jees-19-3.pdf - Journal of Electromagnetic Engineering and ...

JEESJournal of Electromagnetic Engineering and Science

Aims and Scope The Journal of Electromagnetic Engineering and Science (JEES) is an official English-language journal of the Korean Institute of Electromagnetic and Science (KIEES). This journal was launched in 2001 and has been published quarterly since 2003. It is currently registered with the National Research Foundation of Korea and also indexed in Scopus, CrossRef, EBSCO, DOI/Crossref, Google Scholar, and Web of Science Core Collection as Emerging Sources Citation Index (ESCI) journal. The objective of JEES is to publish academic as well as industrial research results and discoveries in electromagnetic engineering and science. The particular scope of the journal includes electromagnetic field theory and its applications; high frequency components, circuits, and systems; antennas; electromagnetic wave environments; and relevant industrial developments. In addition to the regular stream of contributed papers, the journal occasionally publishes special issues on specific topics of current interest, to promote a focused coverage of the selected topics. In addition, JEES regularly posts invited review papers by leading researchers in the relevant fields, which will provide an overview of various subjects from their own perspectives. The editors of JEES are aware of the importance of rapid publication for the delivery of new research results to its readers with minimum delay, and fully recognize the need for a short turnaround time. Manuscripts submitted to JEES take about 3–5 months from the initial submission to publication. This rapid process applies to all three types of manuscripts that JEES publishes: regular papers, letters, and invited papers. The publication of submitted manuscripts is subject to peer review, and the final decision by the Editor-in-Chief is regularly made based on three reviews from experts in the relevant area. This journal actively promotes research in electromagnetic wave-related engineering and sciences, and its ultimate aim is to benefit the global electromagnetic engineering and science community.

Subscription Information JEES is published quarterly in January, April, July, and October. Full text PDF files are available at the official website (http://www.jees.kr). The annual subscription rates for this journal are USD 230 for institutional members, USD 55 for regular members, USD 37 for associate members, and USD 28 for student members.

ISSN 2671-7255 (Print) ISSN 2671-7263 (Online)

Jul. 2019Vol. 19 No. 3

Published byThe Korean Institute of Electromagnetic Engineering and Science

04376, #706 Totoo Valley, 217 Saechang-ro, Yongsan-gu, Seoul, Korea Tel : +82-2-337-9666 / 332-9665Fax : +82-2-6390-7550http://www.kiees.or.krE-mail : [email protected]

Printed July 25, 2019Published July 31, 2019

Printed byGuhmok Publishing CompanyTel : +82-2-2277-3324Fax : +82-2-2277-3390E-mail : [email protected]

This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

@This paper meets the requirements of KS X ISO 9706, ISO 9706-1994 and ANSI/NISO Z.39.48-1992 (Permanence of Paper).

©Copyright The Korean Institute of Electromagnetic Engineering and Science. All Rights Reserved.

KOFST This journal was supported by the Korean Federation of Science and Technology Societies Grant funded by the Korean Government (Ministry of Education).

JEESJournal of Electromagnetic Engineering and Science

REGULAR PAPERS

Miniaturized Multi-Section Power Divider with Parallel RC Isolation CircuitYoung-Chul Yoon · Young Kim

Circularly Polarized Loop-Type Ground Radiation Antenna for IoT ApplicationsZeeshan Zahid · Longyue Qu · Hyung-Hoon Kim · Hyeongdong Kim

A Modified Stripmap SAR Processing for Vector Velocity Compensation Using the Cross-Correlation Estimation MethodChul-Ki Kim · Jung-Su Lee · Jang-Soo Chae · Seong-Ook Park

A Wideband Circularly Polarized Slot Antenna Backed by a Frequency Selective SurfacePuneeth Kumar Tharehalli Rajanna · Karthik Rudramuni · Krishnamoorthy Kandasamy

Health Effects Study and Safe Distance Determination Near an Electrical Distribution NetworkMehdi Nafar

Analysis and Modeling of GPR Signals to Detect Cavities: Case Studies in MoroccoGamil Alsharahi · Ahmed Faize · Carmen Maftei · Mohamed Bayjja · Mohamed Louzazni · Abdellah Driouach · Abdellatif Khamlichi

A Compact Rx Antenna Integration for 3D Direction-Finding Passive RadarHojoo Lee · Jaesik Kim · Jaehoon Choi

Photoacoustic Tomography System for Roughly Painted Micro ObjectsAndreas Setiawan · Fransiscus Dalu Setiaji · Gunawan Dewantoro · Nur Aji Wibowo

A Simple Empirical Model for the Radar Backscatters of Skewed Sea Surfaces at X- and Ku-BandsTaekyeong Jin · Yisok Oh

Design of a Ka Band Multi-Channel (25CH) Millimeter-Wave Downconverter for a Signal Intelligence SystemYuseok Jeon · Jaejin Koo · Hyunkyu Kim

147

153

159

166

172

177

188

197

204

210

This journal was supported by the Korean Federation of Science and Technology Societies Grant funded by the Korean Government (Ministry of Education).

ISSN 2671-7255 (Print) ISSN 2671-7263 (Online)

Jul. 2019Vol. 19 No. 3

JOURNAL OF ELECTROMAGNETIC ENGINEERING AND SCIENCE, VOL. 19, NO. 3, 147~152, JUL. 2019

https://doi.org/10.26866/jees.2019.19.3.147

ISSN 2671-7263 (Online) ∙ ISSN 2671-7255 (Print)

147

I. INTRODUCTION

The power divider is a vital element in communication sys-

tems such as power amplifiers, mixers, and transceivers, owing

to its simple design and superior performance. The conventional

power divider has good matching of input and output, isolation

performance, and low loss in narrow band operation. Due to its

narrow band characteristics, its application has been limited in

the current large-capacity communication systems of WCDMA,

LTE, and 5G.

To address this limitation, a multi-section divider [1–5] has

been proposed. Although the multi-section divider has broad-

band characteristics with an achievable bandwidth greater than

100%, it has the disadvantages of a large circuit and increased

insertion loss.

Recently, various designs to reduce the circuit size of power

dividers have been researched, such as a structure using short

uniform transmission line [6, 7], an isolation circuit positioned

in a quarter-wavelength transformer [8, 9], and artificial trans-

mission lines built up with lumped elements [10].

However, since these dividers use a λ/4 transformer between

the input and output or between the input or output and 50 Ω

termination, they cannot achieve an overall reduction in size. In

addition, the bandwidth is reduced due to the impedance trans-

former, and the parasitic uncertainty of lumped elements is not

desirable.

The power divider proposed in this paper has a multi-section

structure to realize broadband characteristics, with the electrical

length of each power splitting section less than a quarter wave-

length to reduce the circuit size. The characteristic impedance

of each power-splitting section can be controlled within a cer-

tain range while avoiding high impedance. All input and output

ports are designed with 50 Ω termination, so there is no need

Miniaturized Multi-Section Power Divider

with Parallel RC Isolation Circuit Young-Chul Yoon1 · Young Kim2,*

Abstract

This paper proposes a miniaturized multi-section power divider with parallel resistor and capacitor isolation circuits. Each of the section

transmission lines constituting the proposed multi-section power divider has a length less than a quarter wavelength and the impedance of

the transmission lines can be controlled within a certain range while avoiding high impedance. In addition, each circuit connected be-

tween the transmission lines consists of a parallel resistor and capacitor to satisfy the matching and isolation characteristics between out-

put ports. To validate the proposed power divider, two circuits at a frequency of 2 GHz were designed with two- and three-section divid-

ers. In comparison to a conventional power divider, the proposed divider showed broadband performances with a small structure.

Key Words: Miniaturization, Multi-Section, Parallel Resistor and Capacitor Isolation Circuit, Power Divider.

Manuscript received November 26, 2018 ; Revised January 16, 2019 ; Accepted February 28, 2019. (ID No. 20181126-082J) 1Department of Electronics Engineering, Catholic Kwandong University, Gangneung, Korea. 2School of Electronic Engineering, Kumoh National Institute of Technology, Gumi, Korea. *Corresponding Author: Young Kim (e-mail: [email protected])

This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits

unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

ⓒ Copyright The Korean Institute of Electromagnetic Engineering and Science. All Rights Reserved.

JOURNAL OF ELECTROMAGNETIC ENGINEERING AND SCIENCE, VOL. 19, NO. 3, JUL. 2019

148

for an impedance transformer. In addition, a parallel resistor-

capacitor circuit structure is adopted to obtain the broadband

output matching and isolation characteristics of the proposed

power divider.

II. THEORY AND DESIGN

The structure of the proposed N-section power divider is

shown in Fig. 1. It consists of transmission lines with electrical

lengths (θ1, θ2, … , θN) less than 90° and isolation circuits with

parallel resistors (R1, R2, … , RN) and capacitors (C1, C2, … ,

CN). The proposed power divider is analyzed using the even-

and odd-mode analysis methods; Fig. 2(a) and (b) show the

even- and odd-mode equivalent circuits, respectively.

To obtain the electrical lengths and characteristic impedances

of the transmission lines, the matching mechanism of the cas-

cade transmission lines [11] is presented in Fig. 3. As seen in

Fig. 3(a) and (b), it is possible to convert a part of the imped-

ance transformer with Za and 90° to another impedance Zb and

electrical length θb. To match the conversion portion of the im-

pedance transformer, the admittance values of Yin,a and Yin,b must

be the same. By solving the equation for the real and imaginary

parts of the admittance (Yin,a, Yin,b), the impedance Zb and termi-

Fig. 1. Schematic of the proposed power divider.

(a)

(b)

Fig. 2. Equivalent circuits of even-mode (a) and odd-mode (b).

(a)

(b)

(c)

Fig. 3. Matching mechanism of the cascade transmission lines: (a)

conventional λ/4 transformer, (b) proposed transformer

with two sections, and (c) proposed transformer with three

sections.

nating impedance ZLb of the proposed two-section transformer

for θb change can be obtained as:

2 2 2 2 2 2 2(B ) 4 (1 )

2

a La a La a La

a La

BZ Z Z Z B Z Zb Lb Z ZZ Z

(1) 2 2 2 2

2 2 2

(1 ( ) tan )(1 tan )

(1 tan )(1 ( ) tan )a La a b

a b Lb b La

Z ZLb Z Z ZZ

(2)

where

2

2

tan (1 tan )

tan (1 tan )a b

b aB

.

Ahn [12] confirmed that when the electrical length (θa < θb <

90°) and termination conditions (ZLa < ZoS) are satisfied, Zb and

ZLb should satisfy Zb < Za and ZLb < ZLa. Using the above condi-

tions, the impedance Zc and electrical length θd of the three-

section transformer can be obtained in the same way, as seen in

Fig. 3(c). Using this method, all the electrical lengths of the

multi-section transmission line, which is the proposed divider

with each section length less than λ/4 wavelength, can be ob-

tained.

In addition, for the odd-mode equivalent circuit shown in Fig.

2(b), the following equations can be obtained:

, tan N

N

Yin N N NY G jB j

(3)

, 1 1

1 , 1

tan

, 1 1 1 1 tan

in N N N

N in N N

Y jY

in N N N N Y jYY G jB Y (4)

,1

,1

o in

o in

Y Yii Y Y

(5)

YOON and KIM: MINIATURIZED MULTI-SECTION POWER DIVIDER WITH PARALLEL RC ISOLATION CIRCUIT

149

Fig. 4. Equivalent circuit of the entire schematic.

where 2NN RG , 2N NB C , 1

NN ZY , and ii is the reflec-

tion coefficient at port 2 (i = 2) and port 3 (i = 3).

The conductance (G1, G2, …, GN) and susceptance (B1, B2, …,

BN) values can be obtained to satisfy the condition that the re-

flection coefficient of Eq. (5) becomes zero.

To achieve isolation performance between ports 2 and 3, the

S-parameter S32 should be minimized as shown in Fig. 4. To

calculate the isolation characteristic of the divider network,

ABCD and admittance parameters are used. The ABCD ma-

trix of the circuit NetN, which is composed of two transmission

lines (ZN, θN) and a shunt resistor Zo, can be expressed as:

cos sin 1 0 cos sin

sin cos 1 sin cosN N

N N

Net Net N N N N N N

N N N o N N NNet Net

A B jZ jZ

jY Y jYC D

(6)

The admittance matrix of the composite network NetRCN re-

sistor-capacitor can be represented by:

1 1" "11 12" " 1 121 22

N N

N N NRCN N

R R N N

N NR R CNet R

j C j Cy y

j C j Cy y (7)

Since the network NetN is connected in parallel with network

NetRCN, we needed to transform Eq. (6) from an ABCD matrix

to an admittance matrix. Finally, the composite admittance ma-

trix of NetN and NetRCN can be obtained as follows:

' ' " "11 12 11 12 11 12

' ' " "21 22 21 22 21 22

N RCN N RCNNet Net Net Net

y y y y y y

y y y y y y (8)

To obtain the equation of the networks, which are connected

by two cascade transmission lines (ZN-1, θN-1) and the composite

network of NetN and NetRCN, the composite network parameters

are needed to convert the ABCD matrix. We can compute the

ABCD parameters of the networks NetN-1 and NetRCN-1 as pre-

viously calculated. Through this process, the admittance param-

eters of the entire network can be obtained. The S-parameter

value of isolation, S32, is finally obtained by parameter conver-

sion, and the isolation circuit values of the resistor and capacitor

are selected to satisfy the S32 parameter condition of |S32| ≤ -25 dB

at the same time among the values satisfying Eq. (5).

III. SIMULATION AND EXPERIMENTAL RESULTS

To confirm the validation of the design rule, we designed

two- and three-section power dividers operating at a frequency

of 2 GHz and implemented on the substrate of Rogers

RO4350B PCB (εr = 3.48 and h = 0.762 mm). The simulation

was performed using the Microwave Office software developed

by National Instruments.

The two-section power divider configuration process is sum-

marized as follows. First λ/4 impedance transformer of ZoT =

50 Ω, ZLT = 35.35 Ω, ZT = 42.04 Ω and θT = 90° is chosen as

shown in Fig. 5(a).

Based on the decomposition theory of a single transmission

line [13], the impedance transformer can be divided into two

transmission lines of ZoTD = 100 Ω, ZLTD = 70.7 Ω, ZTD =

84.08 Ω, and θTD = 90°, as shown in Fig. 5(b). To constitute the

two-section transmission line of the divider, the following pa-

rameters are set: Za = 84.1 Ω and θa = 60°. The parameters of

Zb = 69.56 Ω, ZLb = 50 Ω, and θb = 76° can be calculated using

Eqs. (1) and (2). To obtain the parameter values for the three-

section lines, we used a similar method as follows:

The following parameters are set: Za = 84.1 Ω and θa = 50°.

The three-section parameters of Zb = 74.44 Ω, ZLb = 59.22 Ω,

θb = 66°, Zc = 67.60 Ω, ZLc = 50 Ω, and θd = 73° can be calcu-

lated as described previously. Finally, the resistor-capacitor val-

ues satisfying the reflection and isolation characteristics of the

output ports are obtained using MATLAB and summarized as

follows: In the case of the two-section divider, R1 = 120 Ω, C1 =

(a)

(b)

Fig. 5. Decomposition of transmission lines: (a) λ/4 transformer

and (b) decomposition of single λ/4 transformer.


150

0.02 pF, R2 = 110 Ω, and C2 = 0.18 pF. In the case of the three-

section divider, R1 = 820 Ω, C1 = 0.01 pF, R2 = 180 Ω, C2 =

0.02 pF, R3 = 150 Ω, and C3 = 0.17 pF.

Fig. 6(a) and (b) show photographs of the two- and three-

section proposed power dividers. The total length of the two-

section divider is 106°, and that of the three-section divider is

179°. Fig. 7(a) and (b) show the measured and simulated S-

parameters of the two-section power divider. The insertion loss-

es of |S21| and |S31| were -3.10 dB, and the isolation of |S32| was

greater than 25 dB. The input return loss of |S11| was better

than -25 dB, and the output return losses of |S22| and, |S33|

were better than -30 dB at the center frequency of 2 GHz. The

measured -15 dB bandwidths of |S11|, |S22|, |S33|, and |S32| were

in the ranges of 1.33–2.46 GHz, 0.5–3.63 GHz, 0.5–3.35 GHz,

and 1.26–2.6 GHz, featuring fractional bandwidths of 56.5%,

156%, 142%, and 67%. In addition, Fig. 8(a) and (b) show the

measured and simulated S-parameters of the three-section pow-

er divider. The insertion losses of |S21| and |S31| were -3.20 dB,

and the isolation of |S32| was greater than 22 dB. The input

return loss of |S11| was better than -19 dB, and output return

losses of |S22| and, |S33| were better than -24 dB at center fre-

quency of 2 GHz. The measured -15 dB bandwidth of |S11|,

|S22|, |S33|, and |S32| were in the ranges of 1.1–3.42 GHz, 0.54–

3.76 GHz, 0.52–3.76 GHz, and 0.95–3.29 GHz, featuring

(a)

(b)

Fig. 6. Photographs of two-section (a) and three-section (b) pro-

posed power dividers.

(a)

(b)

Fig. 7. Measured and simulated S-parameters of the two-section

power divider: (a) |S21|, |S31|, |S32|, and (b) |S11|, |S22|, |S33|.

(a)

(b)

Fig. 8. Measured and simulated S-parameters of the three-section

power divider: (a) |S21|, |S31|, |S32|, and (b) |S11|, |S22|, |S33|.

YOON and KIM: MINIATURIZED MULTI-SECTION POWER DIVIDER WITH PARALLEL RC ISOLATION CIRCUIT

151

Fig. 9. Measured phase difference between the output ports of the

two- and three-section power dividers.

Table 1. Comparison of the proposed divider with conventional

multi-section divider

Ref.

Center

freq.

(GHz)

Section

no.

Total

length (º)

15 dB

BW (%)

Insertion

loss (dB)

[2] 30 2 180 100 -5.0

[3] 1 3/4 270 / 360 110 -2.2 / -5.5

[4] 1 3 183 55 -0.5 / -10

[5] 0.8 / 2 2 90 40 -2.06 / -4.88

This work

2 2 106 156 max. -3.1

3 179 162 max. -3.2

fractional bandwidths of 116%, 161%, 162%, and 117%. Fig. 9

shows that the phase difference between the output ports in the

two- and three-section power dividers was measured within

±2.5°. Table 1 shows a comparison of the proposed divider

with conventional multi-section divider.

IV. CONCLUSION

This paper presents a multi-section power divider with a par-

allel RC isolation circuit. Because each section of the power

divider is less than λ/4 wavelength, the power divider is imple-

mented in a reduced size. In addition, the bandwidth of the

matching and isolation characteristic showed that a fractional

bandwidth above 100% is achievable in the three-section divider.

By choosing the proper source and load impedance, the power

divider can be implemented while avoiding the use of high im-

pedance. By adopting decomposition and the improved trans-

former, the proposed power divider was demonstrated good

performance, and the simulated and measured results were con-

firmed to be in good agreement.

REFERENCES

[1] J. S. Lim, U. H. Park, S. Oh, J. J. Koo, Y. C. Jeong, and D.

Ahn, "A 800- to 3,200-MHz wideband CPW balun using

multistage Wilkinson structure," in Proceedings of 2006

IEEE MTT-S International Microwave Symposium Digest,

San Francisco, CA, 2006, pp. 1141-1144.

[2] Y. Sun and A. P. Freundorfer, "Broadband folded Wil-

kinson power combiner/splitter," IEEE Microwave and

Wireless Components Letters, vol. 14, no. 6, pp. 295-297,

2004.

[3] M. M. Honari, L. Mirzavand, R. Mirzavand, A. Abdipour,

and P. Mousavi, "Theoretical design of broadband multisec-

tion Wilkinson power dividers with arbitrary power split ra-

tio," IEEE Transactions on Components, Packaging and Man-

ufacturing Technology, vol. 6, no. 4, pp. 605-612, 2016.

[4] M. M. Honari, R. Mirzavand, P. Mousavi, and A. Ab-

dipour, "Class of miniaturised/arbitrary power division ratio

couplers with improved design flexibility," IET Microwaves,

Antennas & Propagation, vol. 9, no. 10, pp. 1066-1073, 2015.

[5] I. Sakagami, X. Wang, K. Takahashi, and S. Okamura,

"Generalized two-way two-section dual-band Wilkinson

power divider with two absorption resistors and its minia-

turization," IEEE Transactions on Microwave Theory and

Techniques, vol. 59, no. 11, pp. 2833-2847, 2011.

[6] P. H. Deng and Y. T. Chen, "New Wilkinson power divid-

ers and their integration applications to four-way and filter-

ing dividers," IEEE Transactions on Components, Packaging

and Manufacturing Technology, vol. 4, no. 11, pp. 1828-1837,

2014.

[7] Z. X. Du, X. Y. Zhang, K. X. Wang, H. L. Kao, X. L.

Zhao, and X. H. Li, "Unequal Wilkinson power divider

with reduced arm length for size miniaturization," IEEE

Transactions on Components, Packaging and Manufacturing

Technology, vol. 6, no. 2, pp. 282-289, 2016.

[8] W. Choe and J. Jeong, "Compact modified Wilkinson pow-

er divider with physical output port isolation," IEEE Mi-

crowave and Wireless Components Letters, vol. 24, no. 12, pp.

845-847, 2014.

[9] Y. Kim and Y. C. Yoon, "A modified unequal power divider

with a complex isolation component for enhanced isola-

tion," Microwave and Optical Technology Letters, vol. 57, no.

2, pp. 322-324, 2015.

[10] M. Heydari and S. Roshani, "Miniaturised unequal Wil-

kinson power divider using lumped component elements,"

Electronics Letters, vol. 53, no. 16, pp. 1117-1119, 2017.

[11] T. Zhang, W. Che, and H. Chen, "Miniaturized multiway

unequal power divider with controllable characteristic im-

pedances," IEEE Microwave and Wireless Components Let-

ters, vol. 27, no. 12, pp. 1062-1064, 2017.

[12] H. R. Ahn, "Complex impedance transformers consisting

of only transmission-line sections," IEEE Transactions on

Microwave Theory and Techniques, vol. 60, no. 7, pp. 2073-

2084, 2012.


152

[13] C. Li and D. S. Ricketts, "A low-loss, impedance matched

λ/4 compact T-junction power combiner," in Proceedings

Young-Chul Yoon received his B.S., M.S., and Ph.D. in electronics

engineering from Sogang University, Seoul, Korea,

in 1978, 1982, and 1989, respectively. In 1987, he

joined the Department of Electronics Engineering,

Catholic Kwandong University, Gangneung, Korea,

where he is currently a professor. His areas of interest

are the design of high power amplifiers for the ISM

band, and RF and microwave circuit analysis and

design.

of 2012 7th European Microwave Integrated Circuit Conference,

Amsterdam, The Netherlands, 2012, pp. 147-150.

Young Kim received his B.S., M.S., and Ph.D. in electronics

engineering from Sogang University, Seoul, Korea,

in 1986, 1988, and 2002, respectively. He developed

cellular and PCS linear power amplifiers at Samsung

Electronics Co. Ltd. In 2003, he joined the School

of Electronics Engineering, Kumoh National Insti-

tute of Technology, Gumi, Korea, where he is cur-

rently a professor. His areas of interest are the design

of high power amplifiers and linearization techniques, and RF and micro-

wave circuit analysis and design.




153

I. INTRODUCTION

Circularly polarized (CP) antennas are an attractive choice for

wireless communication because of their reduced polarization

loss factor and multi-path interference [1, 2]. Circular polariza-

tion is also a desirable feature in mobile devices for global posi-

tioning system applications [3, 4]. Moreover, we proposed a

wideband CP monopole slot antenna using a square-shaped

ground plane [5]. Electrically, small antennas of mobile devices

suffer from several inherent problems, such as small bandwidth,

high Q-factor, and low efficiency at smaller sizes [6, 7]. There-

fore, small antennas are often employed to excite the large

ground plane of a mobile device for effective radiation.

A loop-type ground radiation antenna (GradiAnt) showed

good radiation performance in various studies [8–10]. The an-

tenna’s performance is attributed to the antenna’s strong cou-

pling with the ground plane. Enhancing the radiation efficiency

of mobile antennas is a challenging task for antenna engineers.

