techniques to evaluate the current distribution on 84 . The structure being a highly convoluted one. has been presented in that chapter. that may be taken up as a way forward for works in future. the analysis of its radiation pattern is not so straightforward. Chapter 8 dealt with analytical aspect of the antenna. as most of the avenues are still unexplored.1 Conclusion and Scope for Future Works Conclusion The entire thesis deals with an investigative study on microstrip Peano line antennas. viz. A chapter on review of the antenna has been included to portray an update of the latest works done in this field. the analyses of radiation characteristics are yet to be explored. excitation of different modes etc. This provides an interesting area of research. can be a worthy and challenging aspect of research. In the initial chapters. apply the cavity model [1] of microstrip patch antenna on each of them. One suggestive method is to segment the antenna into several rectangular cavities. their causes as well as their remedies. and its reliability has been validated. 26 MHz for biomedical applications.

A. 3rd Edition. 2001 85 . the radiation resistance can be evaluated from the radiated power. P. Ittipiboon. and finally the vector addition of radiation in free space. Antenna Theory: Analysis and Design.. and the VSWR characteristics at the feed can then be accurately determined. Bahl.. Arctec House Inc. due to each of them. A. Balanis. This completes the circuit model. 2005 R. Garg. I. John Wiley and Sons Inc. References [1] [2] C.Conclusion and Scope for Future Works each of the individual segments. Microstrip Antenna Design Handbook. • Once the radiation pattern analysis is possible. Bhartia.

Sign up to vote on this title
UsefulNot useful