To enhance the radiation efficiency of the antennas of a mobile

device that has an electrically small ground plane, ground mode

tuning (GMT) is an effective technique [11, 12]. The GMT

structures are installed at the edge of the ground plane to match

the resonance frequency of the ground mode with the operating

frequency of the antenna [13]. The theory of characteristic

modes can be utilized in a systematic design of antennas. Appli-

cation of the theory to the design of CP antennas has been pro-

posed in [14]. The loop-type GradiAnts proposed in the litera-

ture are linearly polarized antennas. However, GradiAnts with

circular polarization are not well documented in the literature.

In this paper, we propose a CP loop-type GradiAnt for In-

Circularly Polarized Loop-Type Ground Radiation

Antenna for IoT Applications Zeeshan Zahid1 · Longyue Qu2 · Hyung-Hoon Kim3 · Hyeongdong Kim2,*

Abstract

A circularly polarized loop-type ground radiation antenna using a ground mode tuning (GMT) structure is proposed for Internet of

Things (IoT) devices. The antenna is designed to excite two orthogonal modes of equal magnitude on the ground plane. The GMT

structure consists of an inductor and a metallic strip that has been installed at the edge of the ground plane to obtain a 90° phase shift

between the two modes. The proposed antenna generates left-hand circularly polarized waves in the +z-direction and right-hand circu-

larly polarized waves in the -z-direction. The antenna was fabricated to validate the simulation results. The measured -6 dB bandwidth of

the antenna was 150 MHz and the axial ratio bandwidth with reference to 3 dB was 130 MHz, completely covering the 2.4–2.48 GHz

band.

Key Words: Axial Ratio, Characteristic Modes, Circular Polarization, Reflection Coefficient.

Manuscript received November 28, 2018 ; Revised March 07, 2019 ; Accepted April 13, 2019. (ID No. 20181128-083J) 1Department of Electrical Engineering, Military College of Signals, National University of Sciences and Technology, Rawalpindi, Pakistan. 2Department of Electronics and Computer Engineering, Hanyang University, Seoul, Korea. 3Department of Cosmetic Science, Kwangju Women's University, Gwangju, Korea. *Corresponding Author: Hyeongdong Kim (e-mail: [email protected])





154

ternet of Things (IoT) devices with compact-sized ground plane,

operating at 2.45 GHz. To generate CP waves, two orthogonal

current modes of the ground plane need to be excited by the

antenna with a phase difference of 90° [15]. Implementing these

conditions on a mobile device antenna while maintaining good

radiation efficiency is a challenging task. In the antenna design

process, we have employed the characteristic mode analysis of

the proposed ground plane. Two orthogonally polarized modes

with equal magnitude have been excited by designing the an-

tenna at the corner of the ground plane. The GMT structure is

utilized to tune one of the modes so that a phase difference of

90° is achieved between the two modes. The mode can be fine-

tuned by using different values of inductor.

II. ANTENNA DESIGN

The loop-type GradiAnt acts as a magnetic coupler, and its

coupling with the dominant ground mode is maximum when it

is located at the maximum current location of the ground mode.

Therefore, the optimum location of the antenna is at the middle

of the edge of the ground plane. At this location, however, the

antenna excites only one mode. To simultaneously excite two

orthogonal modes on the ground plane with equal magnitude,

the antenna should be located at the corner of the ground plane.

At this location, the electric field of both modes is strongest;

therefore, the electric coupling between the antenna and the

ground modes should be enhanced. This can be expressed as

[16]:

i ge 2 2

o g

1J E d

(1)

where Ji is the impressed electric current density, Eg is the elec-

tric field of the ground plane, ωo is the operating frequency of

the antenna, and ωg is the resonance frequency of the ground

mode. The integration is carried over the volume of the antenna.

Furthermore, the coupling between the GradiAnt and the

ground mode can be enhanced by the optimum impedance level

of the antenna’s resonance loop. The impedance level is ex-

pressed as (Lr/Cr)1/2, where Lr is the inductance of the resonance

loop and Cr is the resonance capacitor. The optimum imped-

ance level depends on the antenna’s location on the ground

plane and the size of the antenna’s resonance loop. If the anten-

na is located at the middle of the ground plane, the impedance

level should be minimum, whereas if the antenna is located at

the edge of the ground plane, the impedance level should be

maximum [17]. The antenna’s impedance level can be enhanced

by increasing the clearance area of the antenna and using a low-

er value of Cr. The impedance matching can be controlled by

the feeding loop that contains Cf [18]. The impedance level of

the feeding loop (Lf/Cf)1/2 must be 50 Ω to match with the RF

(a)

(b)

Fig. 1. Geometry of the proposed antenna with front view (a) and

3D view (b).

source. The operating frequency of the antenna is also deter-

mined by Lr and Cr. The geometry of the proposed antenna is

shown in Fig. 1.

The antenna element is located at the left corner of the 45

mm × 45 mm ground plane by etching a square clearance of 7

mm × 7 mm in size. FR4 material (εr = 4.4, tanδ = 0.02) with

1 mm thickness is used as a substrate material. The resonance

capacitor (Cr) is located at the corner of the outer loop, called

the resonance loop. The feeding loop is 4.5 mm × 4.5 mm in

size and contains the feeding capacitor Cf. The GMT structure

is placed at the bottom of the ground plane and consists of an

inductor (L), a metallic strip 45 mm × 1 mm in size, and a

shorting strip. The strip is oriented along the xz-plane, and the

gap between the metallic strip and the ground plane is 2 mm.

The strip is connected with the ground plane through the short-

ing strip and the inductor. The width of the shorting strip is 2

mm.

III. SIMULATION RESULTS

The resonance capacitor’s location on the resonance loop

plays a critical role in the excitation of orthogonal modes on the

ground [19]. When a low-valued Cr is located at P1 on the res-

ZAHID et al.: CIRCULARLY POLARIZED LOOP-TYPE GROUND RADIATION ANTENNA FOR IoT APPLICATIONS

155

Fig. 2. Simulated current density of the antenna without the GMT

structure.

onance loop, high reactance of the capacitor can be modeled as

an open circuit; thereby the GradiAnt acts as a monopole an-

tenna attached to the location P2. Therefore, according to Eq.

(1), the antenna excites the current along the x-axis (mode 1) on

the ground plane. Similarly, the antenna excites the current

along the y-axis (mode 2) on the ground plane when Cr is locat-

ed at P2. Placing Cr at the corner of the resonance loop virtually

divides the antenna into two monopoles attached at P1 and P2,

respectively, resulting in the excitation of both modes with equal

magnitude. Fig. 2 shows the simulated surface current density

on the ground plane without the GMT structure. The simulat-

ed values of Cr and Cf were 0.15 pF and 0.12 pF, respectively,

where the values have been optimized through full-wave simu-

lations. It can be observed that the current is excited diagonally

on the ground plane, which is the resultant direction of mode 1

and mode 2, demonstrating that both modes have been excited

with equal magnitude.

To achieve circular polarization, the first two modes of the

ground plane have been utilized, where ωg of both modes is

greater than ω0. The GMT structure is employed to decrease

the resonance frequency of mode 2, so that a phase difference of

90° is achieved between modes 1 and 2 while maintaining equal

magnitude in both modes. The resonance frequencies of the

ground modes depend on the size of the ground plane; therefore,

the antenna performance depends critically on the ground plane

size. Simulations have been conducted to observe the effect of

the size of the ground plane on antenna performance where the

sizes of the antenna and the GMT strip have been unchanged.

The observations are shown in Table 1. The data demonstrate

Table 1. Effect of ground size on antenna performance

Ground

size (mm2)

fg

(GHz)

f0

(GHz)

L

(nH)

Bandwidth (MHz)

Matching Axial ratio

40 × 40 4.24 2.47 2 90 60

45 × 45 3.75 2.45 1.5 100 90

50 × 50 3.37 2.43 1 130 110

that the increase in the ground plane size decreases the reso-

nance frequency of both ground modes (fg), as well as the oper-

ating frequency of the antenna (fo). The tabulated fg is without

GMT structure. The decrease in fg is more significant as com-

pared to fo. The operating frequency can be tuned using Cr. Fur-

thermore, it is observed that the antenna’s matching bandwidth

(-6 dB ref.) increases with the increase in the ground size. Ac-

cording to Eq. (1), the coupling between the antenna and the

ground mode increases if ωg approaches ωo. The higher cou-

pling results in the increased matching bandwidth of the anten-

na. To decrease the resonance frequency of mode 2 of the

ground plane, inductor (L) has been utilized, which appears in

series with the ground mode. Therefore, increasing L decreases

the resonance frequency of mode 2. Table 1 indicates that the

lower values of L are used to achieve circular polarization with

the increase in the ground size. Moreover, the axial ratio (AR)

bandwidth increases with the increase in the ground size.

To validate the polarization of the antenna, the simulated

surface current density of the proposed structure is presented in

Fig. 3. Simulations show that the excited currents rotate in

clockwise direction on the ground plane. Fig. 3(a) shows that

the current density on the major portion of the ground plane is

directed along the x-axis at a phase of 0°, i.e., mode 1 is domi-

(a)

(b)

Fig. 3. Simulated surface current density at 2.45 GHz at the phase

of 0° (a) and at the phase of 90° (b).


156

nant.

Similarly, Fig. 3(b) shows that, at the phase of 90°, the cur-

rent is directed along the y-axis, i.e., mode 2 is dominant. It can

be observed that the current density is stronger around the Gra-

diAnt, showing that the antenna acts as an excitation element

for the ground plane. Although the magnitude of current densi-

ty around the antenna is strong, the current distribution on the

ground plane produces effective radiation. Therefore, the time

phase difference between both ground modes generates left-

hand circularly polarized (LHCP) waves along the +z-axis and

right-hand circularly polarized (RHCP) waves along the -z-axis.

IV. EXPERIMENTAL RESULTS AND DISCUSSION

The antenna was fabricated for experimental validation, as

shown in Fig. 4. The reflection coefficient was measured using

Agilent 8753ES vector network analyzer, and the radiation

characteristics were measured using a 3D CTIA-OTA chamber.

The measured and simulated reflection coefficients are present-

ed in Fig. 5.

The simulated bandwidth of the antenna with reference to -6

dB was 100 MHz (2.4–2.5 GHz), whereas the measured band-

width was 150 MHz (2.37–2.52 GHz), showing good agree-

ment with the simulated result. The measured and simulated

ARs in the direction of +z-axis are plotted in Fig. 6 along with

measured total efficiency. The measured AR bandwidth with

Fig. 4. Fabricated antenna.

Fig. 5. Measured and simulated reflection coefficient of the anten-

na.

Fig. 6. Measured axial ratio and total efficiency of the antenna.

reference to 3 dB was 130 MHz (2.38–2.51 GHz) and the sim-

ulated AR was 90 MHz (2.41–2.5 GHz). The minimum value

of measured AR was 1 dB at 2.44 GHz. The average efficiency

of the antenna in the operating band is 65%, and the maximum

efficiency of the antenna is 74% at 2.44 GHz. The efficiency of

the antenna is suitable for wireless applications. Fig. 7 presents

measured LHCP, RHCP, and peak gains as functions of fre-

quency. In accordance with the measured efficiency, the maxi-

mum value of peak gain (1.66 dB) occurs at 2.45 GHz and de-

creases away from it. Measured LHCP and RHCP gains at the

operating frequency are -0.36 dB and 0.64 dB, respectively. The

normalized, simulated, and measured radiation patterns of the

antenna at 2.45 GHz are displayed in Figs. 8 and 9, respectively.

The RHCP and LHCP data in the xz- and the yz-planes are

plotted in Fig. 8. The difference between simulated and meas-

ured results is mainly due to the feeding cable and the fabrica-

tion tolerance of the GMT structure. As shown, the higher val-

ues of LHCP and RHCP gain patterns occur in the upper and

lower hemispheres, respectively, confirming that the antenna

produces LHCP and RHCP waves along the +z- and -z-axes,

respectively. In Fig. 8(a), it can be observed that the LHCP and

RHCP patterns in the xz-plane are symmetric, whereas in the

yz-plane the patterns are tilted towards -30° and -150°, respec-

tively. The asymmetry is due to the asymmetric geometry of the

Fig. 7. Measured gains of the antenna as function of frequency.

ZAHID et al.: CIRCULARLY POLARIZED LOOP-TYPE GROUND RADIATION ANTENNA FOR IoT APPLICATIONS

157

(a) (b)

Fig. 8. LHCP and RHCP gains of the antenna: (a) simulated and

(b) measured.

(a) (b)

Fig. 9. Total gain of the antenna: (a) simulated and (b) measured.

proposed antenna where the GMT structure is installed in the

yz-plane, whereas in the xz-plane there is no GMT structure.

Moreover, as shown in Fig. 3, linear currents are excited in the

GMT structure, causing the patterns to tilt in the yz-plane. The

measured cross polarization levels of the antenna in +z and –z

axes were below -25 dB. The simulated total gain radiation pat-

tern is shown in Fig. 9(a). The pattern is isotropic in the xz-

plane, whereas in the xy-plane, the minimum value of the pat-

tern is -10.5 dBi, which occurs at 90°. Therefore, the antenna

has a quasi-isotropic radiation pattern, suitable for wireless ap-

plications as compared to that of linearly polarized antennas

having nulls in the radiation patterns [19]. A good agreement

can be observed in the measured and simulated gain patterns in

the xz- and yz-planes. The minimum value of the measured

gain was -15.87 dBi, in the yz-plane. The agreement of the

measured and simulated results verifies that the proposed an-

tenna is suitable for IoT applications.

V. CONCLUSION

We proposed a CP loop-type ground radiation antenna using

a ground mode tuning structure. The conditions of circular po-

larization were achieved successfully using the antenna design

and ground mode tuning structure. The simulated and meas-

ured data showed good agreement. The antenna generated

LHCP waves in the +z-axis and RHCP waves in the -z-axis

with cross polarization levels less than -25 dB. The total gain

radiation pattern of the antenna was quasi-isotropic, which is an

attractive feature for internet of things applications. The match-

ing bandwidth of the antenna was 150 MHz and the axial ratio

bandwidth was 130 MHz, covering the complete 2.4–2.48

GHz band. The proposed technique is versatile and can be ap-

plied to other wireless applications as well.

REFERENCES

[1] M. A. Kossel, R. Kung, H. Benedickter, and W. Biichtokd,

"An active tagging system using circular-polarization modu-

lation," IEEE Transactions on Microwave Theory and Tech-

niques, vol. 47, no. 12, pp. 2242-2248, 1999.

[2] R. Garg, P. Bhartia, I. Bahl, and A. Ittipiboon, Microstrip

Antenna Design Handbook. Norwood, MA: Artech House,

2001.

[3] W. J. Liao, J. T. Yeh, and S. H. Chang, "Circularly polar-

ized chip antenna design for GPS reception on hand-

sets," IEEE Transactions on Antennas and Propagation, vol.

62, no. 7, pp. 3482-3489, 2014.

[4] S. H. Chang and W. J. Liao, "A novel dual band circularly

polarized GNSS antenna for handheld devices," EEE Tr-

ansactions on Antennas and Propagation, vol. 61, no. 2, pp.

555-562, 2013.

[5] P. Mousavi, B. Miners, and O. Basir, "Wideband L-shaped

circular polarized monopole slot antenna," IEEE Antennas

and Wireless Propagation Letters, vol. 9, pp. 822-825, 2010.

[6] H. Wheeler, "Small antennas," IEEE Transactions on An-

tennas and Propagation, vol. 23, no. 4, pp. 462-469, 1975.

[7] R. F. Harrington, "Effect of antenna size on gain, band-

width, and efficiency," Journal of Research of the National Bu-

reau of Standards D (Radio Propagation), vol. 64, no. 1, pp.

1-12, 1960.

[8] O. Cho, H. Choi, and H. Kim, "Loop-type ground antenna

using capacitor," Electronics Letters, vol. 47, no. 1, pp. 11-12,

2011.

[9] Y. Liu, X. Lu, H. Jang, H. Choi, K. Jung, and H. Kim,

"Loop-type ground antenna using resonated loop feeding,

intended for mobile devices," Electronics Letters, vol. 47, no.

7, pp. 426-427, 2011.

[10] H. C. Choi, J. Lee, O. Cho, H. Lee, and B. Park, "Gr-

ound radiation antenna," U.S. Patent 8604998, 2013.

[11] L. Qu, R. Zhang, H. Shin, J. Kim, and H. Kim, "Perfor-

mance enhancement of ground radiation antenna for Z-

wave applications using tunable metal loads," Electronics

Letters, vol. 52, no. 22, pp. 1827-1828, 2016.

[12] L. Qu, R. Zhang, H. Shin, J. Kim, and H. Kim, "Mode-

controlled wideband slot-fed ground radiation antenna uti-

lizing metal loads for mobile applications," IEEE Transac-

tions on Antennas and Propagation, vol. 65, no. 2, pp. 867-

872, 2017.


158

[13] Z. Zahid and H. Kim, "Enhancement technique for loop

type ground radiation antenna using ground mode tuning

structure," Microwave and Optical Technology Letters, vol.

59, no. 2, pp. 476-478, 2017.

[14] M. Cabedo-Fabres, E. Antonino-Daviu, A. Valero-

Nogueira, and M. F. Bataller, "The theory of characteristic

modes revisited: a contribution to the design of antennas

for modern applications," IEEE Antennas and Propagation

Magazine, vol. 49, no. 5, pp. 52-68, 2007.

[15] C. A. Balanis, Antenna Theory: Analysis and Design, 3rd ed.

Hoboken, NJ: John Wiley & Sons, 2005.

[16] R. F. Harrington, Time-Harmonic Electromagnetic Fields.

Zeeshan Zahid received his M.S. in Electronics from Quaid-i-Azam

University, Islamabad, in 2006. He joined National

University of Sciences and Technology (NUST) as a

lecturer. He received Best Teacher of the Year award

in 2012. He earned his Ph.D. in electronics and

computer engineering from Hanyang University,

Korea, in 2017. He is currently serving as an assistant

professor at Military College of Signals, NUST. His

research interests include MIMO antennas and circularly polarized anten-

nas for mobile devices.

Longyue Qu received his B.Sc. degree in communication engi-

neering from Yanbian University, China, in 2013,

and his M.Sc. and Ph.D. degrees in microwave en-

gineering from Hanyang University, Seoul, Korea, in

2015 and 2018, respectively. He is currently a re-

search fellow at Hanyang University. His current

research interests include antenna theory and design

especially for 4G/5G communications, massive

MIMO, metamaterials, mmWave, and RF circuits. He serves as a reviewer

for several international journals, such as IEEE Transactions on Antennas

and Propagation, IEEE Antennas and Wireless Propagation Letters, and IEEE

Access.

New York, NY: Wiley-Interscience, 2001.

[17] Z. Zahid and H. Kim, "Performance evaluation of loop-

type ground radiation antenna based on its optimum im-

pedance level," Electronics Letters, vol. 53, no. 7, pp. 446-

448, 2017.

[18] Z. Zahid and H. Kim, "Analysis of a loop type ground

radiation antenna based on equivalent circuit model," IET

Microwaves, Antennas & Propagation, vol. 11, no. 1, pp.

23-28, 2017.

[19] R. Zhang, Y. Liu, and H. Kim, "Effect of capacitor posi-

tion on radiation patterns of ground radiation antenna,"

Electronics Letters, vol. 51, no. 23, pp. 1844-1846, 2015.

Hyung-Hoon Kim received his B.S. degree from Chonnam National

University in 1986, M.S. from Korea Advanced

Institute of Science and Technology in 1988, and his

Ph.D. from Hanyang University in 1997. Since 1994,

he has served as professor in the Department of

Cosmetic Science at Kwangju Women's University,

Korea. His research interest is electromagnetic nu-

merical analysis.

Hyeongdong Kim received his B.S. and M.S. degrees from the Seoul

National University, Seoul, Korea, in 1984 and 1986,

respectively, and his Ph.D. degree from the Univer-

sity of Texas at Austin in 1992. From May 1992 to

February 1993, he was a post-doctoral fellow at the

University of Texas at Austin. In 1993, he worked as

a professor at the Department of Electrical and

Computer Engineering, Hanyang University, Seoul,

Korea. His current research interests are various antenna theories and de-

signs based on ground characteristic mode analysis, namely, wideband,

high-efficiency, circular polarization, MIMO, and high-sensitivity anten-

nas.




159

I. INTRODUCTION

Synthetic aperture radar (SAR) imaging technology is one of

the most promising technologies utilized in various observation

areas such as terrain, resource, and target searches. With the

increasing demand for imaging technology, SAR systems have

been developed to achieve high resolution and wide swath for

accurate target detection, regardless of weather or environmen-

tal changes. To obtain high-resolution images, the SAR system

can be improved in two ways. The first is a budget parameter

optimization of the transmission and reception systems to re-

duce the ambiguity of radar signals and various interference

problems. The second is to use diverse SAR signal processing

algorithms for high-resolution image extraction. However, con-

sidering the cost and physical volume of the SAR system, an

effective way to obtain high-resolution images is with diverse

signal processing techniques.

There are numerous signal processing algorithms such as the

range Doppler algorithm (RDA), chirp scaling algorithm

(CSA), omega-K algorithm (ω-k), back projection algorithm

(BPA). The RDA is widely used for efficient and stable signal

processing because it is simple and easy to use. However, when

the RDA is used for image processing, the final image is not

focused due to random environmental variables. To cope with

environmental variables and acquire high-resolution images,

precise estimations of the Doppler centroid frequency and chirp

rate are required in the algorithm. Many researchers have pro-

posed algorithms to correct the received signal by removing the

unexpected signals, using the magnitude or phase information

A Modified Stripmap SAR Processing for Vector Velocity

Compensation Using the Cross-Correlation

Estimation Method Chul-Ki Kim1 · Jung-Su Lee2 · Jang-Soo Chae2 · Seong-Ook Park1,*

Abstract

This paper proposes a modified stripmap synthetic aperture radar (SAR) signal processing algorithm with an X-band SAR system, using

the practical measurement results of automobile-based SAR (Auto-SAR). The quality of the image is degraded due to an unexpected

direction change or anisotropy of the target position in radar image measurement. To solve the quality problem, signal processing is re-

quired for the vector velocity of each range in the azimuth direction. An X-band chirp pulse system was implemented and optimized by a

signal processing algorithm suitable for high resolution. The stripmap SAR images are produced in five places. The validity of the pro-

posed algorithm is verified by comparing impulse response function analysis and experimental images.

Key Words: Auto-SAR, Compensation Processing, Stripmap SAR, Vector Velocity, X-Band System.

Manuscript received October 11, 2018 ; Revised December 15, 2018 ; Accepted May 07, 2019. (ID No. 20181011-071J) 1School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Korea. 2Satellite Technology Research Center, Korea Advanced Institute of Science and Technology, Daejeon, Korea. *Corresponding Author: Seong-Ook Park (e-mail: [email protected])





160

of the chirp signal based on average cross-correlation coefficient

(ACCC), multi-look cross correlation (MLCC), and multi-

look beat frequency (MLBF) methods [1, 2]. However, the

Doppler parameters extracted through previous algorithms still

have uncertainty because of topographic change of interest area

and interference of unexpected signal. Therefore, many re-

searchers have attempted to increase the accuracy of Doppler

parameters in various SAR experiments [3–6]. This paper pro-

poses a novel method to solve the uncertainty by considering

uncorrected Doppler parameters. In conventional SAR experi-

ments, signal processing was performed by estimating the con-

stant velocity with the lowest ambiguity error. This convention-

al method can obtain a reasonable image quality over the entire

area if the area of interest has a nearly constant height over

whole region. However, if the difference of altitude in the area

and the direction of measurement is changeable, a high-quality

SAR image cannot be obtained with only a constant velocity.

Based on the changes of velocity in the area of interest, the

quality of a SAR image can be improved using cross-correlation

theory, designed for SAR. The estimated velocity is used to

acquire a high resolution SAR image without complicated pro-

cessing techniques. To evaluate the proposed method, the raw

data obtained from practical experiments are used. Finally,

high-resolution SAR images are constructed, applying the pro-

posed algorithm to various environmental changes of automo-

bile-based SAR (Auto-SAR).

II. SAR SIGNAL PROCESSING ANALYSIS OF THE RDA

The RDA is a method of compressing raw data using a

matched filter in the range direction and azimuth direction, re-

spectively. As shown in Fig. 1, the RDA is typically used in

stripmap standard mode with chirp pulse signals. The basic sig-

nal processing procedure consists of range and azimuth com-

pressions and range cell migration, as shown in Fig. 2. Doppler

centroid estimation can be applied to obtain high-resolution

images in the SAR ground experiment. In case of the ground

experiment, the image difference caused by the Doppler cen-

troid variation is insignificant [1]. Therefore, it is possible to

exclude the Doppler centroid estimation in a mathematical

mechanism for SAR signal processing. The chirp pulse signals

in the range and azimuth directions received from the SAR sys-

tem are as follows, respectively.

𝑆 𝜏, 𝜂 𝐴 𝜔 𝜏 𝑒 𝑒

(1)

𝑆 𝜏, 𝜂 𝐴 𝜔 𝜂 𝜂 𝑒 𝑒 (2)

𝜏 and 𝜂 are the time, 𝜔 and 𝜔 are the window func-

tions each chirp pulse signal, 𝐴 and 𝐴 are the amplitude

Fig. 1. Stripmap SAR mode for chirp pulse radar.

Fig. 2. Conventional RDA.

constant values, 𝑓 and 𝑓 are the center frequency of chip

pulse signal, and 𝐾 and 𝐾 are the chirp rate of the signals.

The first and second parameters mentioned above are for the

range and azimuth direction, respectively. 𝑅 𝜂 in (1) is the

instantaneous slant range. Eqs. (1) and (2) are approximated as

(3) through the range and azimuth compressions and range cell

migration. After the range cell migration of SAR data, the per-

formance of resolution for the matched filter in the azimuth

direction can be determined according to 𝐾 and 𝑓 . The

final SAR image signal, 𝑆 𝜏, 𝜂 , is as follows.

𝑆 𝜏, 𝜂 𝐴 𝑝 𝜏2𝑅𝑐

𝑝 𝜂 𝑒 𝑒

(3)

Therefore, it is more important to estimate the exact matched

KIM et al.: A MODIFIED STRIPMAP SAR PROCESSING FOR VECTOR VELOCITY COMPENSATION USING THE CROSS-CORRELATION ESTIMATION …

161

filter in the azimuth direction than the range direction because

the errors of 𝐾 and 𝑓 are vulnerable by the SAR measure-

ment environment, such as the actual speed, changing rate in the

distance to the target, and motion vibrating conditions. However,

the matched filter in the range direction has a low error rate

because the raw data are compressed using the system parame-

ters or loop-back chirp pulse signal. Therefore, our proposed

algorithm estimates the accurate velocity in each range, which is

an important parameter for the azimuth matched filter.

III. PROPOSED SAR PROCESSING ALGORITHM

Based on the RDA, the modified signal process of the Dop-

pler parameter is proposed in this section. To estimate the azi-

muth matched filter that fits with the actual signal, the exact

Doppler centroid, 𝑓 𝜏 and Doppler chirp rate, 𝐾 𝜏 are

required under variable terrain conditions [7–9]. The Doppler

parameters are as follows.

𝑓 𝜏 (4)

𝐾 𝜏 ≅ (5)

𝜃 is the squint angle for measuring the target, 𝜆 is the

wavelength of the center frequency in the chirp pulse signal, and

𝑅 𝜏 is the slant range along to the range time. The Doppler

parameters are addressed as a function of the vehicle velocity, 𝑉 .

Therefore, the main reason for the uncorrected azimuth mat-

ched filter is the misestimated velocity. This velocity leads to the

low quality of the SAR image. The velocity estimation of each

range can lead to more the high-resolution quality and precise

target classification than the conventional results of the SAR

processing. To find the accurate velocity 𝑉 , the proposed algo-

rithm incorporates cross-correlation technique into the estima-

tion of the Doppler parameter [10, 11]. If 𝑆 𝜏, 𝜂 is a 2D sig-

nal constructed in the range and azimuth, the cross-correlation

function is defined as follows.

𝑠 ∗ 𝑠 𝜏 , 𝜂

≝ 𝑠∗ 𝜏 𝑠 𝜏 , 𝜂 𝜔 𝑑𝜔

(n = 1, 2, 3, 4, ⋯, maximum range cell)

(6)

𝜏 is the slow time and, 𝜂 is the fast time. Based on (6), the

chirp pulse frequency of the range-Doppler domain along the

range is divided into two looks (Look1, Look2), having 𝑉Δ𝑉 in phase information. A matched filter of assumed 𝑉

multiplies each look, and then inverse.

FFT is used to generate the images of each look with a phase

error. The cross-correlation values between two images are com-

pared until reaching the smallest correlation value, changing 𝑉 .

Range cell migration correction (RCMC) and azimuth com-

pression are performed, using the 𝑉 defined from finishing the

cross-correlation estimation in each range cell with phase vari-

ance compensation. In the conventional SAR process, each step

of RDA uses its own technology to compensate the phase error

for every range cell. It is time-consuming and complicated, re-

sulting in low efficiency. However, in the proposed technique, all

steps are performed at once without any other process using the

estimated 𝑉 . Unlike airborne- and spaceborne-SAR, which are

sensitive to environmental changes, the proposed algorithm is

suitable for Auto-SAR to make real-time processing fast and

easy, resulting in high efficiency.

(a)

(b)

(c)

Fig. 3. Information of the proposed estimation: (a) the method

from the raw data structure for vector velocity, (b) the chirp

pulse in the azimuth along to the range, and (c) the differ-

ence in the ideal IRF due to the velocity variation.


162

Fig. 4. Proposed RDA for high-resolution stripmap SAR.

From the estimated results, the defined 𝑉 can vary over a

range, such as a vector. Therefore, the 𝑉 is defined as the vector

velocity. The most suitable 𝑉 depends on the direction of the

target. Before the practical experiment, impulse response func-

tion (IRF) analyses using the proposed algorithm are performed

in Fig. 3. Fig. 3(a) and (b) show the analysis structure in 2D and

the Doppler chip pulse signal according to the range, respective-

ly. Fig. 3(c) shows that estimating the correct 𝑉 improves the

peak value and the 3-dB resolution of IRF of the target in the

SAR image. Fig. 4 shows over whole signal processing of the

proposed algorithm by organizing the process described in this

section.

Ⅳ. EXPERIMENT RESULTS AND PERFORMANCE

To evaluate the performance of the proposed algorithm, the

Table 1. X-band SAR system specifications

Parameter Value

Center frequency X-band

Frequency bandwidth (MHz) 300

Pulse repetition frequency (Hz) 1,000

Measurement velocity (km/hr) 80

Fig. 5. The picture of the SAR measurement setup in an auto-

mobile.

Fig. 6. The measurement area for verifying proposed SAR (loca-

tion of Scenes #1 and #2, Gyeo-nam bridge; Scene #3,

Chonchon bridge; Scenes #4 and #5, Magoksa IC).


163

(a) (b)

(c)

(d)

(e)

Fig. 7. Signal processing results by the proposed algorithm: (a)

Scene #1, (b) Scene #2, (c) Scene #3, (d) Scene #4, and (e)

Scene #5.

X-band SAR system was implemented and the practical exper-

iment was conducted. Table 1 shows the specifications of the X-

band SAR system.

As shown in Fig. 5, two antennas are installed as the trans-

mitter and receiver, respectively. The stop-and-go method is

applied to measure the practical SAR experiment. To prove the

validity of the proposed SAR algorithm, five different regions

are selected. In particular, two curved terrains are included in

the practical experiment regions. Fig. 6 shows the practical

measurement locations, which are taken while driving through

an express highway. Scenes #1 and #2 are places to drive

through curved loads, Scene #3 is a lower elevation terraced

region as the range increases along the distance. Scenes #4 and

#5 are rice fields measured in both directions at a height of 60

m. Applying vector velocity estimation, the quality of the SAR

image is significantly improved over conventional RDA in all

regions, as shown in Fig. 7. Fig. 8 shows the difference before

and after applying the vector velocity to Scenes #1 and #2, re-

spectively. As shown in the images in Fig. 8, we can see the per-

formance of the algorithm through the focusing improvement

in practical experiments. In addition, as confirmed in Figs. 7

and 8, the proposed technique can be applied to various terrain

and experimental conditions in Auto-SAR. In Fig. 9, analysis of

(a) (b)

(c) (d)

Fig. 8. The results of Scenes #1 and #2 when processed by the orig-

inal RDA (a), (c) and the proposed RDA (b), (d), respec-

tively.


164

(a)

(b)

Fig. 9. IRF analysis using the proposed algorithm: (a) the loca-

tion and image of corner reflector and (b) the impulse re-

sponse function of a corner reflector according to velocity.

the corner reflector in Scene #5 shows that the vector estimation

improves the 3-dB resolution of the IRF and increases the peak

energy of the main lobe. Through the estimated velocity is clos-

er to the accurate velocity, the energy balance of the sinc-

function is addressed more clearly. These results suggest that the

desired resolution and high-quality SAR image can be obtained

by the correction of the vector velocity.

Ⅴ. CONCLUSION

The proposed algorithm is suitable for measuring the image

in Auto-SAR. The recent technology has complex process to

compensate for the phase error in each range cell for airborne-

and spaceborne-SAR. It is inefficient in creating SAR images in

Auto-SAR. However, the proposed technology is easy to handle

and provides fast performance, allowing an immediate response

to a variety of situations that can occur on the ground. Applying

the precise estimation of the vector velocity can provide more

accurate signal processing for high-quality SAR images. To

prove the effectiveness of the proposed algorithm, high-quality

results are acquired from practical experiments. In addition, the

proposed technique of vector velocity is expected to be applied

in various conditions, such as sub-aperture correction, spotlight

mode, and autofocus.

This work was supported by Institute for information &

communications Technology Promotion (IITP) grant funded

by the Korea government (MSIT) (No. 2018-0-01658, Key

Technologies Development for Next Generation Satellites).

REFERENCES

[1] I. G. Cumming and F. H. Wong, Digital Processing of Syn-

thetic Aperture Radar Data. London: Artech House Inc.,

2005.

[2] J. C. Curlander and R. N. McDonough, Synthetic Aperture

Radar Systems and Signal Processing. New York, NY: John

Wiley & Sons Inc., 1991.

[3] B. Liu, C. Liu, and Y. Wang, "Slice convolution based slope

estimation for SAR Doppler ambiguity resolver," in Proceed-

ings of 2013 IEEE International Geoscience and Remote Sens-

ing Symposium, Melbourne, Australia, 2013, pp. 2055-2058.

[4] J. Yang, Y. Zhang, and X. Kang, "A Doppler ambiguity toler-

ated algorithm for airborne SAR ground moving target im-

aging and motion parameters estimation," IEEE Geoscience

and Remote Sensing Letters, vol. 12, no. 12, pp. 2398-2402,

2015.

[5] Q. Liu, Z. Wang, W. Xia, and B. Chen, "Ground moving

target positioning with Doppler ambiguity and scene ambi-

guity in SAR," in Proceedings of 2017 40th International Con-

ference on Telecommunications and Signal Processing (TSP),

Barcelona, Spain, 2017, pp. 465-469.

[6] W. Li, Y. Huang, J. Yang, and J. Wu, "Comparison of geome-

try-based Doppler ambiguity resolver in squint SAR," in

Proceedings of 2011 IEEE International Geoscience and Remote

Sensing Symposium, Vancouver, Canada, 2011, pp. 4443-4445.

[7] H. Zheng, J. Wang, X. Liu, Y. Gao, and L. Zhang, "Ve-

locity estimation of the moving target for high-resolution

wide-swath SAR systems," in Proceedings of 2017 IEEE In-

ternational Geoscience and Remote Sensing Symposium

(IGARSS), Fort Worth, TX, 2017, pp. 2054-2057.

[8] X. Liu, J. Xiong, Y. Huang, J. Wu, J. Yang, and W. Pu,

"SAR Doppler frequency rate estimation based on time-

frequency rate distribution," in Proceedings of 2009 2nd

Asian-Pacific Conference on Synthetic Aperture Radar, Xian,

China, 2009, pp. 926-930.

[9] A. Moreira, J. Mittermayer, and R. Scheiber, "A SAR auto-

focus technique based on azimuth scaling," in Proceedings of

1997 IEEE International Geoscience and Remote Sensing

Symposium Proceedings. Remote Sensing-A Scientific Vision for

Sustainable Development, Singapore, 1997, pp. 2028-2030.

[10] Y. Yue, X. Zhang, and S. Jun, "Motion compensation

method for translational variant bistatic SAR using auto-

correlation variance," in Proceedings of IET International Ra-

dar Conference, Guilin, China, 2009.

[11] K. Ouchi and S. I. Hwang, "Improvement of ship detection

accuracy by SAR multi-look cross-correlation technique us-

ing adaptive CFAR," in Proceedings of 2010 IEEE Interna-

tional Geoscience and Remote Sensing Symposium, Honolulu,

HI, 2010, pp. 3716-3719.


165

Chul-Ki Kim was born in Gangneung, Korea, in July 1989. He

obtained his B.S. degree in Electronic Engineering

from Soongsil University, Seoul, Korea, in 2014, his

M.S. degree in Electrical Engineering from Korea

Advanced Institute of Science and Technology,

Daejeon, in 2016, and is currently the candidate-

student, working toward a Ph.D. in Electrical En-

gineering at Korea Advanced Institute of Science

and Technology (KAIST). His current research interests include synthetic

aperture radar (SAR).

Jung-Su Lee was born in Busan, Korea in February 1978. He

obtained his B.S. degree in Electronic Engineering

from Korea University, Seoul, Korea in 2001 and his

M.S. degree in Radio Sciences and Engineering,

Korea University, Seoul, Korea in 2003. Since July

2003, he has been a Senior Research Engineer at the

Satellite Technology Research Center, Korea Ad-

vanced Institute of Science and Technology, Dae-

jeon, Korea. His research interests include satellite wireless communication

and ranging systems.

Jang-Soo Chae was born in Suncheon, Korea, in August 1959. He

obtained his B.S. degree from In Ha University,

Korea in 1987, his M.S. degree from Seoul National

University, Seoul, Korea in 1989, his Ph.D. from

POSTECH in 1992, and another Ph.D. from Ajou

University, Suwon, in 2004. From March 1990 to

August 1995, he was a Research Engineer with Ko-

rea Aerospace Research Institute, Daejeon, working

with LEO satellite system engineering, and is currently working at Satellite

Research Center in KAIST. He is engaging in research on the develop-

ment of passive antenna for synthetic aperture radar (SAR) onboard a small

satellite and dynamics and the control of flexible and space structure.

Seong-Ook Park was born in Kyungpook, Korea, in December 1964.

He obtained his B.S. degree from Kyungpook Na-

tional University, Korea, in 1987, his M.S. degree

from Korea Advanced Institute of Science and

Technology, Daejeon, Korea, in 1989, and his Ph.D.

degree from Arizona State University, Tempe, AZ,

in 1997, all in electrical engineering. From March

1989 to August 1993, he was a Research Engineer

with Korea Telecom, Daejeon, working with microwave systems and net-

works. He later joined the Telecommunication Research Center, Arizona

State University, until September 1997. Since October 1997, he has been

with the Information and Communications University, Daejeon, and cur-

rently is a Professor at the Korea Advanced Institute of Science and Tech-

nology. His research interests include mobile handset antenna and analyti-

cal and numerical techniques in the area of electromagnetics. Dr. Park is a

member of Phi Kappa Phi.




166

I. INTRODUCTION

Circularly polarized (CP) antennas are popular these days be-

cause of their superior performance over linearly polarized an-

tennas. Single-feed CP antennas are much preferred in wireless

communication because of their easy fabrication and low cost.

Many techniques have been used to convert linearly polarized

antennas to exhibit circular polarization behavior [1–5]. By in-

troducing strip lines into coplanar waveguide (CPW)-fed slot

antennas, a broadband CP antenna is achieved in [1–3]. In [4],

an inverted L-shaped asymmetric CPW-fed square slot antenna

with small strips is presented to achieve a wideband CP. A

strip-loaded monopole antenna is designed to obtain a wide-

band CP in [5]. As strip loading is sensitive to the design pro-

cess, achieving optimum wideband CP characteristics using a

stub or the strip-loading method is difficult. The gain of a slot

antenna is low because its radiation characteristic is bidirectional.

A perfect reflector or a backed cavity can be used to improve the

gain, but both have a limited axial ratio bandwidth [6–8].

Therefore, metasurfaces (MTSs) are the potential structures

that can be used to achieve a wide CP bandwidth and gain.

MTSs are periodic subwavelength metal or dielectric patches

used to improve the performance of primary source antennas

such as a patch or a slot. MTSs are used as reflectors to achieve

a wide CP bandwidth. Artificial magnetic conductors (AMCs),

high impedance surfaces, frequency selective surfaces (FSS), and

electromagnetic band gap (EBG) structures are some of the

MTSs used as reflectors to improve the gain, CP bandwidth,

and pattern purity. A simple linearly polarized dipole antenna is

utilized to generate a circularly polarized wave using an EBG

structure in [9, 10]. Different slot antennas with a single- and

dual-layer AMCs as reflectors are investigated in [11, 12]. A

loop antenna is designed with an AMC as a reflector to improve

the axial ratio bandwidth in [13]. These MTS-based antennas

have a low axial ratio bandwidth. An AMC or an EBG struc-

A Wideband Circularly Polarized Slot Antenna Backed

by a Frequency Selective Surface Puneeth Kumar Tharehalli Rajanna* · Karthik Rudramuni · Krishnamoorthy Kandasamy

Abstract

This paper presents the design of a coplanar waveguide (CPW)-fed single-layer wideband circularly polarized (CP) planar slot antenna

backed by a frequency selective surface (FSS). The planar slot antenna is fed with a stub-loaded modified CPW feed line to tune and

optimize the impedance bandwidth. The corners of the square slot antenna are perturbed to produce two orthogonal degenerate modes

required for a wideband CP operation. The FSS layer is placed under the slot antenna to increase the gain and axial ratio bandwidth. The

measured results of the proposed antenna provide an impedance bandwidth of 63.22% and an axial ratio bandwidth of 31.14% with a

peak gain of 4.87 dB. The proposed antenna has a simple geometry, a wide CP bandwidth, and a good gain.

Key Words: Circular Polarization, Frequency Selective Surface, Slot Antenna.

Manuscript received December 05, 2018 ; Revised February 07, 2019 ; Accepted May 07, 2019. (ID No. 20181205-084)

Department of Electronics and Communication, National Institute of Technology Karnataka, Mangalore, India. *Corresponding Author: Puneeth Kumar Tharehalli Rajanna (e-mail: [email protected])




RAJANNA et al.: A WIDEBAND CIRCULARLY POLARIZED SLOT ANTENNA BACKED BY A FREQUENCY SELECTIVE SURFACE

167

ture along with a slot antenna improves the gain and axial ratio

bandwidth. The careful design of a primary source antenna with

MTS as a reflector can achieve wideband CP characteristics.

Corner truncated microstrip patch antennas have been discussed

in the literature, but they have a narrow CP bandwidth [14]. A

truncated slot antenna loaded with a pair of split-ring resonators

(SRRs) is presented in [15] for dual-band CP applications. The

disadvantage is that the corner truncated slot antenna alone

cannot achieve a CP. The loading of an SRR on the truncated

slot antenna gives a CP in both bands. In our proposed design,

the perturbed slot antenna alone can produce a wideband CP by

increasing the depth of perturbation.

In this paper, a novel perturbed slot antenna is proposed to

achieve a wideband CP. Primarily, the square slot antenna is

designed to radiate at 2.5 GHz but with a linear polarization.

The corner of the slot antenna is truncated to provide two or-

thogonal degenerate modes to produce a wide CP bandwidth.

The truncated slot antenna achieves an axial ratio bandwidth of

20.68% at 4.35 GHz resonance, but the gain of the perturbed

slot antenna is 1.88 dB. Furthermore, an effort is made to in-

crease the axial ratio bandwidth and gain of the truncated planar

slot antenna by backing the FSS. A 6 × 10 array of the FSS is

used to improve the gain and axial ratio bandwidth of the per-

turbed slot antenna. The FSS improve the axial ratio bandwidth

up to 31.14% at 4.33 GHz resonance and a peak gain of up to

4.87 dB.

II. ANTENNA GEOMETRY AND ITS OPERATION

The proposed antenna consists of a single-layer CPW line-

fed truncated corner slot antenna and an FSS layer, as shown in

Fig. 1. A low-cost FR4 substrate with εr = 4.3 and loss tangent

Fig. 1. Geometry of the proposed FSS-based antenna.

(a)

(b)

Fig. 2. Simulated reflection coefficient and axial ratio of the per-

turbed slot antenna: (a) |S11| and (b) axial ratio.

tanδ = 0.025 is utilized for both slot antenna and the FSS. The

FSS backs the slot antenna with an air gap of ham = 22 mm. The

resonance frequency of the slot antenna is calculated using Eq.

(1) [15].

𝑓 , (1)

where c is the speed of light, and 𝑆 = (S + a) is the total length

of the square slot antenna. The proposed antenna parameters

are Wp = 7 mm, Lp = 13 mm, Wg = 70 mm, Wm = 3 mm, Lm =

29 mm, n = 0.1 mm, m = 0.2 mm, W = 89 mm, L = 100 mm,

hs = 1.6 mm, hf = 0.508 mm, a = 8.4 mm, L1 = 7 mm, L2 = 5

mm, and W1 = 1 mm.

The conventional slot is perturbed to obtain the optimized

performance of the axial ratio. The simulated reflection coeffi-

cient and axial ratio of the slot antenna with different truncated

values are shown in Fig. 2. The dimension of truncation is op-

timized to a = 8.4 mm, as the axial ratio value deteriorates with

the further increase in truncation. The electric field distribution

of the truncated slot antenna at different time intervals is shown

in Fig. 3. In the figure, the truncated slot antenna radiates left-

hand circularly polarized (LHCP) waves. The opposite corners

of another diagonal of the slot antenna should be truncated to

obtain right-hand circularly polarized (RHCP) waves. To fur-


168

(a)

(b)

Fig. 3. Simulated electric field distribution of the truncated slot

antenna at 3.47 GHz: (a) t = 0 and (b) t = T/4.

ther improve the gain and axial ratio bandwidth of the truncated

slot antenna, an FSS consisting of rectangular unit cells are used.

The proposed FSS acts as a polarization-dependent reflective

surface. To enhance the gain of the slot, the FSS is chosen to

reflect the signal from the slot antenna back to the forward di-

rection to improve the directivity of the antenna. The maximum

gain is obtained for the 6 × 10 array. The FSS has a size of W

× L and a rectangular cell size of Wp and Lp. The gap between

patches in the horizontal and vertical directions is m and n, re-

spectively.

The perturbed slot gives a CP bandwidth of 3.9–4.8 GHz.

The polarization-dependent reflective property of the FSS is

used to improve the axial ratio bandwidth. The reflection phase

curves for the x- and y-polarized normally incident waves are

shown in Fig. 4. The proposed FSS gives a reflection phase of

±90° for the x- and y-polarized waves from 3.6 GHz to 5 GHz.

As explained in [9], this 90° phase difference produces a CP

radiation in the forward direction. This phase difference en-

hances the CP bandwidth of 3.6–5 GHz of the perturbed slot

antenna. The total field E is the sum of the direct radiating field

(Ef) from the slot and the reflected field (Er) from FSS. It is

given by

Fig. 4. Reflection phase characteristics of the FSS unit cell.

𝐸 𝐸⃗ 𝐸⃗√

𝑒 𝑥 𝑦 𝑗 𝑥 𝑦 (2)

where 𝐸⃗√

𝑥. 𝑒 𝑦. 𝑒 and

𝐸𝐸

√2 𝑥. 𝑒 . 𝑒 𝑦. 𝑒 . 𝑒 ,

where θx and θy are the polarization-dependent reflection phases

for the x- and y-polarized waves, respectively.

The design process of the proposed antenna is shown in Fig.

5. The CPW-fed conventional slot antenna (case a) resonates at

2.5 GHz with linearly polarized waves. The axial ratio plot in

Fig. 5(b) shows that the conventional slot has a high value of

axial ratio at its resonance frequency. The corner perturbed slot

(a)

(b)

Fig. 5. Simulated reflection coefficient and axial ratio of the anten-

na design process: (a) |S11| and (b) axial ratio.


169

(case b) resonates at 3.66 GHz and produces a wideband circu-

lar polarization. To improve the axial ratio further, the FSS is

placed below the slot antenna (case c), which enhances the axial

ratio bandwidth.

The comparison among the conventional slot antenna, the

slot antenna over perfect electric conductor (PEC), and the

FSS-backed slot antenna provides the performance evaluation

of FSS is shown in Fig. 6(a) and (b). The S11 plot in Fig. 6(a)

shows that the resonance of the slot over PEC and the slot over

FSS is shifted compared with the perturbed slot because of the

loading of PEC or FSS. In Fig. 6(b), the perturbed slot antenna

has a lesser axial ratio bandwidth than the FSS-backed slot an-

tenna because of the forward signal from the slot, and the re-

flected signal from the FSS produces orthogonal fields in the

far-field region producing a wideband CP.

The reflection coefficient of the designed antenna with and

without tuning stubs is shown in Fig. 6(c). The role of the tun-

ing stubs in the CPW feed is to tune and optimize the imped-

ance bandwidth, so that the CP bandwidth lies within the

specified impedance bandwidth. The slot antenna directivity is

improved by nearly 3 dB for both PEC and FSS, but improving

the axial ratio bandwidth is not possible for PEC. The gain

comparison of the perturbed slot and the perturbed slot with

FSS is illustrated in Fig. 6(d). In this figure, the gain of the per-

turbed slot is less which is 1.88 dB at 4.2 GHz because of the

bidirectional radiation pattern. The FSS is placed below the

perturbed slot antenna, which increases the gain up to 4.87 dB.

The above comparison infers that the FSS-backed perturbed

slot antenna improves the axial ratio bandwidth and gain.

III. EXPERIMENTAL RESULTS AND DISCUSSION

The FSS-backed slot antenna is fabricated, and the simula-

tion results are exper imentally verified. Fig. 7 presents a photo-

graph of the fabricated prototype. The simulated and measured

reflection coefficient magnitudes are plotted in Fig. 8(a). A sim-

ulated impedance bandwidth of 55.23% is achieved from 2.87

GHz to 5.06 GHz. The measured impedance bandwidth of

63.22% is obtained from 2.65 GHz to 5.10 GHz. A simulated

3 dB axial ratio bandwidth of 30.94% is achieved from 3.66

GHz to 5 GHz, and the measured axial ratio bandwidth of

(a) (b) (c)

Fig. 7. Prototype of the fabricated antenna: (a) slot antenna, (b)

FSS, and (c) proposed antenna.

(a) (b)

(c) (d)

Fig. 6. Simulated reflection coefficient, axial ratio, and gain plot of the slot, the slot with PEC, and the slot with FSS: (a) |S11|, (b) axial ratio,

(c) |S11| of the proposed antenna with and without tuning stubs, and (d) gain of the proposed antenna with and without FSS.


170

(a)

(b)

Fig. 8. Simulated and measured reflection coefficient, axial ratio,

and gain of the proposed antenna: (a) |S11| and (b) axial ra-

tio and gain.

(a) (b)

(c) (d)

Fig. 9. Simulated and measured LHCP and RHCP radiation pat-

terns in (a) the xz plane at 3.98 GHz, (b) yz plane at 3.98

GHz, (c) xz plane at 4.42 GHz, and (d) yz plane at 4.42

GHz.

31.14% is obtained from 3.66 GHz to 5.01 GHz. The simulat-

ed and measured broadside gain and axial ratio plots are shown

in Fig. 8(b). The broadside peak gain of 4.87 dB is obtained at

4.2 GHz with a 2 dB variation over the 3 dB axial ratio band-

width. The simulated and measured radiation patterns are given

at different frequencies in both the xz and yz planes in Fig. 9.

The simulated and measured radiation patterns agree well for

both the RHCP and the LHCP. The LHCP level is 16 dB

below the RHCP level in both the xz and yz planes. The

LHCP is obtained at an inclined angle of 20° with respect to

the broadside direction. The tilted LHCP radiation is due to

the perturbation effect of the slot antenna.

IV. CONCLUSION

A wideband CP slot antenna with FSS loading is proposed in

this study. The truncated slot antenna achieves an optimized

axial ratio but with a low gain. To improve the gain and the

axial ratio of the truncated slot antenna, an FSS is used as a re-

flector. The proposed antenna achieves a measured impedance

bandwidth of 2.44 GHz and a 3 dB axial ratio bandwidth of

1.35 GHz with a peak gain of 4.87 dB. The proposed antenna

is useful for CP applications, such as radar and satellite, over the

CP frequency range of 3.66–5.01 GHz.

This work was supported by the Indian Space Research

Organization under the RESPOND program.

REFERENCES

[1] R. Cao and S. C. Yu, "Wideband compact CPW-fed circu-

larly polarized antenna for universal UHF RFID read-

er," IEEE Transactions on Antennas and Propagation, vol. 63,

no. 9, pp. 4148-4151, 2015.

[2] S. Karamzadeh, V. Rafii, M. Kartal, and H. Saygin, "Com-

pact UWB CP square slot antenna with two corners con-

nected by a strip line," Electronics Letters, vol. 52, no. 1, pp.

10-12, 2016.

[3] J. Pourahmadazar and S. Mohammadi, "Compact circularly-

polarised slot antenna for UWB applications," Electronics

Letters, vol. 47, no. 15, pp. 837-838, 2011.

[4] J. Y. Sze, J. C. Wang, and C. C. Chang, "Axial-ratio band-

width enhancement of asymmetric-CPW-fed circularly-

polarised square slot antenna," Electronics Letters, vol. 44, no.

18, pp. 1048-1049, 2008.

[5] R. K. Saini, S. Dwari, and M. K. Mandal, "CPW-fed dual-

band dual-sense circularly polarized monopole antenna,"

IEEE Antennas and Wireless Propagation Letters, vol. 16, pp.

2497-2500, 2017.

[6] S. Fu, Q. Kong, S. Fang, and Z. Wang, "Broadband circularly


171

polarized microstrip antenna with coplanar parasitic ring

slot patch for L-band satellite system application," IEEE

Antennas and Wireless Propagation Letters, vol. 13, pp. 943-

946, 2014.

[7] S. P. Pan, J. Y. Sze, and P. J. Tu, "Circularly polarized square

slot antenna with a largely enhanced axial-ratio band-

width," IEEE Antennas and Wireless Propagation Letters, vol.

11, pp. 969-972, 2012.

[8] R. Li, B. Pan, A. N. Traille, J. Papapolymerou, J. Laskar, and

M. M. Tentzeris, "Development of a cavity-backed broad-

band circularly polarized slot/strip loop antenna with a sim-

ple feeding structure," IEEE Transactions on Antennas and

Propagation, vol. 56, no. 2, pp. 312-318, 2018.

[9] F. Yang and Y. Rahmat-Samii, "A low profile single dipole

antenna radiating circularly polarized waves," IEEE Trans-

actions on Antennas and Propagation, vol. 53, no. 9, pp. 3083-

3086, 2015.

[10] H. Yi and S. W. Qu, "A novel dual-band circularly polar-

ized antenna based on electromagnetic band-gap struc-

ture," IEEE Antennas and Wireless Propagation Letters, vol.

12, pp. 1149-1152, 2013.

Puneeth Kumar Tharehalli Rajanna received his B.E. degree in telecommunication engi-

neering and M.Tech. degree in digital communica-

tion engineering from Visveswaraya Technological

University, Belgaum, Karnataka, India, in 2009 and

2013, respectively. He is currently working toward

his Ph.D. in electronics and communication at the

National Institute of Technology Karnataka, Man-

galore, India. His field of research includes met-

amaterial-based antennas, microwave antennas, and microstrip antennas.

Karthik Rudramuni received his B.E. degree in electronics and commu-

nication engineering and M.Tech. degree in RF and

microwave engineering from Visveswaraya Techno-

logical University, Belgaum, Karnataka, India, in

2009 and 2013, respectively. He is currently working

toward his Ph.D. in electronics and communication

at the National Institute of Technology Karnataka,

Mangalore, India. His field of research includes

metamaterial-based antennas, Goubau line-based leaky wave antennas, and

microstrip antennas.

[11] K. Agarwal and A. Alphones, "Unidirectional wideband

circularly polarised aperture antennas backed with artificial

magnetic conductor reflectors," IET Microwaves, Antennas

& Propagation, vol. 7, no. 5, pp. 338-346, 2013.

[12] K. Agarwal and A. Alphones, "Wideband circularly polar-

ized AMC reflector backed aperture antenna," IEEE

Transactions on Antennas and Propagation, vol. 61, no. 3, pp.

1456-1461, 2013.

[13] H. Li, K. W. Xu, X. B. Wang, R. Kanan, and L. X. Ran,

"Low-profile circularly polarised loop antenna assisted with

an effective PMC bandwidth," Electronics Letters, vol. 49,

no. 16, pp. 978-979, 2013.

[14] P. C. Sharma and K. C. Gupta, "Analysis and optimized

design of single feed circularly polarized microstrip anten-

nas," IEEE Transactions on Antennas and Propagation, vol.

31, no. 6, pp. 949-955, 1983.

[15] K. Kandasamy, B. Majumder, J. Mukherjee, and K. P. Ray,

"Dual-band circularly polarized split ring resonators loaded

square slot antenna," IEEE Transactions on Antennas and

Propagation, vol. 64, no. 8, pp. 3640-3645, 2016.

Krishnamoorthy Kandasamy received his B.E. degree in electronics and commu-

nication engineering from Bharathiar University,

Coimbatore, India, in 2003, his M.E. degree in

communication systems from the College of Engi-

neering, Guindy, Anna University, Chennai, India,

in 2007, and his Ph.D. in electrical engineering from

IIT Bombay, Mumbai, India, in 2016. He is cur-

rently an assistant professor at the Department of

Electronics and Communication Engineering, National Institute of Tech-

nology Karnataka, Surathkal, India. His current research interests include

metamaterials, antenna engineering, microwave integrated circuits (MICs),

and monolithic MICs.




172

I. INTRODUCTION

Previous studies on the effect of the electric and magnetic

fields (EMFs) of power lines on health have produced incon-

sistent and often contradictory results [1]. Many people are con-

cerned about power frequency (50 Hz) EMFs produced by elec-

trical distribution lines (EDLs) because the research indicates

possible harmful effects of exposure to these fields. According to

previous studies and electromagnetic fields standards, the max-

imum allowable magnetic field is 0.04 μT, and the maximum

allowable electric field is 4–5 kV/m [1, 2]. Many studies have

addressed the evaluation of the electrically induced voltage and

current in living organisms [1, 2]. The conclusion is that power

lines are generally assumed to be harmless. However, this as-

sumption has never been adequately tested. Low-level harmful

effects could be missed, even though they might be an im-

portant issue for the population as a whole, particularly as elec-

tric lines are ubiquitous. Other studies have evaluated the EMFs

of different power lines that are considered major sources of

pollution [3–9]. These fields may affect the operation of nearby

electric and electronic devices and appliances and may affect

various living organisms [1–5]. Electric distribution systems are

becoming increasingly prevalent worldwide, but despite this the

environmental effects of EMFs have not been adequately stud-

ied. In recent years, there has been significant public concern

prompted by reports on the adverse health effects of EMFs [6].

As stated, existing studies on the environmental pollution of

EMFs [1–9] report the maximum allowable electric field to be

4–5 kV/m and the maximum allowable magnetic field to be

0.04 μT for small animals and about 4–100 μT for humans,

depending on body area and the value of the line voltage and

Health Effects Study and Safe Distance Determination

Near an Electrical Distribution Network Mehdi Nafar*

Abstract

Living organisms are affected by environmental pollution, one main source of which is the electric and magnetic fields (EMFs) of electri-

cal distribution lines that produce high-voltage currents. Nowadays, electrical power transmission lines and distribution systems are be-

coming increasingly prevalent, with power distribution lines passing through cities and being located very close to buildings. For this rea-

son, there has been significant public concern, which has escalated because of reports about the adverse health effects of the EMFs of

power distribution systems. In this paper, the environmental pollution produced by several different configurations of electrical distribu-

tion systems in Iran is studied. The paper presents new software to calculate EMFs around electric power lines. Using this software and

based on maximum allowable EMFs, a safe margin around selected power lines has been determined. According to the study results, it

seems that in Iran this safe distance is not met. Therefore, the standards relating to power distribution lines should be reviewed.

Key Words: Electric Field, Electrical Distribution Systems, Magnetic Field, MATLAB, Safe Margin, Software Package.

Manuscript received December 17, 2018 ; Revised March 29, 2019 ; Accepted May 07, 2019. (ID No. 20181217-087J)

Department of Electrical Engineering, Marvdasht Branch, Islamic Azad University, Marvdasht, Iran. *Corresponding Author: Mehdi Nafar (e-mail: [email protected])




NAFAR: HEALTH EFFECTS STUDY AND SAFE DISTANCE DETERMINATION NEAR AN ELECTRICAL DISTRIBUTION NETWORK

173

current [1, 2]. Unfortunately, in Iran power lines in cities have

been developed irregularly and are very close to the place of hu-

man life. In this paper, we have developed a new software pack-

age using MATLAB to calculate EMFs around electrical lines.

By analyzing the results, the environmental pollution level

around electrical distribution systems (EDSs) can be determi-

ned and the safe distance calculated.

II. ELECTROMAGNETIC FIELDS

Electric fields are created only when there is electric voltage.

The magnitude of an electric field is proportional to the voltage

of an electric line. There is no relationship between electric field

strength and current. In the literature, electric field data are

presented in units of V/m. Electric fields are caused by electric

charges and are described by Gauss’s law. This is used to find the

electric field generated by voltage in highly symmetric geo-

metries, such as an infinitely long wire. Gauss’s law may be

expressed as:

o

QsdE

. , (1)

where Q is the electric flux through a closed surface S, E is the

electric field, 𝑑𝑠 is a vector representing an infinitesimal ele-

ment of area of the surface enclosing any volume V, and ε0 is the

electric constant. Using Gauss’s law, electric fields around

electrical lines can be determined. The equations for calculating

the electric field around a power transmission line are described

in [9]. Magnetic fields are created only when there is an electric

current [1]. The magnitude of a magnetic field is proportional to

the current flow through an electric line and not the voltage. As

the current increases, so does the magnetic field. There is no

relationship between magnetic field strength and voltage. Re-

garding electric transmission lines, it is not uncommon for a 20

kV electric line to have a higher magnetic field than a 115 kV

line. High-voltage 400 kV lines can carry large currents and

therefore may produce relatively high magnetic fields, but

primary distribution lines with voltages less than 63 kV can

produce fields similar to those measured around a transmission

line if they carry enough current. Magnetic fields rapidly be-

come weaker with distance from the source. However, they do

pass through most non-metallic materials, so they are more

difficult to shield. In the literature, magnetic field data are

presented in either units of gauss (G) or tesla (T). A milligauss

(mG) is equal to one thousandth of a gauss (G). One tesla is

equal to 10,000 gauss. A micro-tesla (μT) is equal to one

millionth of a tesla or 10 mG. A useful law is Ampere’s Law

that relates the magnetic field around a closed loop to the

electric current passing through the loop. This law is used to

find the magnetic field generated by currents in highly sy-

mmetric geometries, such as an infinitely long wire or a solenoid.

According to Ampere’s Law, the integral of B around any closed

mathematical path equals μ0 times the current intercepted by the

area spanning the path. Eq. (2) illustrates this concept [1]:

IdlB 0. , (2)

where the line integral is over any arbitrary loop, I is the current

enclosed by that loop, and r is the distance from the center of

the wire. Using Ampere’s Law, the magnetic fields of power

transmission lines can be calculated [1].

III. HISTORY OF THE HEALTH EFFECTS OF EMFS

After more than three decades of research, there is still public

concern regarding exposure to EMFs and the increased risk of

childhood cancer. This concern largely derives from studies of

people living near power lines or working in electrical-related

occupations. Some of these studies appear to show a weak

association between exposure and power-frequency EMFs and

the incidence of some cancers. Concerns about the EMFs of

power lines were first raised in a 1979 study that associated an

increased risk of childhood leukemia with residential proximity

to power lines [2]. In 1998, an international panel of experts

convened by the National Institute of Environmental Health

Sciences reported that EMFs like those surrounding electric

power lines should be regarded as a possible human carcinogen

[1–3]. Some diseases reportedly caused by electromagnetic

radiation include childhood leukemia, brain tumors, and neu-

rodegenerative diseases, such as Alzheimer disease. Health risks

from high-voltage power lines are not limited to children, and a

study published in 2008 reported an increase in the incidence of

Alzheimer disease in a population living near high-voltage

power lines [3]. Recent laboratory studies suggest that EMFs

with intermittent or transient characteristics may produce

biological effects [4]. Many studies [1–7] have attempted to

determine whether childhood cancer can be linked to magnetic

fields from power lines. Different confounding factors, such as

air pollution, water contamination, exposure to hazardous

chemicals, heavy traffic, and housing density, could conceivably

contribute to cancer. Experimental studies [10–14] conducted

on more than 2,500 animals of different species (mice, cats,

rabbits, and monkeys) have aimed to evaluate the most sen-

sitivities, body nervous and immune systems, and also vital

activity. The results of this research confirm a relatively

biological activity of 50 Hz magnetic fields. It was concluded

that the biological effects of 50 Hz magnetic fields depend on

the exposure intensity and duration. The obtained data on

changes in the immune and other systems indicated that a value

of 0.04 μT magnetic flux density during 4 hours per day of

exposure had adverse health effects on the experimental animals.


174

This value is higher for humans and depends on body area and

the value of the line current. All these studies clearly reveal that

the health risks associated with high-voltage power lines and

can affect a significant amount of people living in proximity to

these lines.

IV. SOFTWARE PACKAGE

MATLAB is one of the most powerful engineering applica-

tions. In this work, we used MATLAB to develop a program to

calculate the EMFs around transmission lines in three-phase

systems. The input parameters included the voltage of wires, the

current, the height of transmission towers, the diameter of wires,

and the position of the phases. By using input data and based on

the equations in Section II, EMFs around distribution systems

at a constant height from the ground or in a fixed width of the

transmission tower can be calculated. The main menu of the

developed software is shown in Fig. 1.

V. CALCULATION OF EMFS AROUND ELECTRICAL

DISTRIBUTION LINES

In Iran, several line configurations are installed for different

distribution voltage levels. Fig. 2 shows several typical lattice-

type configurations for the 20 kV level.

The data for these cases are listed in Table 1.

Fig. 3 shows the most common configurations for 400 V

lines.

The data for these cases are listed in Table 2.

VI. RESULTS AND SAFE MARGIN CALCULATION

According the previous studies that mentioned above, maxi-

mum allowable electric fields is 4–5 kV/m and maximum allow-

Fig. 1. The main menu of the developed software.

Fig. 2. Line conductor configurations of 20 kV lines.

Table 1. The data for 20 kV cases

Voltage (kV) 20

Diameter of wires (mm) 13.2

Maximum current (A) 200

Fig. 3. Line conductor configurations of 400 V lines.

Table 2. The data for 400 V cases

Voltage (V) 400

Diameter of wires (mm) 3.9

Maximum current (A) 150

Fig. 4. Power distribution lines close to residential buildings.

able magnetic field is 0.04 μT for small animals and about 4–

100 μT for humans, and it will depend on the percentage area

of the human body with respect to the animal’s body area and

also on the value of the line voltage and current. The results of

NAFAR: HEALTH EFFECTS STUDY AND SAFE DISTANCE DETERMINATION NEAR AN ELECTRICAL DISTRIBUTION NETWORK

175

Fig. 5. The curve of an electric field near a 20 kV line.

Fig. 6. The curve of an electric field near a 400 V line.

Fig. 7. The curve of a magnetic field near a 20 kV line.

this study are listed in Table 3.

Based on these results, it can be concluded that:

• Environmental pollution in the form of electric fields is

minimal and can be contained; therefore, the health effects

relating to electric fields are negligible.

Fig. 8. The curve of a magnetic field near a 400 V line.

Table 3. Safe margin around EDLs

Safe distance (m)

Based on electric field Based on magnetic field

20 kV 0.9 10

400 V 0.27 7.5

• The magnetic fields of EDLs are a form of environmental

pollution, and based on previous studies, living near power

lines can be detrimental to human health.

• Environmental pollution in the form of magnetic fields is

much greater than environmental pollution in the form of

electric fields and should be studied carefully.

• Living close to EDLs may be hazardous.

• Based on electric distribution design standards in Iran,

there must be a minimum horizontal distance of 3 m be-

tween any part of a building and the closest 20 kV line and

1.5 m between any part of a building and the closest 400 V

line. It seems that the safe distance standards for power

distribution lines should be reviewed.

VII. CONCLUSION

This paper investigated environmental pollution caused by

EMFs near EDSs in Iran. A software package was developed

using MATLAB to calculate these EMFs. Using this software,

EMFs near 20 kV and 400 V distribution lines were deter-

mined. Based on the results and on safe EMF levels, safety

margins around EDLs were calculated. It was shown that envi-

ronmental pollution in the form of electric fields is minimal, but

environmental pollution in the form of magnetic fields is signif-

icant enough to warrant careful study. It seems that the safe

distance standards for power distribution lines should be re-

viewed.

REFERENCES

[1] M. Nafar, "Magnetic field calculation around 230kV bun-


176

dled transmission lines," International Journal of Engineering

Innovations and Research, vol. 2, no. 6, pp. 463-466, 2013.

[2] N. Wertheimer and E. D. Leeper, "Electrical wiring config-

urations and childhood cancer," American Journal of Epide-

miology, vol. 109, no. 3, pp. 273-284, 1979.

[3] R. M. Lowenthal, D. M. Tuck, and I. C. Bray, "Residential

exposure to electric power transmission lines and risk of

lymphoproliferative and myeloproliferative disorders: a case-

control study," Internal Medicine Journal, vol. 37, no. 9, pp.

614-619, 2007.

[4] A. Ahlbom, "Neurodegenerative diseases, suicide and de-

pressive symptoms in relation to EMF," Bioelectromagnet-

ics, vol. 22, no. S5, pp. S132-S143, 2001.

[5] A. Huss, A. Spoerri, M. Egger, and M. Roosli, "Residence

near power lines and mortality from neurodegenerative dis-

eases: longitudinal study of the Swiss population," American

Journal of Epidemiology, vol. 169, no. 2, pp. 167-175, 2009.

[6] V. C. Motrescu and U. V. Rienen, "Computation of cur-

rents induced by ELF electric fields in anisotropic human

tissues using the Finite Integration Technique (FIT),"

Advances in Radio Science, vol. 3, pp. 227-231, 2005.

[7] A. Ahlbom, A. Green, L. Kheifets, D. Savitz, and A.

Swerdlow, "Epidemiology of health effects of radiofrequen-

cy exposure," Environmental Health Perspectives, vol. 112, no.

17, pp. 1741-1754, 2004.

[8] J. R. Reitz and F. J. Milford, Foundations of Electromagnetic

Mehdi Nafar received his B.S., M.S., and Ph.D. degrees in electri-

cal engineering in 2002, 2004, and 2011 from

PWIT University, Amirkabir University of Tech-

nology, and Islamic Azad University, Science and

Research Branch, all in Tehran. He graduated from

all programs with First Class Honors. He is currently

Assistant Professor in the Department of Electrical

Engineering, Marvdasht Branch, Islamic Azad Uni-

versity in Marvdasht, Iran. He is the author of more than 50 journal and

conference papers. His teaching and research interests include power sys-

tem and transformer transients, lightening protection, and optimization

methods in power systems.

Theory, Reading, MA: Addison-Wesley, 1960.

[9] B. Florkowska, A. Jackowicz-Korczynski, and M. Timler,

"Analysis of electric field distribution around the high-

voltage overhead transmission lines with an ADSS fiber-

optic cable," IEEE Transactions on Power Delivery, vol. 19,

no. 3, pp. 1183-1189, 2004.

[10] I. Katsuo and N. Hayashi, "Characterization of electric and

magnetic fields at ground level under EHV transmission

lines," in Proceedings of International Symposium on High

Voltage Engineering, Dresden, Germany, 1991.

[11] M. O. Melo, L. C. Fonseca, E. Fontana, and S. R. Naidu,

"Electric and magnetic fields of compact transmission

lines," IEEE Transactions on Power Delivery, vol. 14, no. 1,

pp. 200-204, 1999.

[12] L. L. Dvorak and L. Romero, "Evaluation of grounding

methods used for maintenance of transmission lines,"

IEEE Transactions on Power Delivery, vol. 12, no. 3, pp.

1379-1384, 1997.

[13] P. S. Maruvada, A. Turgeon, D. L. Goulet, and C. Cardi-

nal, "An experimental study of residential magnetic fields

in the vicinity of transmission lines," IEEE Transactions on

Power Delivery, vol. 13, no. 4, pp. 1328-1334, 1998.

[14] R. G. Olsen and P. S. K. Wong, "Characteristics of low

frequency electric and magnetic fields in the vicinity of

electric power lines," IEEE Transactions on Power Deliv-

ery, vol. 7, no. 4, pp. 2046-2055, 1992.




177

I. INTRODUCTION

In civil engineering, cavities and fragile areas are the most im-

portant obstacles in construction zones and quarries when de-

ciding suitable places for using explosives [1]. In environmental,

industrial, and urban construction points of view, the existence

of cavities underground is a potential source of dangerous phe-

nomena [2]. Ground penetrating radar (GPR) is a non-invasive

subsurface prospecting technique based on the propagation of

electromagnetic waves [3]. Its measurements are sensitive to

dielectric permittivity variations between the cavity and the ex-

ternal environment [4]. The development of imaging to esti-

mate the electromagnetic properties of the medium include the

dielectric permittivity ε (F/m) and the electrical conductivity σ

(S/m), which are critical issues for a physical interpretation of

the target structures. Moreover, GPR is based on the transmis-

sion and reception of 10 MHz–2.6 GHz electromagnetic waves

into the ground. The dielectric permittivity of the target over-

whelms the permittivity of other soil components, and the pres-

ence of water in the soil principally governs GPR wave propaga-

tion. The system transmits a pulse in very short duration waves

in the material and records the amplitude and time of arrival of

each wave. Reflection waves are produced to the right of any

changes in the properties of conduction of electric current in the

middle (dielectric constant). Its amplitude is determined from

the contrast of dielectric permittivity between the banking and

the target. One part of the transmitted wave energy continues to

Analysis and Modeling of GPR Signals to

Detect Cavities: Case Studies in Morocco Gamil Alsharahi1,* · Ahmed Faize2 · Carmen Maftei3 · Mohamed Bayjja1 ·

Mohamed Louzazni4 · Abdellah Driouach1 · Abdellatif Khamlichi5

Abstract

This research studied the detection of cavities and other abnormalities in subsoil utilizing ground penetrating radar (GPR) and an algo-

rithm developed under MATLAB. The study also analyzed and modeled GPR signals to find the exact location of cavities in the subsoil.

The algorithm was based on the finite difference time domain (FDTD) method to simulate cavities and compare them with an experi-

mental study. This study used GPR with 400 MHz and 200 MHz to investigate and detect cavities. Three different areas were chosen

with different proprieties of subsoil to detect the cavities: mountainous, coastal, and lowland areas. The results show high accuracy in the

detection of cavities using GPR.

Key Words: Algorithm, Detection, FDTD, Ground Penetrating Radar, Subsoil.

Manuscript received November 24, 2018 ; Revised April 17, 2019 ; Accepted May 15, 2019. (ID No. 20181124-080J) 1Department of Physics, Faculty of Sciences, Abdelmalek Essaadi University, Tetouan, Morocco. 2Department of Physics, Polydisciplinary Faculty, Mohammed 1st University, Nador, Morocco. 3Department of Civil Engineering, Faculty of Civil Engineering, Ovidius University of Constanta, Constanta, Romania. 4Department of Electrical Engineering, National School of Applied Sciences, Abdelmalek Essaadi University, Tangier, Morocco. 5Department of Industrial and Civil Engineering, National School of Applied Science (ENSA), Tetouan, Morocco. *Corresponding Author: Gamil Alsharahi (e-mail: [email protected])





178

spread in the middle until it is attenuated to be detected. The

attenuation of the transmitted signal is extremely variable and

depends greatly on the electrical conductivity of the materials.

Any field with high electrical conductivity will very strongly

attenuate the radar waves and vice versa [5, 6]. There are nu-

merous applications for GPR radar in various research areas:

geology, mining, environmental, geotechnical, civil engineering,

hydrology, archeology, and others [1–3]. A high-resolution seis-

mic reflection was used to detect and characterize underground

cavities [2]. An experimental detection was also used [3] as a

geophysical survey routine for the detection of doline areas in

the surroundings of Zaragoza, using two techniques. The first

was magnetometry and electromagnetic multi-frequency radia-

tion to examine the entire survey area to determine anomalous

subsoil behaviors; the second was GPR electric tomography, or

seismic methods, to generate an indirect view of the subsoil

structure to detect cavities [7, 8]. Cavity detection has been con-

sidered one of the primary targets in geophysical prospecting,

representing a principal goal for the prevention of risk. For ex-

ample, cavity detection in urban areas is important to prevent

accidents related to possible collapses, and in archeology, these

cavities are even more important [9]. Most of the techniques

proposed lie in the detection of water or air-filled voids whose

physical properties (density, electrical resistivity and conductivity,

dielectric constant, magnetic susceptibility) are considerably

different from the host sediment or rock [3]. In previous studies,

the detection of cavities could not fully distinguish those cavities

from others due to the similarity between certain materials and

the properties of constant dielectric cavities and iron. In this

work, we have developed a numerical program to present the

amplitude of the buried object’s signal during exploration. The

program has been effective distinguishing the cavities. In order

to conduct a preliminary study of the signal’s characteristics of

GPR reflections from cavities, numerical simulations were first

studied using the gprMax program. This program is a GPR

simulator based on the finite difference time domain (FDTD)

method developed by Antonis Giannopoulos [4]. This method

is widely used in various fields due to its generality and simplici-

ty [10], such as using optical analysis for an optical device, in-

cluding waveguide array and antenna analysis. The FDTD

method discretized the electrical and magnetic fields into the

dual lattice in space and time coordinates. However, the FDTD

method is essentially based on the integral of Maxwell’s equa-

tions.

This paper examines cavity detection in different fields with

experimental and theoretical simulation. The use of geophysical

methods, such as GPR and seismic methods, provides an effec-

tive alternative for cavity detection. Our analyses of cavity detec-

tion were conducted with high precision at several location areas

with different physical properties. To validate the accuracy of

the experimental results, a numerical simulate was developed to

detect the cavities. The theoretical and experimental results

show high accuracy for using GPR and simulation for real ap-

plications in various fields. Also, GPR and the developed code

can be used in several areas of civil engineering, such as for

buildings and bridges, or to detect human bodies under ruins

after earthquakes.

II. THEORETICAL BACKGROUND

FDTD or Yee’s method is a numerical analysis technique

widely used for modeling the dynamical system based in a dif-

ferential equation [10]. To implement the FDTD method in

Maxwell’ equations, a computational time domain must first be

determined. Generally, the computational domain is the physical

region over which the simulation will be performed or the area

scanned with GPR.

( , )( , ) ( , )

H r tE r t B r t

dt t

(1)

( , )( , ) ( , ) ( , ) ( , )

E r tH r t j r t D r t E r t

dt t

(2)

( , ) . ( , )

( , ) . ( , )

B r t H r t

D r t E r t

(3)

The FDTD method is based on the numerical resolution of

Maxwell’s equations expressed in cylindrical coordinates by fol-

lowing the adapted Yee scheme [11, 12], shown in Fig. 1(a).

The chart of the FDTD method is given in Fig. 1(b). The re-

flection of the strength created by GPR in a perpendicular inci-

dent wave depends on the contrast of relative dielectric con-

stants across the reflecting boundary. The reflection coefficient

RC can be expressed as in Davis and Annan [5, 13]:

(a) (b)

Fig. 1. Positions of the electric and magnetic field components

according to the cylindrical Yee scheme in a 3D case (a) and

the FDTD method (b).

ALSHARAHI et al.: ANALYSIS AND MODELING OF GPR SIGNALS TO DETECT CAVITIES: CASE STUDIES IN MOROCCO

179

1 2

2 1

r r

r r

RC

, (4)

where εr1 and εr2 are the relative dielectric constants of mediums

1 and 2, respectively. These constants are used in the calculation

of dielectric constant εr, according to the following equation

using c as the speed of light (0.3 m/ns) and h as the depth:

2

2rct

h

. (5)

III. MATERIALS AND METHODS

1. Materials

The GPR experimental data were collected using a GSSI

SIR-3000 device with two antennas, 400 MHz and 200 MHz,

for the center frequency. These antennas were adapted to be

used in different geological conditions. The main adjusted set-

tings were the acquisition gains that were measured in signal/

noise and time. A simulation was made to detect several cavities

and distinguish them from others using different targets and

different forms using RADAN 7 for data processing.

2. Description of the Antennas

The antennas used for this study were an RC loaded bowtie

designed for the 300 MHz to 3 GHz frequency range, as de-

scribed in [14]. Fig. 2(a) presents the geometry of the GPR an-

(a)

(b)

Fig. 2. (a) Geometry of RC bowtie antenna and (b) S11 variation.

Fig. 3. Mode scan and offset constant: (a) A-scan, (b) B-scan, and

(c) C-scan.

tennas, constituted by two half-ellipses. The capacitive loading

was designed in periodic slots cued on the antenna arms, and a

graphite sheet of 1 mm thickness was placed over each arm to

provide the resistive load. The antenna was fed using a vertical

microstrip to parallel stripline transition. Fig. 2(b) illustrates the

simulated and measured magnitude of the S11 parameter for the

RC bowtie antenna.

3. Mode Scan and Method of Prospection

The experimental data obtained by GPR were analyzed and

interpreted as an image of a vertical slice of the basement. The

image was received by the concatenation of the A-scan and B-

scan data recorded by GPR along the measured area [15]. Ob-

jects that fled the basement were marked by the presence of the

hyperbole in the B-scan (Fig. 3(b)) and C-scan (Fig. 3(c)) [1].

IV. RESULTS AND DISCUSSION

1. Cavity Detection in the Laboratory and Simulation

1.1 Comparison of simulation and experimentation results

(1,600 MHz)

In order to simulate and measure the electromagnetic wave

reflected by cavities and to validate the theoretical results ob-

tained by the simulation, we created a laboratory experiment.

We buried in the ground two iron bars (diameter of 16 mm)

and two plastic tubes with vide (diameter 32 mm) that have

dielectric permittivity of 7 F/m and conductivity of 0.001 S/m

as shown in Fig. 4. The antenna we used had a central frequency

of 1,600 MHz. Because the iron bar and the tubes had the same

dielectric permittivity, we used empty tubes.

Fig. 5(b) shows the direct waves: the first strong amplitude is

the combined reflection between the antenna and the reflected

signal. It has a negative reflection coefficient due to the air be-

tween the antenna and the land surface. The amplitude differ-

ence at t = 1.4 ns is (1.4, 500) and at t = 1.8 is (1.8, -800)

shows the effects of a cavity on the GPR signal.

1.2 Simulation of the cavity’s detection using 400 MHz

In this part of the study, we detected the cavity of a rectangu-


180

(a)

(b)

Fig. 4. (a) Image shows an empty tube buried and (b) radargram of

experimental results.

(a)

(b)

(c)

Fig. 5. (a) Simulation results of the empty tube, (b) amplitude sig-

nal obtained from modeling, and (c) zoom of (b).

Fig. 6. Schema of the rectangular target buried in soil (50 cm × 25 cm).

lar object (50 cm × 25 cm) presented in Fig. 6 using a simula-

tion of GPR measurements. The response was from the surface

buried in a homogeneous medium with a dielectric permittivity

of 7, a conductivity of 10-3, and a depth of 0.87 m as shown in

Fig. 5 by gprMax2D and MATLAB programs. Fig. 5(a) pre-

sented in the radargram shows the presence of hyperbola in Fig.

7(a) due to the total diffraction of the wave inside the cavity and

its reflection at the end. Fig. 7(b) shows the evolution of the

electric field. The amplitude Ez of the electric field for the first

travel time starts the diffraction for negative direction and then

has a very strong diffraction in the other direction, due to the

presence of the cavity, which has no conductivity.

Fig. 8(a) shows the simulation result for the detection of a

square conductor target. We note that the hyperbola in this fig-

ure is clearly visible because of the conductivity of the target. Fig.

8(b) shows the results of modeling for detection of the conduc-

tor targets. The total reflection of the electric field amplitude Ez

is shown in the first travel time for the positive direction and

then has a very strongly reflection in the other direction caused

by high electrical conductivity.

Fig. 9(a) presents the simulated detection of a rectangular die-

lectric target. We observe that the hyperbolas in this case is less

clear compared to the case of the conductor target. This is be-

cause the reflection and diffraction waves are approximately

equal, and because the positive and negative amplitude are ap-

proximately equal, as shown in Fig. 8(b). In order to detect the

difference between cavities and hyperbolas, the following condi-

tions are used:

(1) The dielectric constant of the cavity is equal to one and

the conductivity is zero;

(2) The hyperbolas of the cavity are in white and black;

(3) The amplitude of the signal has a minimal value in the

negative direction and a maximum value in the positive

direction.

2. Results of Investigation of Cavities in a Subsurface in Homo-

geneities using Data Acquisition

This section uses three GPR surveys perpendicular to the

ground in three different areas. We measured the GPR data in

Tube with vide

Direct wave

Reflection from of cavity

Medium of soil Depth=0.87 m


181

(a)

(b)

(c)

Fig. 7. (a) Simulation to detect rectangular cavities, (b) trace ampli-

tude of the wave reflection and refraction of the evolution of

the electric field, and (c) zoom of (b).

(a)

(b)

Fig. 8. (a) Simulation to detect a rectangular conductor and (b)

trace amplitude of the wave reflection and refraction.

(a)

(b)

Fig. 9. (a) Simulation to detect rectangular plastic and (b) trace

amplitude of the wave reflection and refraction.

the mode of prospection (common measure point). The results

shown in the radargrams are represented by the distance on axis

X and the depth on axis Y.

2.1 Cavity detection in a mountainous area: a quarry in Khamis

Anjara

The importance of this cavity detection in quarries is to iden-

tify suitable areas for the placement of explosives during the

production of gravel, sand, and other elements. This is the basis

of all construction in civil engineering. It is necessary to locate

these cavities and to record their coordinates with precision on

the ground so as to designate future areas for explosives and to

avoid fragile zones and cavities. If the explosives are placed in

cavities or fragile ground, there will be no explosion, causing loss

of money, effort, and time. The stone quarry located next to the

village of Khamis Anjara in a region of Tetouan city was used in

this study.

In Fig. 10(a), the cavity extends from 0.5 m to 5.3 m and the

depth ranges from 1 m to more than 2.5 m.

2.2 Cavity detection in a flat area (road project)

The aim of this survey in this area is to detect cavities in an

urban area. A cavities were found during work on the construc-

tion of a new road project. In this zone, a GPR survey was con-

ducted at a survey distance of 1,000 m and a depth of 6 m.

2.2.1 Methodology

In this part of the study, an antenna with 200 MHz was used

to detect an object at 8 m depth in the road. The results are giv-

1.8 2 2.2 2.4 2.6

x 10-8

-200

-150

-100

-50

0

50

100

150

200

t(s)

mV

/m

Distance(m)

t(s)

0.5 1 1.5 2

0

2

4

6

8

10

12

x 10-9

0 0.2 0.4 0.6 0.8 1 1.2

x 10-8

-1000

-500

0

500

1000

1500

t(s)

Am

pltide

Direct wave

cavity

Reflection from of cavity

Direct wave


182

(a)

(b)

Fig. 10. Radargrams of the profiles detecting the cavities (site

Khamis Anjara): (a) Profile 1 and (b) Profile 2.

(a)

(b)

Fig. 11. (a) Localization of the project in El Jadida city and (b) a

diagram of the work area survey.

en in B-scan mode. The road (El Jadida-Safi, 143 km) was di-

vided into four 4-km-long profiles for each investigation (Safi-

El Jadida). The distance between each profile was 2 m (Fig. 11).

The radar profiles show an overview of the studied site, with

3D profiles acquired along the road and illustrating the contrast

below the surface on different sections of the road. In order to

facilitate the reading of the radargram, the red squares of the

zones of reflections and diffractions represent cavities. The three

areas where we found a long cavity were more than 5 km. The

specific areas that were considered in this study have a length of

1 km per area, divided into several radargrams, where we could

not present all the data (information about targets). Two-

dimensional and 3D investigations are described and presented

below. Furthermore, the simulation of cavity detection in the

fourth section by the signals and radargram was applied, and the

work confirmed the results.

2.2.2 Investigations in zone 1B-scan mode (2D)

At the beginning, we carried out 2D investigations that be-

gan from 246 km to 247 km of the road. In the first zone (Fig.

12), we chose to investigate deep cavities at a depth different

from 2.2 m to 6 m.

Fig. 12 shows a GPR image obtained on the ground. At the

travel time of 120 ns and depth of 6 m, a strong reflection was

(a)

(b)

(c)

Fig. 12. Radar data showing cavities on the road project: (a) Profile

1, (b) Profile 2, and (c) Profile 3.

Study zone

cavity


183

detected, which is considered the electromagnetic wave reflected

from the cavity. In Profile 1, there are cavities from 0 m to 30 m.

In Profile 2, there are cavities from 600 m to 660 m. In Profile 3,

there are cavities from 485 m to 520 m.

2.2.3 3D investigation of zone 2 by C-scan modes

In this part of the study, we carried out 3D investigations on

lengths of 1 km and 6 km, respectively, divided into several ra-

dargrams and at a depth of 5m and a width of 2 m.

The second zone begins from 251 km and 252 km as a length

of the road at 3 km shown in Fig. 13. The coordinates of zone 2

are at depth different from 2.2 m to 6 m, length along 251–252

km (1 km), and width is 2 m.

2.3 Cavity detection in a coastal area: Harhoura

For this part of the study, we positioned the radar profiles as

shown in the map below. On the basis of indications of the plan

in Fig. 14, we indicate the various anomalies by area. The differ-

ent profiles performed in this area show that several areas have

empty attendance indices, soil destructuration, or simple frac-

tures with small voids. The profile in Fig. 14 shows an evolution

of anomalies with a fracturing of the subsoil that connects the

two anomalies. The first anomaly evolves in depth with the

presence of empty franc, an increase of the surface of the third

anomaly with the presence of a vacuum, and the presence of a

1m deep fracture that connects the two anomalies with a scat-

tered void.

The profile in Fig. 15 shows an evolution of anomalies and

fractures with the presence of voids throughout the profile

length and shows the presence of a significant cavity starting

from a depth of 4.5 m. The profile of Fig. 16 shows continuity

of the deep cavity with an increase in the volume of the anomaly

detected at the surface, mainly for the anomaly above the cavity,

with ramifications between them. We will indicate the results of

the additional auscultation that we have achieved to determine

the maximum depth of the cavity detected in zone 1.

We carried out radar profiles of deep cavity detection areas in

order to define in depth as seen in Figs. 14–16. The radar pro-

files carried out on that cavity could reach 15 m deep into usable

signals on this depth, and we have delimited the cavity at 8.20 m

depth. Beyond 8.20 m deep, the pitch becomes homogeneous.

V. CONCLUSION

In this work, we conducted numerical and simulated results,

as well as experiments and surveys in mountainous, flat, and

(a) (b)

(c) (d)

Fig. 13. Profiles of radargrams in zone 2 (profiles 18 and 23) along 251–52 km and depth 5 m: (a) from 45 m to 85 m, (b) from 485 m to

530 m, (c) from 730 m to 775 m, and (d) from 930 m to 975 m.


184

Fig. 14. Profile 4 next to zone 1.



Cavity

Cavity

Cavity Cavity Cavity

Cavity Abnormal Abnormal

Cavity

Cavity


185

coastal areas. We found great similarity between the results from

our modeling and the experimental results and field surveys. We

have been able to identify cavities in subsoil in an effective and

successful way. We observed that the hyperbola resulting from

the reflection of waves from the cavities in coastal areas is very

thin (faint) and different from the other regions due to the high

humidity in this area. Also noteworthy is that the hyperbolas are

insufficient to identify cavities so the use of an amplitude of

reflected waves is needed. The results obtained show the pres-

ence of cavities at different distances and depths in the places

studied. It is noted in this study that the depth of the investiga-

tions reached 8 m using a 400 MHz antenna, which can indi-

cate that the ground has distinctive characteristics and very low

conductivity. The detection of the cavity in the coastal area was

difficult compared to the mountainous and flat areas, which

were easier and more obvious.

REFERENCES

[1] G. Alsharahi, A. Faize, M. Louzazni, A. M. M. Mostapha,

M. Bayjja, and A. Driouach, "Detection of cavities and frag-

ile areas by numerical methods and GPR applica-

tion," Journal of Applied Geophysics, vol. 164, pp. 225-236,

2019.

[2] L. Driad and B. Piwakowski, "Detection and characterisa-

tion of underground cavities using high resolution seismic

reflection (HRSR)," in Proceedings of the 8th General Annual

Meeting of the European Section of the Environmental and

Engineering Society (EEGS-ES), Aveiro, Portugal, 2002.

[3] O. Pueyo-Anchuela, A. M. Casas-Sainz, M. A. Soriano,

and A. Pocovi-Juan, "A geophysical survey routine for the

detection of doline areas in the surroundings of Zaragoza

(NE Spain)," Engineering Geology, vol. 114, no. 3-4, pp.

382-396, 2010.

[4] A. Giannopoulos, "Modelling ground penetrating radar by

GprMax," Construction and Building Materials, vol. 19, no.

10, pp. 755-762, 2005.

[5] G. Alsharahi, M. M. A. Mint, A. Faize, and A. Driouach,

"Modelling and simulation resolution of ground-penetrating

radar antennas," Journal of Electromagnetic Engineering and

Science, vol. 16, no. 3, pp. 182-190, 2016.

[6] G. Alsharahi, A. Driouach, A. Faize, and A. Khamlichi,

"Effect of electrical conductivity and dielectric constant on

the performance of ground penetrating radar," International

Journal of Microwave and Optical Technology, vol. 10, no. 6,

pp. 458-463, 2015.

[7] G. Grandjean and D. Leparoux, "The potential of seismic

methods for detecting cavities and buried objects: experi-

mentation at a test site," Journal of Applied Geophysics, vol.

56, no. 2, pp. 93-106, 2004.

[8] Y. V. Kang and H. C. Hsu, "Application of ground pene-

trating radar to identify shallow cavities in a coastal

dyke," Journal of Applied Science and Engineering, vol. 16, no.

1, pp. 23-28, 2013.

[9] N. Barraca, M. Almeida, H. Varum, F. Almeida, and M. S.

Matias, "A case study of the use of GPR for rehabilitation

of a classified Art Deco building: the InovaDomus hou-

se," Journal of Applied Geophysics, vol. 127, pp. 1-13, 2016.

[10] W. Zhao, G. Tian, E. Forte, M. Pipan, Y. Wang, X. Li, Z.

Shi, and H. Liu, "Advances in GPR data acquisition and

analysis for archaeology," Geophysical Journal Internation-

al, vol. 202, no. 1, pp. 62-71, 2015.

[11] A. Benedetto and L. Pajewski, Civil Engineering Applica-

tions of Ground Penetrating Radar. Cham: Springer, 2015.

[12] V. Perez-Gracia and M. Solla, "Inspection procedures for

effective GPR surveying of buildings," in Civil Engineering

Applications of Ground Penetrating Radar. Cham: Springer,

2015, pp. 97-123.

[13] Davis, J. L. and A. P. Annan, "Ground penetrating radar

for high resolution mapping of soil and rock stratigraphy,"

Geophysical Prospecting, vol. 37, no. 5, pp. 531-551, 1989.

[14] K. K. Ajith and A. Bhattacharya, "Printed compact lens

antenna for UHF band applications," Progress In Electro-

magnetics Research C, vol. 62, pp. 11-22, 2016.

[15] A. Shakas and N. Linde, "Apparent apertures from ground

penetrating radar data and their relation to heterogeneous

aperture fields," Geophysical Journal International, vol. 209,

no. 3, pp. 1418-1430, 2017.


186

Gamil Alsharahi was born in Amran City, Yemen, in 1979. He ob-

tained a Bachelor of Mathematics and Master of

Electronic & Telecommunication degrees at Uni-

versity Sana’a, Faculty of Science, Amran, and Ab-

delmalek Essaadi University, Faculty of Science,

respectively. He also holds a Ph.D. in Electronic &

Telecommunications at the Faculty of Science, Te-

touan. As a Ph.D. student, he visited Ovidius Uni-

versity of Constanta, Romania, at the Faculty of Civil Engineering in 2019.

He has two books and over 28 scientific papers published in different jour-

nals and conference proceedings.

Ahmed Faize

was born in Morocco in 1985. Currently, he is a

professor and researcher at Mohammed 1st Univer-

sity, Faculty Polydisciplinarly, Nador, Morocco.

Carmen Maftei is Professor at the Civil Engineering Department of

Ovidius University in Constanta, Romania. She

attended the Polytechnic Institute of Iasi where she

majored in Land Reclamation in 1988. From 1996

to 1998, she specialized in water resources and pro-

tection of water resources, equivalent to a Master of

Science. In 2002, she finished her Ph.D. studies in

Water Science in Continental Environment. Her

research activities focus on: hydrology and hydrological modeling; geo-

graphic information systems applications in environmental sciences; and

remote sensing applications in environmental sciences. She has written 11

books and book chapters, over 90 scientific papers published in different

journals or conference proceedings, one patent, three international projects,

17 national grants, and 13 research contracts with economic partners. She

is a member of many editorial boards of scientific journals. Her 25 years of

teaching experience includes hydrology, hydraulics, and geographic infor-

mation systems and applications at Ovidius University of Constanta and

Transilvania University of Brasov in Romania.

Mohamed Bayjja was born in Errachidia, Morocco, in 1982. He ear-

ned mastery (Bac +4) in Computer Science, Elec-

tronics, Electrical, and Automatic from the Faculty

of Science and Technology, Errachidia, Morocco, in

2007. He also received a Master degree in 2011 and

a PhD degree, both in Electronic & Telecommuni-

cations from Abdelmalek Essaadi University, Mo-

rocco. He is preparing a postdoc at the Faculty of

Electronics, Telecommunications and Information Technology, Cluj-

Napoca in Romania. He authored or coauthored more than 13 publications

in international journals and conference proceedings.

Mohamed Louzazni obtained a Bachelor of Electronic and Master of

Electronic & Telecommunication degrees at the

University Ibn Tofail, Faculty of Science, Kenitra,

Morocco, and Abdelmalek Essaadi University, Fac-

ulty of Science, respectively. He also holds a Ph.D.

in electrical and renewable energy engineering at the

National School of Applied Science, Tangier. He

was a Ph.D. student visiting the University Poly-

technic of Bucharest at the Faculty of Electrical Engineering in 2015. In

2017, he started research at the University Polytechnic of Milan, Italy at the

Solar Tech laboratory in the Energy Department. In 2018, he was a Ph.D.

visitor at the University of Cadiz at the Higher Polytechnic School of Alge-

ciras, Spain, for 5 months. Currently, he is a researcher at the University

Polytechnic in Milan.

Abdellah Driouach was born in Al Hocei-ma City, Morocco, in 1954.

He received his license in electronic physics at the

University of Rabat (Morocco) in 1979, his doctor-

ate in spectronomiehertzienne at the University of

Bordeaux 1 (France) in 1983, and his Ph.D. in sci-

ences at the University of Grenade (Spain). He has

been a professor and researcher in the Faculty of

Science, Abdelmallek Essaadi University since 1983.

He has participated in several scientific research studies, including the dis-

persion of electromagnetic waves on obstacles to arbitrary structures. Cur-

rently, he is a member of the research team of communication systems.


187

Abdellatif Khamlichi was born in 1963 in Northern Morocco. He has

made his studies in Morocco and France. He was

graduated as Ph.D. in Mechanics at Ecole Centrale

of Lyon in France in 1992. Before that he was grad-

uated in 1998 as Civil Engineer at the National

School of State Public works which is located in

Vaulx-en-Velin in France. He is now professor at

the National School of Applied Sciences of Tetouan

in Morocco. His main research activity was dedicated to civil and mechani-

cal engineering. Among the topics he has worked on there are: waterham-

mer in piping systems, fluid-structure interaction, piezoelectric actuators,

crashworthiness, friction products, composites, stability of shells, internal

erosion, thermomechanical fatigue, control of wind turbines, seismic vul-

nerability, building reinforcement, inverse problem, structural health moni-

toring, non-destructive evaluation, GPR, soil-structure interaction. He

has published more than 44 scientific articles in Scopus indexed journals

and about 54 Scopus indexed conference papers. He has also advised more

than 20 PhD theses.




188

I. INTRODUCTION

With the rapid development of commercial and military avi-

onic radar systems, interest in antenna design technology for

guidance and control systems for avionic vehicles has increased

explosively during the past few years. To meet the performance

requirements of the antennas mounted on an object for direc-

tion-finding, many studies have been conducted [1–6]. To re-

duce power consumption, the antenna usually operates in pas-

sive Rx mode, with an additional Tx base station on the ground

[7], as shown in Fig. 1. The Tx station first transmits a signal to

the target in the air. When the reflected signal from the target

hits the Rx antenna, the front-end receiver module deduces the

amplitude of it as DC voltage [8, 9]. After processing and com-

Ground base station (Tx)

Target

Aviating platform (Rx)

Transmitted signal

Reflected signal

Fig. 1. Operating principal of the application.

puting that voltage value, the direction of the incoming signal

can be found by analyzing the differential amplitude of the volt-

A Compact Rx Antenna Integration for 3D

Direction-Finding Passive Radar Hojoo Lee1 · Jaesik Kim2 · Jaehoon Choi1,*

Abstract

A compact receiving (Rx) antenna integration for 3D direction-finding passive radar is proposed. The proposed antenna integration con-

sists of two Yagi antennas and two waveguide slot antennas. The antennas are mounted on four sides of a parallelepiped polycarbonate

module with the same type of antennas facing each other on the opposite sides of the module. The antenna is designed for a direction-

finding radar system as an Rx terminal on an avionic vehicle with a separate Tx base station on the ground. The proposed antenna inte-

gration demonstrates a maximum return loss of 9.6 dB (voltage standing wave ratio 2:1) ranging from 9.7 to 10.2 GHz in the X band and

has suitable radiation patterns for direction-finding system in both the xz and yz planes. In addition, the proposed antenna integration

has the compact size and light gross weight of 37 mm × 46 mm × 50 mm and 97 g, respectively, making it suitable for application in an

avionic system.

Key Words: Direction-Finding, Radar, Waveguide Slot Antenna, Yagi Antenna.

Manuscript received December 05, 2018 ; Revised March 27, 2019 ; Accepted May 17, 2019. (ID No. 20181205-085J) 1Department of Electronics and Communications Engineering, Hanyang University, Seoul, Korea. 2Agency for Defense Development, Daejeon, Korea.

*Corresponding Author: Jaehoon Choi (e-mail: [email protected])




LEE et al.: A COMPACT Rx ANTENNA INTEGRATION FOR 3D DIRECTION-FINDING PASSIVE RADAR

189

(a)

x z

y

(b)

Fig. 2. Four antennas mounted on a paralleled piped module for

direction-finding application: (a) perspective view and (b)

front view.

(a)

(b)

Fig. 3. A brief example of ideally suitable radiation patterns on xz

plane (a) and yz plane (b).

age measured from each Rx channel. Thus in the industrial field,

multi-channel Rx antenna integrations are generally used [10].

Each antenna is mounted on four faces of a parallelepiped an-

tenna module on an avionic vehicle as shown in Fig. 2. When

using the amplitude comparison scheme for direction-finding,

ambiguity of the gain amplitude can become an issue that af-

fects the accuracy of the system [11]. To eliminate that ambigui-

ty in a four-channel direction-finding system, the radiation pat-

terns of the four antennas in Fig. 2 should be similar to the pat-

terns shown in Fig. 3 [12]. In the ideal of that case, the gain

difference between ANT1–ANT2 and ANT3–ANT4 in xz

and yz planes can be graphically deduced as the plots given in

Fig. 4. To satisfy all those properties, we here propose a compact

and light-weight Rx antenna integration suitable for a 3D direc-

tion-finding radar system.

II. DESIGN OF SINGLE ANTENNA ELEMENTS

With reference to the radiation patterns in Fig. 3, four anten-

nas need to be chosen to generate the required beam patterns.

However, because the Tx antenna of a ground base station sys-

tem usually has a sharp beam with linear polarization [13, 14]

(e.g., dipole or patch array), the linearly polarized planes of the

(a) (b)

Fig. 4. A brief example of ideal gain difference between ANT1–

ANT2 and ANT3–ANT4 in (a) xz plane (b) yz plane.

(a) (b)

Fig. 5. Comparison of space occupancy between two types of an-

tenna integration using (a) four identical antennas and (b)

two types of antenna pairs with polarization orthogonal to

each other.

four Rx antennas must coincide with one another to receive the

reflected signal from the target. One way to obtain that property

is to use four identical antennas, such as a dipole antenna on

each side of the platform, as shown in Fig. 5(a). However, that

configuration requires a significant amount of space for the an-

tennas. Therefore, we propose two Yagi antennas and two wave-

guide slot antennas for ANT1–ANT2 and ANT3–ANT4 in

Fig. 2, respectively, as shown in Fig. 5(b), which requires much

less space.

1. Microstrip Yagi Antenna

The single element for a microstrip Yagi antenna is illustrated

in Fig. 6. The antenna consists of an SMA connector, dipole

element, rear reflector, front driver, polycarbonate (PC) spacer,

bottom reflector, and two PC bolt screws. Parameters of the

antenna were calculated using the basic properties of a general

Yagi antenna [15], as shown in Table 1, and then fine-tuning

them using HFSS software (version 2016.2, Ansys Corporation,

Pittsburgh, PA, USA). The simulated return loss of the antenna

is shown in Fig. 7. The bandwidth of the antenna to achieve

voltage standing wave ratio (VSWR) 2:1 is 9.25–11.8 GHz. As

a conventional Yagi antenna, the antenna has a directive and

nearly symmetric radiation pattern in the xz plane [16], as

shown in Fig. 8. On the other hand, the antenna exhibits a tilted

beam from the +z to +y direction at 50°, as shown in Fig. 8,


190

(a)

(b)

Fig. 6. Geometry of the proposed Yagi antenna element: (a) top

view and (b) side view.

Frequency [GHz]9.0 9.5 10.0 10.5 11.0

Ret

urn

Los

s [d

B]

0

5

10

15

20

25

Fig. 7. Simulated return loss of the Yagi antenna element.

due to the bottom reflector made of FR4 substrate with a full

metal pattern on its upper side. In addition, the antenna is line-

arly polarized in the xz-plane according to the nature of a dipole

antenna [17].

2. Waveguide Slot Antenna

The single element of a waveguide slot antenna is designed as

shown in Fig. 9. The resonant frequency is determined by the

half-wavelength (λ/2) central slot on the top of the waveguide.

Table 1. Theoretical values for parameter of Yagi antenna

Parameter Theoretical value

Distance

Reflector to dipole 0.2-0.35λDirector to director 0.2-0.35λ

Width

Dipole 0.015-0.025λDirector 0.015-0.025λ

Length

Dipole 0.45-0.49λDirector 0.4-0.45λReflector ≈0.5λ

Theta [deg]-90 -60 -30 0 30 60 90

Rea

lize

d g

ain

[d

Bi]

-20

-15

-10

-5

0

5

10

xz-plane(w reflector)yz-plane(w reflector)xz-plane(w/o reflector)yz-plane(w/o reflector)

Fig. 8. Simulated radiation pattern of the Yagi antenna element at

10 GHz with and without bottom reflector.

(a) (b)

Fig. 9. Geometry of the proposed waveguide slot antenna element:

(a) top view and (b) cross-sectional side view.

Impedance matching was achieved by placing a conducting

metal block inside the waveguide and adjusting the gap between

the feeder pin and the block. The simulated result shows that

the bandwidth of the antenna to achieve VSWR 2:1 is 9.85–

10.23 GHz, as shown in Fig. 10. Unlike conventional wave-

guide slot antennas [18–20], the waveguide slot antenna in this


191

Frequency [GHz]9.0 9.5 10.0 10.5 11.0

Ret

urn

Los

s [d

B]

0

5

10

15

20

Fig. 10. Simulated return loss of the waveguide slot antenna ele-

ment.

paper has three corrugation grooves on the top wall of the

waveguide. The one in the rear is for suppressing the surface

wave toward –z direction, and the two in the front induce more

surface wave toward the +z direction. The former is placed λ/2

away from the slot and the latter are located close to the slot to

obtain the tilted main beam [21–23]. Fig. 11 illustrates the ef-

fect of each corrugation on the main beam direction of the an-

tenna in the xz plane. As a final optimized result, the main

beam of the waveguide slot antenna in the xz plane is tilted 45°

from the +x direction to the +z direction while being symmet-

ric in the yz plane, as shown in Fig. 12. In addition, the antenna

is linearly polarized in the xz plane according to the nature of a

waveguide slot antenna [17].

III. ANTENNA INTEGRATION DESIGN

As illustrated in the previous section, the antenna integration

consists of two Yagi antennas (Y1 and Y2) and two waveguide

slot antennas (B1 and B2) mounted on each side of the plastic

module, as shown in Fig. 13. Considering the position of the

mounted antennas, the linearly polarized plane of the four an-

tennas is identical (i.e., coincides with the xz plane) due to the

Fig. 11. The effect of corrugation grooves on the radiation pattern

of the antenna in xz plane.

Theta [deg]-90 -60 -30 0 30 60 90

Rea

lize

d g

ain

[d

Bi]

-30

-25

-20

-15

-10

-5

0

5

10

xz-planeyz-plane

Fig. 12. Simulated radiation pattern of the waveguide slot antenna

element in both xz and yz planes at 10 GHz.

(a)

(b)

Fig. 13. Geometry of the proposed antenna integration: (a) per-

spective view and (b) rear view.

polarization characteristics of the antenna types. The module for

mounting the integrated antennas is made of PC (𝜀 = 2.8),

which is economical, easily machinable, and lightweight. The

simulated return loss characteristics of the antenna mounted on

the module are shown in Fig. 14. All four antennas satisfy the

maximum return loss level of 9.6 dB (i.e., VSWR 2:1) from

9.85 GHz to 10.15 GHz. Fig. 15 briefly illustrates the E-field

distributions of both antennas mounted on the module to verify

the origination of the tilted main beams in both the xz and yz

Frequency[GHz]9.0 9.5 10.0 10.5 11.0

Ret

urn

Los

s[dB

]

0

5

10

15

20

25

30

B1&B2Y1&Y2

Fig. 14. Simulated return loss of the proposed antennas.


192

y

x

z

(a)

y

x z

(b)

Fig. 15. E-field distributions of two antenna types at 10 GHz: (a)

waveguide slot antenna in xz plane and (b) Yagi antenna

in yz plane.

planes. Compared with the proposed Yagi antenna in the yz

plane, the waveguide slot antenna has a broader radiation pat-

tern in the xz plane due to the nature of the antenna mechanism.

Therefore, to obtain a steeper gradient—i.e., larger value of

(dBi/theta degree)—for the gain plot in the xz plane, both

waveguide slot antennas are rotated about 3° toward the z-axis,

which makes the radiation pattern similar to that in Fig. 3. As

shown in Fig. 16, the gradient of the gain decrement versus

theta becomes steeper in general as the antenna is rotated but

starts to flatten from 4°. Fig. 17 illustrates the optimized result

for simulated radiation patterns of the proposed antenna inte-

gration at 10 GHz in the xz and yz planes. As shown in the

figure, the radiation patterns of the B1 and B2 are identical in

the yz plane and symmetric in the xz plane. On the other hand,

the radiation patterns of Y1 and Y2 are identical in the xz plane

and symmetric in the yz plane. In addition, the maximum gains

of the B-type antennas in xz plane and the Y-type antennas in

yz plane are 7.5 dBi and 8.1 dBi, respectively.

To analyze the suitability of the radiation pattern illustrated

in Fig. 3, we conducted a graphical manipulation. The plots in

Theta [deg]-90 -60 -30 0 30 60 90

Rea

lize

d g

ain

[d

Bi]

-10

-5

0

5

10

rotation=0degrotation=1degrotation=2degrotation=3degrotation=4deg

Fig. 16. Effect of the rotation of the waveguide slot antenna on the

gradient of the gain versus theta in xz plane.

(a)

Theta[deg]

-90 -60 -30 0 30 60 90-25

-20

-15

-10

-5

0

5

10

B1 B2Y1Y2

(b)

Fig. 17. Simulated radiation patterns of the proposed antenna at 10

GHz: (a) xz plane and (b) yz plane.

Fig. 18 indicate the gain difference between the antennas with

identical types in the xz and yz planes at 10 GHz. The gradient

of the gain difference plots between B1 and B2 in the xz plane

and Y1 and Y2 in the yz plane are steep, whereas that between

B1 and B2 in the yz plane and Y1 and Y2 in the xz plane are

relatively flat. Thus, the antenna integration generally satisfies

the property explained in Fig. 4. However, there is a certain

range within which its performance deteriorates (e.g., |theta| 40° in both the xz and yz plane). Outside that range, the

radiation patterns of the proposed antenna integration offer good

performance for direction-finding application. On the other

hand, the slope of gain differences of Y1 and Y2 in the yz plane

and B1 and B2 in the yz plane are not identical to each other.

However, this can be overcame by employing calibration factor

to the gain difference slope at the signal processing stage in or-

der to eliminate possible error of the direction finding system.

IV. EXPERIMENTAL RESULTS

1. Fabrication of the Proposed Antenna

Because the proposed antenna integration combines four an-

tennas onto a plastic assemblage module, three mechanical pro-

Theta[deg]

-90 -60 -30 0 30 60 90

Rea

lized

Gai

n[dB

i]

-20

-15

-10

-5

0

5

10

B1 B2Y1Y2


193

Theta[deg]

-90 -60 -30 0 30 60 90

Gai

n di

ffer

ence

[dB

]

-20

-10

0

10

20

B1-B2Y1-Y2

(a)

Theta[deg]

-90 -60 -30 0 30 60 90

-30

-20

-10

0

10

20

30

B1-B2Y1-Y2

(b)

Fig. 18. Gain difference between B1–B2 and Y1–Y2 at 10 GHz vs.

theta: (a) xz plane and (b) yz plane.

x

z y (a)

y

x

z

(b)

Fig. 19. Fabricated antenna integration: (a) perspective view and (b)

rear view.

cessing techniques were used to fabricate the whole structure.

First, computer numerical control (CNC) machining was used

to manufacture the PC module and spacer. PC was used because

it is inexpensive, easily machinable, and has a relatively high tol-

erance to heat and impact [24]. Second, an etching process was

used to fabricate the Yagi antenna and FR4 reflector. For the

waveguide slot antennas, electroforming was applied because the

bent inner space is difficult to realize using CNC machining

[25]. To minimize unexpected degradations in performance, all

the bolts used to fasten the subparts were made of PC. Fig. 19

illustrates the finalized antenna integration. The total dimen-

sions of the proposed antenna integration are 37 mm × 46 mm

× 50 mm and its gross weight is 97 g. Therefore, the proposed

antenna integration is both compact and light, which are great

features for being mounted onto an avionic vehicle system.

2. Measured Results

The return loss and radiation pattern of the assembled anten-

Frequency[GHz]9.0 9.5 10.0 10.5 11.0

Ret

urn

Los

s[d

B]

0

5

10

15

20

25

30

35

40

45

B1B2Y1Y2

Fig. 20. Measured return loss of the proposed antennas.

Theta[deg]-60 -30 0 30 60

Rea

lized

Gai

n[dB

]

-20

-15

-10

-5

0

5

10

B1 (sim) B2 (sim)Y1 (sim)Y2 (sim)B1 (meas) B2 (meas) Y1 (meas) Y2 (meas)

(a)

Theta[deg]-60 -30 0 30 60

Rea

lized

Gai

n[dB

]

-25

-20

-15

-10

-5

0

5

10

B1 (sim)B2 (sim)Y1 (sim)Y2 (sim)B1 (meas)B2 (meas)Y1 (meas)Y2 (meas)

(b)

Fig. 21. Simulated and measured radiation patterns of the proposed

antenna at 10 GHz: (a) xz plane and (b) yz plane.


194

na integration were experimentally verified. Fig. 20 shows the

measured return loss characteristics of the proposed antennas.

As shown in Fig. 20, the 9.6 dB return loss bandwidths are

from 9.7 to 10.2 GHz for B1–B2 and from 9.1 to 10.2 GHz for

Y1–Y2. Fig. 21 shows the measured radiation patterns of the

proposed antenna. As shown Fig. 21, the radiation patterns for

B1–B2 are symmetric and mostly identical in the xz and yz

planes, respectively. On the other hand, the radiation patterns of

Y1–Y2 are mostly identical and symmetric in the xz and yz

planes, respectively as illustrated in Fig. 22. By evaluating the

measured gain difference between the two antennas of the same

type, the plot of B1–B2 (i.e., gain difference between B1 and B2

vs. theta) and Y1–Y2 (i.e., gain difference between Y1 and Y2

vs. theta) in both the yz and xz planes are similar to those in Fig.

4. In addition, the integrity of the proposed design is verified

because the simulated and measured results agree well with each

other. Therefore, the proposed antenna integration is suitable

for direction-finding. Furthermore, based on the result shown

in Fig. 22, the ranges of the detection angle (theta) for direc-

tion-finding in the yz and xz planes are excellent: |theta|30° and |theta| 45°, respectively. In addition, as shown in

Fig. 21, the simulated and measured results agree well with each

other except for a small range of angles, verifying the integrity of

the antenna integration design. The overall performance of the

proposed antenna integration is summarized in Table 2.

V. CONCLUSION

In this paper, we have proposed a compact, lightweight Rx

antenna integration suitable for 3D direction-finding radar. The

proposed antenna integration exhibits a return loss value under

9.6 dB in the frequency band range from 9.7 to 10.2 GHz. The

radiation patterns of the proposed antenna in both the yz and xz

planes are generally suitable for a direction-finding system. In

addition, the proposed antenna integration is both compact and

light. Therefore, when combined with an appropriate receiver

module for extracting DC voltage, it is a good antenna candi-

date for a direction-finding system for an avionic object.

This research was supported by the Agency for Defense

Development, Korea.

Theta[deg]

-60 -30 0 30 60

Gai

n di

ffer

ence

[dB

]

-15

-10

-5

0

5

10

15

B1-B2(measured)Y1-Y2(measured)B1-B2(simulated)Y1-Y2(simulated)

xz-plane

Theta[deg]

-60 -30 0 30 60

Gai

n di

ffer

ence

[dB

]

-30

-20

-10

0

10

20

30

B1-B2(measured)Y1-Y2(measured)B1-B2(simulated)Y1-Y2(simulated)

(a) (b)

Fig. 22. Comparison between simulated and measured gain difference between B1–B2 and Y1–Y2 at 10 GHz vs. theta: (a) xz plane and (b)

yz plane.

Table 2. Summary of the performance of the proposed antenna integration

B1 B2 Y1 Y2

VSWR 2:1 bandwidth (GHz) 9.8-10.3 9.7-10.2 9.1-10.2 9.2-10.4

Main beam (dBi@deg) 7.1@43 7.0@-42 7.5@50 7.4@-52

Half power beamwidth (deg)

xz-plane 59 61 49 50

yz-plane 65 63 57 55

Gradient of gain difference (dB/deg)

xz-plane Average 0.27 (relatively steep) Average 0.05 (relatively flat)

yz-plane Average 0.02 (relatively flat) Average 0.77 (relatively steep)


195

REFERENCES

[1] E. Ferrar and T. Parks, "Direction finding with an array of

antennas having diverse polarizations," IEEE Transactions

on Antennas and Propagation, vol. 31, no. 2, pp. 231-236,

1983.

[2] J. Yeo and J. I. Lee, "Miniaturized LPDA antenna for port-

able direction finding applications," ETRI Journal, vol. 34,

no. 1, pp. 118-121, 2012.

[3] J. H. Choi, C. S. Park, S. P. Nah, and W. Jang, "A multi-

channel correlative vector direction finding system using ac-

tive dipole antenna array for mobile direction finding appli-

cations," Journal of Electromagnetic Engineering and Sci-

ence, vol. 7, no. 4, pp. 161-168, 2007.

[4] A. Farina and H. Kuschel, "Guest editorial special issue on

passive radar (Part I)," IEEE Aerospace and Electronic Systems

Magazine, vol. 27, no. 10, pp. 5-5, 2012.

[5] H. D. Griffiths and C. J. Baker, "Passive coherent location

radar systems. Part 1: Performance prediction," IEE Pro-

ceedings-Radar, Sonar and Navigation, vol. 152, no. 3, pp.

153-159, 2005.

[6] C. J. Baker, H. D. Griffiths, and I. Papoutsis, "Passive co-

herent location radar systems. Part 2: Waveform proper-

ties," IEE Proceedings-Radar, Sonar and Navigation, vol. 152,

no. 3, pp. 160-168, 2005.

[7] P. Knott, "Design of a printed dipole antenna array for a

passive radar system," International Journal of Antennas and

Propagation, vol. 2013, article ID. 179296, 2013.

[8] H. Peng, Z. Yang, and T. Yang, "Design and implementa-

tion of a practical direction finding receiver," Progress In

Electromagnetics Research Letters, vo. 32, pp. 157-167, 2012.

[9] N. D. Au and C. Seo, "A novel design of an RF-DC con-

verter for a low–input power receiver," Journal of Electro-

magnetic Engineering and Science, vol. 17, no. 4, pp. 191-196,

2017.

[10] N. M. Mitrovic and M. M. Ponjavic, "Multichannel 2-D

direction finding based on differential amplitude detec-

tion," IEEE Sensors Journal, vol. 15, no. 9, pp. 5064-5070,

2015.

[11] Q. Zhu, S. Yang, R. Yao, and Z. Nie, "Direction finding

using multiple sum and difference patterns in 4D antenna

arrays," International Journal of Antennas and Propagation,

vol. 2014, article no. 392895, 2014.

[12] R. B. Dybdal, "Monopulse resolution of interferometric

ambiguities," IEEE Transactions on Aerospace and Electronic

Systems, vol. 22, no. 2, pp. 177-183, 1986.

[13] H. S. Setiadji, F. Y. Zulkifli, and E. T. Rahardjo, "Radia-

tion characteristics of microstrip array antenna for x-band

radar application," in Proceedings of 2016 Progress in Elec-

tromagnetic Research Symposium (PIERS), Shanghai, China,

2016, pp. 3992-3995.

[14] I. Slomian, K. Wincza, and S. Gruszczynski, "Low‐cost

broadband microstrip antenna array for 24 GHz FMCW

radar applications," International Journal of RF and Micro-

wave Computer‐Aided Engineering, vol. 23, no. 4, pp. 499-

506, 2013.

[15] C. Balanis, Antenna Theory, 3rd ed. Hoboken, NJ: John

Wiley & Sons, 2005.

[16] V. Bankey and N. A. Kumar, "Design of a Yagi-Uda an-

tenna with gain and bandwidth enhancement for Wi-Fi

and Wi-Max applications," International Journal of Anten-

nas, vol. 2, no. 1, pp. 1-14, 2016.

[17] Y. T. Lo and S. W. Lee, Antenna Handbook. New York,

NY: Van Nostrand Reinhold, 1993.

[18] S. Trinh-Van, S. C. Song, S. H. Seo, and K. C. Hwang,

"Waveguide slot array antenna with a hybrid-phase feed

for grating lobe reduction," International Journal of Anten-

nas and Propagation, vol. 2016, article ID. 4825924, 2016.

[19] Q. Fang, L. Song, and M. Jin, "Design and simulation of a

waveguide slot antenna," in Proceedings of 2012 5th Global

Symposium on Millimeter-Waves, Harbin, China, 2012, pp.

131-134.

[20] S. Pandit, A. Mohan, and P. Ray, "A low-profile high-

gain substrate-integrated waveguide-slot antenna with

suppressed cross polarization using metamaterial," IEEE

Antennas and Wireless Propagation Letters, vol. 16, pp.

1614-1617, 2017.

[21] Y. Jin, H. Lee, D. Woo, and J. Choi, "Low‐profile tilted‐

beam waveguide slot antenna using corrugations for UAV

applications," Microwave and Optical Technology Letters, vol.

60, no. 2, pp. 341-347, 2018.

[22] M. Beruete, I. Campillo, J. S. Dolado, J. E. Rodrguez-

Seco, E. Perea, F. Falcone, and M. Sorolla, "Very low pro-

file and dielectric loaded feeder antenna," IEEE Antennas

and Wireless Propagation Letters, vol. 6, pp. 544-548, 2007.

[23] M. Beruete, I. Campillo, J. S. Dolado, J. E. Rodriguez-

Seco, E. Perea, F. Falcone, and M. Sorolla, "Very low-

profile" Bull's Eye" feeder antenna," IEEE Antennas and

Wireless Propagation Letters, vol. 4, pp. 365-368, 2005.

[24] J. G. Chen, T. Zhang, J. S. Zhu, X. Y. Zhang, J. L. Zhou,

J. F. Fan, and G. H. Hu, "Low‐loss planar optical wave-

guides fabricated from polycarbonate," Polymer Engineering

& Science, vol. 49, no. 10, pp. 2015-2019, 2009.

[25] J. A. McGeough, M. C. Leu, K. P. Rajurkar, A. K. M. De

Silva, and Q. Liu, "Electroforming process and application

to micro/macro manufacturing," CIRP Annals, vol. 50, no.

2, pp. 499-514, 2001.


196

Hojoo Lee received his B.S. and M.S. degrees in Electronics

and Computer Engineering from Hanyang Univer-

sity in Seoul, Korea, in 2010 and 2015. He is cur-

rently working toward a Ph.D. degree in the De-

partment of Electronics and Computer Engineering

at Hanyang University in Seoul, Korea. His current

research interest focuses on various RF devices,

waveguide components and antenna designs, mainly

in the field of 5G mm wave and automotive application.

Jaesik Kim received B.S. and Ph.D. degrees in radio wave engi-

neering from Kwangwoon University and electrical

and electronic engineering from Yonsei University,

Seoul, Korea in 2011 and 2017, respectively. He

joined Agency for Defense Development, Daejeon,

Korea in 2017, where he is currently a Senior Re-

searcher. His current research interests include direc-

tion-finding antennas and phases array antennas.

Jaehoon Choi received the B.S. degree from Hanyang University,

Korea in 1980 and the M.S. and Ph.D. degrees from

Ohio State University, Ohio, USA in 1986, and

1989, respectively. From 1989 to 1991, he was a

research analyst with the Telecommunication Re-

search Center at Arizona State University, Tempe,

Arizona. He had worked for Korea Telecom as a

team leader of the Satellite Communication Division

from 1991 to 1995. Since 1995, he has been a professor in the Department

of Electronics and Computer Engineering at Hanyang University, Korea.

He has published more than 290 refereed journal articles and numerous

conference proceedings. He also holds over 90 patents. His research inter-

ests include antenna, microwave circuit design, and EMC. Currently, his

research is mainly focused on the design of a compact multiband antenna

for mobile wireless communication and antennas for biomedical applica-

tions.




197

I. INTRODUCTION

Optical imaging plays a significant role in observing defects in

objects [1–5]. Its main advantage lies in its broad observation

spectrum characteristics and high resolution. However, optimal

imaging faces difficulties in observing an object at subsurface

levels due to scattering photons and stops at a tissue depth of

about 10 μm. This means that the bulk of photons propagating

through a tissue slice will experience at least one scattering event,

resulting in image blur [6].

Acoustic modality has also been commonly used in subsur-

face imaging because of its better subsurface attenuations, which

utilize piezoelectric materials as transducers for both transmit-

ters and receivers. However, the selection of piezo material must

be carefully considered to avoid mismatched impedance and

yield optimal efficiency.

To yield optimal efficiency, the air-couple transducer has

been developed by combining several impedance-matching ma-

terials so that the increase of impedance can be adjusted gradu-

ally. However, the minimum focal diameter of such technology

is still greater than 1 mm [7, 8]. Therefore, the air-couple tech-

nique for acoustic imaging still leads to problems when dealing

with micro objects.

Photoacoustic (PA) methods have overcome the challenges

found in acoustic coupling by using laser light to excite and

probe acoustic disturbances without contacting the sample [9,

10]. Moreover, the laser-driven piezoelectric technology with

femtosecond laser pulse has successfully generated acoustic

Photoacoustic Tomography System for Roughly

Painted Micro Objects Andreas Setiawan1,2,* · Fransiscus Dalu Setiaji3 · Gunawan Dewantoro3 · Nur Aji Wibowo1,2

Abstract

Subsurface imaging is challenging; it is difficult to detect objects visually. In this research, a novel non-contact photoacoustic (PA) imag-

ing system was developed to detect subsurface objects. The Rosencwaig-Gersho (RG) model was successfully employed to capture micro-

object images covered by rough paint. The experiments were conducted using a copper ring with a 1-mm diameter fully coated by rough

paint with an average thickness of 2.3 μm. The resulting PA images exhibited up to 72% consistency despite the rough paint; the shapes

of the objects were clearly recognized before and after coating. To conduct the experiment, simulations and image acquisitions were ar-

ranged. Then, the system capability to produce tomographic images was improved by adjusting the thermal diffusion lengths, and subsur-

face object images were successfully acquired at depths of 2.0, 2.6, 9.8, and 52 μm. The detailed composition of image slices displayed the

structure profile of subsurface objects appropriately.

Key Words: Imaging, Mesoscopic, Photoacoustic, Subsurface, Tomography.

Manuscript received January 24, 2019 ; Revised May 03, 2019 ; Accepted June 10, 2019. (ID No. 20190124-005J) 1Department of Physics, Universitas Kristen Satya Wacana, Salatiga, Indonesia. 2Study Center for Multidisciplinary Applied Research and Technology, Universitas Kristen Satya Wacana, Salatiga, Indonesia. 3Faculty of Electronic and Computer Engineering, Universitas Kristen Satya Wacana, Salatiga, Indonesia. *Corresponding Author: Andreas Setiawan (e-mail: [email protected])





198

waves in order of terahertz [11].

PA has been reported to detect surface defects in metals [12–

14]. With an optical resolution photoacoustic microscopy (OR-

PAM) system configuration and water coupling, a full-width at

half-maximum (FWHM) of up to 2.5 μm has been accom-

plished. Nevertheless, defects can also occur at the subsurface

level, and these defects are challenging to detect due to invisibil-

ities [15, 16]. Therefore, PA imaging has been used satisfactori-

ly at subsurface levels for applications such as detecting hidden

underdrawings in paintings [17]. Indeed, the drawing of a pen-

cil coated with paint can be clearly restored through PA imag-

ing.

In this experiment (subsurface imaging of hidden underdraw-

ings in paintings), direct contact with the object via a water-

couple transducer was necessary. Some alternative methods have

been employed to create non-contact PA imaging; for example,

the all-optical PA microscopy method uses a Fabry-Perot

transducer customization [18]. In this study, a novel non-

contact PA imaging system was proposed to capture subsurface

images without transducer customizations. A PA imaging

mesoscopic system and tomography for micro objects using a

focused laser beam was developed [19], and the application of

the Rosencwaig-Gersho (RG) model to acquire subsurface im-

ages was successfully demonstrated [20].

By means of thermal diffusion length adjustment, imaging at

a certain depth was accomplished, which generated non-contact

tomographic images for micro objects. The experiment object

was a copper ring coated by rough paint, and the results showed

that the original object shape could be restored completely.

Comparisons between pre-coated and coated objects were also

performed. The PA images gave superior consistency compared

to that of the optical images. Additionally, the system capability

to capture tomographic images of an object was also demon-

strated. By varying the thermal diffusion length, images could

be constructed layer by layer, generating the subsurface structure

of the object. The complete experiment results are presented in

Section III.

II. EXPERIMENTAL METHOD

1. Materials

The arrangements of the systems are shown in Fig. 1(a). A

modulated laser diode was used to generate thermal contraction

of the object. To adjust the thermal diffusion length, a comput-

er-controlled signal generator was employed. Fig. 1(b) shows

the interaction between the laser and the coarse surface. There

were numerous possible paths, including path 1 and 2.

Throughout path 1, the interaction between the laser beam and

the material occurred only once, which resulted in a normal

Fig. 1. The proposed PA imaging system. (a) The focused laser

formed a mesoscopic system in an open cell configura-

tion. (b) For coarse objects, the surface contour resulted

in an increasing absorption coefficient (path 2).

absorption coefficient (i.e., dielectric function of the Drude

model).

Although the electronic interactions were unchanged, the ab-

sorption coefficient increased significantly along path 2. This

was because the surface geometrical profile captured the light,

which led to repeated reflections. A more detailed experiment of

this kind was performed in [21], in which the surface roughness

level more significantly influenced the absorption coefficient

compared to the laser wavelength variation. This conclusion

remained consistent even when tested under wider ranges (532–

1,064 nm).

Regarding the PA process, this absorption coefficient deter-

mined the acoustic initial pressure: p0 = ΓηthβF, where Γ, ηth, β,

and F are the Gruneisen parameter, energy conversion efficiency,

optical absorption coefficient, and laser optical fluence, respec-

tively [22]. Therefore, the PA response on the rough surface

was assumed to be insensitive to the laser wavelength variation,

and the experiments were conducted using only one wavelength

(805 nm, ML520G12; Mitsubishi Electric). A Behringer

ECM8000 microphone was utilized to acquire the acoustic

waves with a sensitivity of -39.2 dB (11 mV/Pa). The micro-

phone was mounted 10 mm away from the object, forming a 45°

angle with the normal line. The collimator lens used was the

Mitutoyo Objective Lens BD Plan Apo 7.5×, whose parame-

ters are specified in Table 1.

The tested material was a copper ring with a thickness of 0.05

mm and an outer diameter of 1 mm. The ring lay on the surface

of an acrylic printed circuit board. Different object and back-

ground materials were chosen to acquire better PA image con-

Table 1. Optical parameters of the lens

Numerical aperture 0.21

Lens focal length (mm) 26.7

Beam divergence (mrad) 87.5

Beam spot size (μm) 9

SETIAWAN et al.: PHOTOACOUSTIC TOMOGRAPHY SYSTEM FOR ROUGHLY PAINTED MICRO OBJECTS

199

trast resulting from different thermal properties. As a compari-

son, a standard air-couple-based imaging system—namely, the

Examet Union 81837 digital microscope (Ogawa Seiki Co.

Ltd.)—was used.

2. PA Multilayer Analysis

According to the RG model, the PA amplitude (QPA) corre-

lates with the thermal diffusion length of the material. Thus,

QPA specifies the complex envelope of the sinusoidal pressure

variation in the form of acoustic magnitude generated by the

PA process. For opaque materials (μa << l, μa << μs), QPA is pro-

portional to two states, as follows:

12

12

j

j 1

1

j

PA

j j

j j

, for 2

, for 2

j

lf k

lf k

Q

(1)

where f, α, and k are laser modulation frequency, diffusivity, and

thermal conductivity of the material, respectively [23]. Indices a

and s in μ represent air and solid. Afterward, index j is used to

represent the order of solid layers. Meanwhile, μj and lj are

thermal diffusion length and material thickness, respectively.

The first state occurs when the material is under thermally thin

conditions (e.g., μj > lj), whereas the contrary condition is called

thermally thick. The two conditions can be achieved by adjust-

ing laser modulation frequency using the following equation:

1 2

jj f

(2)

Eq. (1) shows that the PA amplitude depends on the com-

parison of the thermal diffusion length and the thickness of the

material. Therefore, any alteration of layers can be discovered by

gradually varying the modulation frequency. When the modula-

tion frequency satisfies the condition μj < lj, the PA amplitude

can be determined by the material at layer j. However, when the

frequency is varied until it satisfies the condition μj > lj, the am-

plitude changes drastically according to the (j + 1)th layer’s char-

acteristics.

III. RESULTS AND DISCUSSION

1. Modeling and Simulation

The interaction scheme between the thermal diffusion length

and the paint depth is shown in Fig. 2(a). Rough paint was

characterized by the thickness of the paint (lpaint) that enclosed

the object surface. According to Eq. (1), three possible condi-

tions existed: μpaint < lpaint, μpaint = lpaint, and μpaint > lpaint. The first

Fig. 2. (a) The resulting amplitude QPA due to various surface rou-

ghness. (b) The cross section of the roughly-coated copper

ring model on 500 × 500 pixels. Simulations of PA images

with a roughness mean of 2.0 μm (c) and 3.0 μm (d), where

μpaint = 2.46 μm.

condition took place when the paint coat was thicker than its

thermal diffusion length and the PA amplitude was determined

by the first material layer, QPA (μpaint, kpaint). When the paint

coating layer was thinner than its thermal diffusion length, the

third condition was satisfied, and the amplitude became QPA

(μobject, kobject). Theoretically, despite the variation of the paint

depth, the PA amplitude did not respond directly. Any altera-

tion occurred only when the paint coat thickness exceeded the

threshold of the thermal diffusion length. This became the

principal difference between optical and PA imaging due to

insensitivity to the surface roughness. The second condition

occurred when the paint coat thickness was the same as the

thermal diffusion length. Then, the QPA experienced a transi-

tion from the preceding two conditions.

To run simulations, a number of computations were per-

formed using GNU Octave version 5.1.0 programming. To

generate a PA image, Eq. (1) was employed with each associat-

ed material parameter. To establish the object, an array was

constructed with a dimension of 500 × 500 pixels, on which a

copper ring was laid as shown in Fig. 2(b). For simulation pur-

poses, the rough paint was assumed to have a normal distribu-

tion with a mean of 2 μm and 3 μm. These values were chosen

to approach the thermal diffusion length of the paint, which

was 2.46 μm (αpaint = 2.1 × 10-7 m2/s at modulation frequency

of 11 kHz). Thus, the distributions above and below the mean

were expectedly balanced.

The PA image resulting from the simulation using these pa-

rameters is depicted in Fig. 2(c) for μpaint > 𝑙 , where 𝑙

was the average thickness of the paint. The copper region gave

weaker QPA in the order of about 0.39 × 10-9 (a.u). At this state,

the shape of the copper ring was displayed clearly because its


200

thermal property significantly affected the QPA. Since the ther-

mal conductivity of copper was far different than that of the

background (kcopper = 386 W/mK and kacrylic = 0.40 W/mK), the

pixel intensity level could be distinctly discriminated and there-

fore yielded a higher image contrast.

When 𝑙 was increased to 3.0 μm, the results changed,

as shown in Fig. 2(d). At this state, the condition of μpaint <

𝑙 was more likely to happen, and the amplitude QPA was

dominated only by the paint’s thermal property. Since the ther-

mal conductivity of the paint (kpaint = 1.45 W/mK) was much

closer to that of the background (acrylic), the resulting QPA pa-

rameters were similar in the order of 4.5 × 10-9 (a.u). Due to

this similarity of pixel intensity levels, the image contrast de-

creased, and the copper ring image became blurred.

2. Roughly Painted Object Imaging

Fig. 3(a) shows the object image prior to being coated. By

adhering to the prior simulation parameters, a modulation fre-

quency of 11 kHz and a rough paint coating with an average

depth of 2.3 μm were used. A commercial acrylic epoxy spray

paint (Dull Black-4, RJ London) was applied gradually. The

subsequent layer was applied immediately after the previous

layer to avoid multilayer coating. After accomplishing the coat-

ing process, the average paint depth was measured using the

Mitutoyo 193-111 micrometer.

At the initial stage, a single object point was measured, and

the signal was recorded for 0.5 seconds. Then, Fourier trans-

formation was performed to obtain the signals spectrum (Fig.

3(b)). Since an open cell configuration was used, the position of

the microphone could be arranged to obtain the optimum result.

The signal QPA was intensely detected with a peak of 11 kHz in

accordance with the modulation frequency. The spectrum ex-

hibited a microphone response up to 20 kHz with a background

intensity of -70 dB and a peak of up to -57 dB. Due to this

large difference, the PA signal could be detected directly.

Prior to measurements, the object image was captured by

Fig. 3. (a) Photograph of the copper ring prior to being coated, as

indicated by the arrow. (b) The PA signal response with a

modulation frequency of 11 kHz.

means of a microscope. The optical pre-coated object image is

shown in Fig. 4(a). Visually, the image possessed an excellent

quality with high and evenly distributed contrast so that the

object shape could be acquired clearly. However, a different

result arose for the coated object, as depicted in Fig. 4(b). The

rough paint coat created many disturbances in the form of light

reflections and randomly distributed high absorbance; as a con-

sequence, the object was barely displayed.

Next, the experiments for PA imaging were conducted using

same object. The first scanning result is shown in Fig. 4(d). The

PA image showed a high contrast, as was expected from the

prior simulation. The object shape was easy to recognize and,

accordingly, could be shown as a whole.

Fig. 4(e) shows the PA image of the roughly coated object.

Both PA images did not indicate significant changes visually,

regardless of the presence of coating. To test this consistency,

image subtractions for both modalities were conducted after

normalizing each image. The result for the optical images is

shown in Fig. 4(c). The inconsistency of the object intensities

before and after the coating process were shown; thus, the im-

age subtraction failed to display the object completely. Unlike

the optical image, the PA images gave better consistency, and

the corresponding image subtraction displayed the real object

shape, as shown in Fig. 4(f).

Furthermore, line spread function (LSF) curves were estab-

lished for all images, as shown in Fig. 4(g) and (h). As indicated

before, the optical image had a large change of LSF curves (Fig.

Fig. 4. Object image captured by microscope (a, b, c) and PA (d, e,

f). (a) and (d) are real object images. (b) and (e) are images

of the coated objects. (c) and (f) are subtractions of the two

previous images, followed by normalization. (g) and (h) are

LSF curves for optical and PA image, respectively. Index

“sub” corresponds to LSF for roughly painted and coated

objects.


201

4(g)). Therefore, it could be inferred that the curve widened due

to the rough paint coating. The FWHM measurement implied

a change of values from 4 μm to 34 μm, giving a consistency of

only 12%. The PA images, however, had a smaller change (i.e.,

from 55 μm to 76 μm), giving a consistency of up to 72%.

3. Tomography Object under Paint

The tomography process was carried out by adjusting the

penetration depth by changing the value of the thermal diffu-

sion length of the first layer (paint), as given by Eq. (2). In this

experiment, four different frequencies were employed: 15.8 kHz,

12.5 kHz, 9.7 kHz, and 1.5 kHz. These corresponded to ther-

mal diffusion lengths of 2.0 μm, 2.6 μm, 9.8 μm, and 52 μm.

These frequencies were chosen so that the image slices were

distributed to distinct layer depths. Then, the laser beam energy

for each modulation was measured using a laser power meter

(Sanwa LP1); the results are shown in Table 2.

The resulting tomographic images associated to object depths

are shown in Fig. 5(a). When the thickness of the paint coating

and copper ring were 2.3 μm and 50 μm, respectively, the image

slice for the inner part of the paint coat could be obtained with a

modulation of 15.8 kHz (Fig. 5(b)). When the frequency was

decreased to 12.5 kHz, the image slice arose from the paint–

ring junction, as shown in Fig. 5(c). Since it the image slice was

closer to the ring surface, several scratches were seen on the im-

age, as indicated by the arrows in Fig. 5(c). Furthermore, when

the frequency was decreased to 9.7 kHz, the middle of the cop-

Table 2. Modulated laser beam energy

Frequency (kHz) Energy per time (μJ/s)

1.5 944

9.7 47

12.5 31

15.8 22

Fig. 5. The tomography process used four different frequencies,

which were 15.8, 12.5, 9.7, and 1.5 kHz, corresponding to

μpaint of 2.0, 2.6, 9.8, and 52 μm, respectively.

per ring could be observed (Fig. 5(d)).

At this depth, the copper’s thermal property dominated the

establishment of the PA amplitude, as in Eq. (1). Therefore, the

resulting image was different from the two preceding images.

Compared to the previous images, the intensity of the ring was

darker than its background because the thermal conductivity of

copper was higher than that of the paint, and the magnitude

QPA decreased. This transition process also verified that the to-

mography process occurred at the desired layer. When the fre-

quency was further decreased to 1.5 kHz, the thermal diffusion

length of the copper ring surpassed its thickness so that PA im-

aging produced the base layer image, as shown in Fig. 5(e). At

this frequency, the thermally thin condition was satisfied.

Therefore, the copper ring became photo-acoustically transpar-

ent. In this transparent condition, PA imaging penetrated

through the ring and captured the image below it.

IV. CONCLUSION

A PA imaging system was developed for opaque micro ob-

jects coated by rough paint. This imaging method was robust

against disturbances and was able to display the objects clearly.

Furthermore, it demonstrated effectiveness in providing tomo-

graphic images by means of thermal diffusion length according

to the RG model. By using this simple method, the three-

dimensional condition of the object could be reconstructed in

accordance to the desired depth. Despite the aforementioned

facts, the PA FWHM value was still much lower than that of

the optical image. This was mainly due to the X-Y stage accura-

cy used during the scanning process; this accuracy enhancement

was expected to improve the captured PA image.

This work was supported by the Universitas Kristen Satya

Wacana through Fundamental Research Grant with S.K.

No. 062/Penel./Rek./3/V/2019.

REFERENCES

[1] J. Rupil, S. Roux, H. Hild, and L. Vincent, "Fatigue mi-

crocrack detection with digital image correlation," The Jour-

nal of Strain Analysis for Engineering Design, vol. 46, no. 6,

pp. 492-509, 2011.

[2] M. Suzuki, K. Ogawa, and T. Shoji, "Quantitative NDE of

surface cracks in ceramic materials by means of a high-

frequency electromagnetic wave," Materials Transactions, vol.

47, no. 6, pp. 1605-1610, 2006.

[3] T. Hoshimiya, "Nondestructive evaluation of surface defects

under dry/wet environment by the use of photoacoustic and

photothermal electrochemical imaging," NDT & E Interna-


202

tional, vol. 32, no. 3, pp. 133-137, 1999.

[4] M. Feng, L. Zou, and M. Imai, "Detection of ceramic

cracks using a distributed high-resolution Brillouin fiber op-

tic sensor," SICE Journal of Control, Measurement, and System

Integration, vol. 3, no. 4, pp. 279-284, 2010.

[5] F. T. Dagefu and K. Sarabandi, "High resolution subsurface

imaging of deep targets based on distributed sensor net-

works," in Proceedings of 2009 IEEE International Geoscience

and Remote Sensing Symposium, Cape Town, South Africa,

2009, pp. 258-261.

[6] V. Ntziachristos, "Going deeper than microscopy: the opti-

cal imaging frontier in biology," Nature Methods, vol. 7, no.

8, pp. 603-614, 2010.

[7] T. E. G. Alvarez-Arenas, J. Camacho, and C. Fritsch, "Pas-

sive focusing techniques for piezoelectric air-coupled ultra-

sonic transducers," Ultrasonics, vol. 67, pp. 85-93, 2016.

[8] R. Kazys, R. Sliteris, and J. Sestoke, "Air-coupled low fre-

quency ultrasonic transducers and arrays with PMN-32%

PT piezoelectric crystals," Sensors, vol. 17, article no. 95,

2017.

[9] J. A. Rogers, A. Maznev, and K. A. Nelson, "Impulsive

stimulated thermal scattering," in Characterization of Mate-

rials. New York, NY: John Wiley & Sons Inc., 2012.

https://doi.org/10.1002/0471266965.com063.pub2.

[10] X. L. Dean-Ben, G. A. Pang, F. Montero de Espinosa,

and D. Razansky, "Non-contact optoacoustic imaging

with focused air-coupled transducers," Applied Physics Let-

ters, vol. 107, article no. 051105, 2015.

[11] A. A. Maznev, K. J. Manke, K. H. Lin, K. A. Nelson, C.

K. Sun, and J. I. Chyi, "Broadband terahertz ultrasonic

transducer based on a laser-driven piezoelectric semicon-

ductor superlattice," Ultrasonics, vol. 52, no. 1, pp. 1-4,

2012.

[12] A. Setiawan, G. B. Suparta, M. Mitrayana, and W.

Nugroho, "Surface crack detection with low-cost photoa-

coustic imaging system," International Journal of Technolo-

gy, vol. 9, no. 1, pp. 159-169, 2018.

[13] A. Setiawan, G. B. Suparta, M. Mitrayana, and W. Nug-

roho, "Subsurface corrosion imaging system based on LA-

SER generated acoustic (LGA)," International Journal on

Advanced Science, Engineering and Information Technolo-

gy, vol. 7, no. 6, pp. 2189-2196, 2017.

[14] S. Jeon, J. Kim, J. P. Yun, and C. Kim, "Non-destructive

photoacoustic imaging of metal surface defects," Journal of

Optics, vol. 18, article no. 114001, 2016.

[15] B. Saleh, Introduction to Subsurface Imaging. Cambridge,

UK: Cambridge University Press, 2011.

[16] W. Geary and J. Hobbs, "Catastrophic failure of a carbon

steel storage tank due to internal corrosion," Case Studies in

Engineering Failure Analysis, vol. 4, no. 1, pp. 257-264,

2013.

[17] G. J. Tserevelakis, I. Vrouvaki, P. Siozos, K. Melessanaki,

K. Hatzigiannakis, C. Fotakis, and G. Zacharakis, "Photo-

acoustic imaging reveals hidden underdrawings in paint-

ings," Scientific Reports, vol. 7, article no. 747, 2017.

[18] S. L. Chen, L. J. Guo, and X. Wang, "All-optical photoa-

coustic microscopy," Photoacoustics, vol. 3, no. 4, pp. 143-

150, 2015.

[19] D. Soliman, G. J. Tserevelakis, M. Omar, and V. Ntzia-

christos, "Combining microscopy with mesoscopy using

optical and optoacoustic label-free modes," Scientific Re-

ports, vol. 5, article no. 12902, 2015.

[20] A. Rosencwaig and A. Gersho, "Theory of the photoa-

coustic effect with solids," Journal of Applied Physics, vol. 47,

no. 1, pp. 64-69, 1976.

[21] D. Bergstrom, "The absorption of laser light by rough

metal surfaces," Ph.D. dissertation, Lulea Tekniska Uni-

versitet, Sweden, 2008.

[22] J. Yao and L. V. Wang, "Sensitivity of photoacoustic mi-

croscopy," Photoacoustics, vol. 2, no. 2, pp. 87-101, 2014.

[23] S. I. Yun and H. J. Seo, "Photoacoustic measurements of

thermal diffusivity or thickness of multi-layer solids," Le

Journal de Physique Colloques, vol. 44, no. C6, pp. C6.459-

C6.462, 1983.


203

Andreas Setiawan is a lecturer at the Department of Physics, Universi-

tas Kristen Satya Wacana. He received a bachelor’s

degree in physics instrumentation from the Universi-

tas Kristen Satya Wacana, Indonesia, in 1999 and a

master’s degree in instrumentation and control from

the Institute of Technology, Bandung, Indonesia. He

received a doctorate at the Department of Physics,

Gadjah Mada University, Indonesia, focusing on

photoacoustic instruments for practical applications.

Fransiscus Dalu Setiaji received a bachelor’s degree in electrical engineering

at the Universitas Kristen Satya Wacana, and his

master’s degree is in the same field from Gadjah

Mada University. He currently works as a lecturer at

the Department of Electrical and Computer Engi-

neering, Universitas Kristen Satya Wacana. His

interests include teaching and research in photoa-

coustic imaging.

Gunawan Dewantoro received his bachelor’s degree from Gadjah Mada

University, Yogyakarta, in 2008. He earned his mas-

ter’s degree from National Taiwan University of

Science and Technology in 2011. His interests in-

clude control theory, optimizations, and intelligent

systems. He has been with Universitas Kristen Satya

Wacana since 2012.

Nur Aji Wibowo received his bachelor’s and master’s degrees in phys-

ics from Sebelas Maret University, Indonesia, in

2007 and 2011, respectively. He currently works as a

lecturer in the Physics Department of the Faculty of

Science and Mathematics in Universitas Kristen

Satya Wacana, Indonesia. His fields of research

interest include optical thin film, spintronic, nano-

magnetic particles, and magnetized fuel.




204

I. INTRODUCTION

The radar remote sensing of the sea status is crucial for wea-

ther forecasting, navigation, and scientific studies. Radar back-

scatter from the ocean surface depends on the surface roughness

and the dielectric constant of the water, in which the roughness

of the sea surface mainly depends on wind speed and wind di-

rection [1]. The measurement and characterization of the sea

surface roughness is difficult, and so is the computation of the

backscattering coefficient of a sea surface.

The backscattering coefficients of various sea surfaces were

measured using spaceborne and airborne scatterometers [2–4].

The satellite synthetic aperture radar systems also provided a

tremendous amount of radar backscatter data for various fre-

quencies, angles, polarizations, and wind vectors [5–7]. Owing

to the difficulty of theoretical modeling, attempts have been

made for empirically modeling the backscattering coefficient of

sea surfaces based on the radar measurements for various wind

speeds and wind directions, the so-called “geophysical model

functions (GMF)” such as SASS-2 for Ku-band [8], XMOD2

for X-band [7], CMOD5 for C-band [9], and L-GMF for L-

band [6]. These empirical models are based on the functional

form of σ0=A+Bcosϕ+Ccos2ϕ, where ϕ is the azimuth angle

of the wind direction relative to the radar look direction. The

coefficients A, B, and C of each GMF model are determined

empirically based on the experimental data sets for each fre-

quency band as functions of radar parameters (frequency, angle,

and polarization) and wind speed. However, each of these em-

pirical models has limited ranges of frequencies, polarizations,

incidence angles, and wind vectors.

To obtain a wider validity region of a radar scattering model

for various radar and sea-surface parameters, theoretical models

of the backscattering coefficients for sea surfaces were developed

on the basis of the approximated surface roughness characteris-

tics and the integral equation method (IEM) model [10–13].

Surface roughness moments are related to its second-order and

third-order cumulant functions known as correlation and bi-

A Simple Empirical Model for the Radar Backscatters

of Skewed Sea Surfaces at X- and Ku-Bands Taekyeong Jin · Yisok Oh*

Abstract

This paper presents a new simple empirical model for the backscattering coefficients of skewed sea surfaces. Instead of using a complicat-

ed bispectrum function for the backscattering coefficient of a skewed sea surface, a simple modifying term is multiplied to the auto-

correlation length of the surface. The unknown constants of the modifying term are obtained by com-paring the simple empirical model

with an extensive database for various wind speeds, incidence angles, polarizations, and frequencies. The accuracy of the new model is

verified using measurement datasets.

Key Words: Backscattering, Simple Empirical Model, Skewed Sea Surfaces, Wind Direction.

Manuscript received November 23, 2018 ; Revised March 27, 2019 ; Accepted June 12, 2019. (ID No. 20181123-079J)

Department of Electronic and Electrical Engineering, Hongik University, Seoul, Korea. *Corresponding Author: Yisok Oh (e-mail: [email protected])

Current affiliation of the first author (Taekyeong Jin): Hanwha Systems, Gumi, Korea.




JIN and OH:A SIMPLE EMPIRICAL MODEL FOR THE RADAR BACKSCATTERS OF SKEWED SEA SURFACES AT X- AND Ku-BANDS

205

coherence functions, respectively. The Fourier transforms of the

correlation and bicoherence functions are the roughness spec-

trum and bispectrum. The sea surface can be modeled as a time-

varying, anisotropic, and skewed rough surface, in which “skew-

ness” is defined by the difference in the backscattering coeffi-

cients between the upwind (ϕ=0°) and downwind (ϕ=180°)

directions, and “peakedness” is defined by the difference be-

tween the upwind (ϕ=0°) and crosswind (ϕ=90°) directions.

The surface skewness is described by the bicoherence or the

bispectrum. Then, the backscattering coefficient of a sea surface

can be estimated by adding a term associated with the rough-

ness spectrum (correlation function) and the other term with

the bispectrum (bicoherence) [1, 10]. The first term can be ob-

tained by the traditional IEM, with the roughness spectrum

corresponding to a correlation function with a root-mean-

square (RMS) height and a correlation length. The second term

may be obtained with a complicated bispectrum computation

for the skewness effect. The bispectrum calculation involves one

unknown parameter s0 (skewness factor) that may be empirically

determined for a given wind speed.

To avoid computing the complicated bispectrum function

B(k)=BS(k)+jBa(k) with an unknown parameter (skewness fac-

tor), which should be empirically determined for each wind

speed, we attempted to generate an empirical expression for

representing the skewness effect. In the computation of the first

term of the theoretical model [10] associated with the correla-

tion function, the correlation length of a sea surface is represent-

ed by the function of the correlation length of the upwind (Lu)

and crosswind (Lc) directions and the azimuth angle of wind

direction ϕ. We found that the correlation length could also

include the skewness effect without the complicated second

term when the correlation length is multiplied by a Gaussian-

type functional form of the wind azimuth angle ϕ. In other

words, the skewness effect can be practically accounted for by

controlling the correlation length of a sea surface. The coeffi-

cients of the new multiplying function can be obtained by com-

paring the new model with an extensive experiment database.

Unlike the existing model, which is complicated and has skew-

ness factors that are determined for each wind speed, this new

model is simple without the complicated bispectrum function

and needs only the empirical determination of the coefficients

for each frequency and polarization for a wide range of wind

speeds. The accuracy of this new model was verified with the

extensive experiment database at X-band. Surprisingly, the

model agreed well with the Ku-band independent radar meas-

urements [14].

II. THE EXISTING MODEL

The IEM model has been widely used for estimating the

backscattering coefficients of rough surfaces including rough sea

surfaces [11]. The IEM model was applied to the measurement

data in [2, 4] at X- and Ku-bands for VV and HH polarizations

for various wind speeds and incidence angles using a simple

exponential correlation function [11]. The IEM consists of two

parts: 𝜎 𝜎 (N)+𝜎 (S), where the first part represents

the backscattering coefficients of the non-skewed surfaces, and

the second part represents the skewness effect.

2

0 2 2 211exp 2 cos

2pp rms

kN k h

2

( )

1

| |,

!

npp n

n

IW K

n

,

(1a)

2

20 2 2 211exp 4 cos

2pp pp rms

kS f k h

3 31 ( )

1

8 cos,

!

n

na

n

kB K

n

2

2 2 211Re * exp 3 cos

2 pp pp rms

kf F k h

3 31 ( )

1

3 cos,

!

n

na

n

kB K

n

3 311 (2 )

cos1 ,

2 !

n

n ns

kB K

n

2

2 2 2 211exp 2 cos

8 pp rms

kF k h

3 31 ( )

1

cos,

!

n

na

n

kB K

n

,

(1b)

where Ipp, W(K, ϕ), fpp, Fpp, and B(K,ϕ) are given in [11], k1is the

wavenumber, θ is the incidence angle, and hrms is the RMS

height of a sea surface. The W(n)(K, ϕ) is the nth-order surface

roughness spectrum, which is the Fourier transform of the nth-

power of a correlation function. For an exponential correlation

function, the roughness spectrum has the form of W(n)(K,

ϕ)=nlc[n2+(Klc)2]-1.5 with K=2k1sinθ and lc=Lucos2ϕ+Lcsin2ϕ,

where the correlation length lc can be obtained from the upwind

and crosswind correlation lengths, Luand Lc, respectively, with

the azimuth angle ϕ. The upwind and crosswind correlation

lengths are given in [11] for the various wind speeds. As previ-

ously mentioned, a skewness factor s0 is used when describing

the skewness effect. This constant is empirically determined at a

wind speed, and thus it does not have the physical characteris-

tics for wind speed. Moreover, Eq. (1b) is complicated, and the

accessibility is poor. Therefore, we proposed a new simple mod-

el that does not need Eq. (1b) and has the physical characteris-

tics for wind speed.

JOURNAL OF ELECTROMAGNETIC ENGINEERING AND SCIENCE, VOL. 19, NO. 3,JUL. 2019

206

III. NEW SIMPLE EMPIRICAL MODEL

In general, the RMS height and correlation length are mainly

used to describe the roughness of a rough surface. As the

roughness increases, the backscattering increases. As the RMS

height increases and/or the correlation length decreases, the

roughness increases. Therefore, if the correlation length de-

creases, the backscattering increases. We also used the rough-

ness characteristic of the correlation length increasing in the

order of upwind, downwind, and crosswind. Using these char-

acteristics, we proposed a modified correlation length, which is

the function of the azimuth angle. The peakedness and skew-

ness of the backscattering coefficients at the azimuth angles of 0°

≤ ϕ≤ 360° can be controlled by the surface correlation length.

Therefore, we proposed the following form of a modified corre-

lation length with unknown constants: a, b, and c.

2

2 2cos sin expc u cl L L a cb

(2)

For example, Fig. 1 shows the typical angular responses of the

backscattering coefficient 𝜎 (N) at a wind speed of 9.3 m/s,

f=10 GHz, and θ=32° with various values of the unknown

constants a, b, and c. The constant a in Eq. (2) mainly controls

the skewness effect, as shown in Fig. 1—the solid line (a=0.5)

and the dashed line (a=0.1) with fixed b and c. The comparison

between the solid line (b=3) and the dotted line (b=1) shows

that the constant b controls the peakedness (Fig. 1). The role of

the constant c is mainly to control the level of the backscattering

coefficient, as shown in Fig. 1—the solid line (c=0.7) and the

dash-dot line (c=0.3) with fixed a and b. Therefore, the skew-

ness and peakedness can be simply controlled by selecting the

optimum values of a, b, and c.

We used measurement data [1] to propose the model using

Eq. (2). The data used were measured in the sea near Japan and

Fig. 1. Backscattering coefficients 𝜎 (N) with various values of

unknown constants a, b, and c.

the Pacific Ocean using an airborne microwave scatterometer–

radiometer system and showed how the backscattering coeffi-

cients changed for the wind direction. The backscattering coef-

ficients were measured as combined functions of the microwave

frequency (10.0 GHz and 13.9 GHz), polarization (VV and

HH), incident angle (0°–70°), azimuth angle (0°–360°), and

wind speed (3.2–19.4 m/s). In the data processing procedure for

obtaining one point of data, the standard deviations for the

wind direction were less than 1°, and the standard deviations for

the backscattering coefficients were 0.3–1.0 dB.

An extensive database was obtained from [11] (originally [2,

4]), which consisted of 462 measurement data-points for VV

and HH polarizations at various wind speeds (i.e., 3.2, 9.3, and

14.5 m/s at 10 GHz measurements at 32° and 52°, and 4.2, 5.5,

7.5, 12, 15, and 19.4 m/s for 13.9 GHz measurements at 40°).

The optimum values of the unknown constants were obtained

using the minimum root-mean-square errors (RMSEs) between

the scattering model and the database as functions of wind

speed u in the following forms:

0.01 0.47

0.12 4.90 for VV-polarization, and

0.80

a u

b u

c

(3a)

0.02 0.92

0.09 2.08 for HH-polarization.

0.73

a u

b u

c

(3b)

If we use the correlation length using Eqs. (2) and (3), we can

control the skewness effect without using the complex Eq. (1b).

In the existing model, the skewness factor s0, which was ob-

tained empirically for a wind speed, was applied, but the new

model with (3) became much simpler because the correlation

length is a function of wind speed. We can also predict the

backscattering coefficients for continuous wind speeds.

IV. VERIFICATION OF THE NEW MODEL

To verify the accuracy of the new simple model that was em-

pirically developed with Eqs. (2) and (3), we compared the new

model with the existing model and the experimental datasets.

For the existing model, the complicate skewness term 𝜎 (S)

with the unknown skewness factor s0 was used. Conversely, the

new model is simpler than the existing model for computing the

backscattering coefficients of skewed sea surfaces.

Figs. 2 and 3 show a typical comparison among the new sim-

ple empirical model, the existing model, and the measurement

data. Both the new and existing models agree well with the

measurements except for several points at the crosswind direc-

tion, as shown in Fig. 2. The accuracies of both models are

comparable, as shown in Figs. 2 and 3.

back

scat

terin

g co

effic

ient

, dB


207

Fig. 2. Comparisons between the existing model and the new mod-

el with 10 GHz measurements for a wind speed of 9.3 m/s

at 52°.

(a)

(b)


el with 10 GHz measurements at 32° for VV-polarization (a)

and HH-polarization (b).

Table 1 shows the RMSEs between the measurement data

and the new model and those between the measurement and

the existing model. For example, the RMSEs for the data at 10

Table 1. RMSEs under various conditions

Freq.

(GHz)

Incidence

angle (°)Polarization

Wind

speed

(m/s)

RMSE (dB)

Existing

model [11]

New

model

10 32 VV 3.2 0.65 0.49

9.3 0.40 0.44

14.5 0.57 0.49

HH 3.2 0.61 0.79

9.3 0.80 0.78

14.5 0.73 0.81

52 VV 3.2 0.42 0.48

9.3 0.84 0.85

14.5 0.75 0.63

HH 3.2 0.52 0.36

9.3 0.49 0.33

14.5 0.33 0.82

13.9 40 VV 4.2 0.26 0.44

5.5 0.64 0.59

7.5 0.68 0.74

12 0.80 0.69

15 0.53 0.68

19.4 0.46 0.34

HH 4.2 1.16 1.30

5.5 0.61 0.59

7.5 0.47 0.55

12 0.61 0.75

15 0.72 0.52

19.4 0.26 0.39

Average 0.60 0.61

GHz and 3.2 m/s for the VV polarization (crosses in Fig. 2) are

0.65 dB and 0.49 dB for the existing model and the new model,

respectively, and for the data at 14.5 m/s (circles in Fig. 2), the

RMSEs are 0.57 dB and 0.49 dB for the existing and new

models, respectively, as shown in Table 1. Also Table 1 shows

the comparable RMSE values for other datasets between the

existing and the new model.

Fig. 4 shows the comparison of the new simple model with

independent datasets [14] and the existing model. The RMSEs

are the same at 0.71 dB for the existing and new models. The

new empirical model agrees well with the experimental data,

although the model is simpler than the existing model.

This new model has two main advantages over the existing

model with comparable accuracies. First, this model is easily

understandable and conveniently applicable because of its sim-

plicity. Second, the calculation time is dramatically reduced with

this new model. For example, when we compute the backscat-

tering coefficients of sea surfaces for various conditions with 10

wind speeds, 360 wind directions, 90 incidence angles, 10 fre-

quencies, and 2 polarizations (total of 6,480,000 data) using a

personal computer, it takes only 73 seconds with this model and

back

scat

terin

g co

effic

ient

, dB

JOURNAL OF ELECTROMAGNETIC ENGINEERING AND SCIENCE, VOL. 19, NO. 3,JUL. 2019

208


el with independent datasets acquired at 13.9 GHz for a

wind speed of 12.8 m/s at 40°.

7 hours with the existing model. In short, the calculation time is

reduced to 1/345.

V. CONCLUSION

A new empirical scattering model for estimating the back-

scattering coefficients of skewed sea surfaces was proposed. This

model is simpler and more practical than the existing model.

Instead of using the complicated skewness term 𝜎 (S), a mod-

ified correlation length was used in the new model to include

the skewness and peakedness effects. The new simple empirical

model agreed well with experimental datasets, and its accuracy

was comparable with that of the existing model.

This research was supported by the Basic Science Research

Program of the National Research Foundation of Korea (No.

2016R1D1A1A09918412).

REFERENCES

[1] F. T. Ulaby and D. G. Long, Microwave Radar and Radio-

metric Remote Sensing. Norwood, MA, Artech House, 2014.

[2] L. Schroeder, P. Schaffner, J. Mitchell, and W. Jones,

"AAFE RADSCAT 13.9-GHz measurements and analysis:

wind-speed signature of the ocean," IEEE Journal of Oceanic

Engineering, vol. 10, no. 4, pp. 346-357, 1985.

[3] J. R. Carswell, S. C. Carson, R. E. McIntosh, F. K. Li, G.

Neumann, D. J. McLaughlin, J. C. Wikerson, P. G. Black,

and S. V. Nghiem, "Airborne scatterometers: investigating

ocean backscatter under low-and high-wind conditions,"

Proceedings of the IEEE, vol. 82, no. 12, pp. 1835-1860, 1994.

[4] H. Masuko, K. I. Okamoto, M. Shimada, and S. Niwa,

"Measurement of microwave backscattering signatures of the

ocean surface using X band and Ka band airborne scat-

terometers," Journal of Geophysical Research: Oceans, vol. 91,

no. C11, pp. 13065-13083, 1986.

[5] F. M. Monaldo, D. R. Thompson, W. G. Pichel, and P.

Clemente-Colon, "A systematic comparison of QuikSCAT

and SAR ocean surface wind speeds," IEEE Transactions on

Geoscience and Remote Sensing, vol. 42, no. 2, pp. 283-291,

2004.

[6] O. Isoguchi and M. Shimada, "An L-band ocean geophysi-

cal model function derived from PALSAR,"IEEE Transac-

tions on Geoscience and Remote Sensing, vol. 47, no. 7, pp.

1925-1936, 2009.

[7] F. Nirchio and S. Venafra, "XMOD2: an improved geo-

physical model function to retrieve sea surface wind fields

from Cosmo-Sky Med X-band data," European Journal of

Remote Sensing, vol. 46, no. 1, pp. 583-595, 2013.

[8] F. J. Wentz, S. Peteherych, and L. A. Thomas, "A model

function for ocean radar cross sections at 14.6 GHz,"

Journal of Geophysical Research: Oceans, vol. 89, no. C3, pp.

3689-3704, 1984.

[9] H. Hersbach, A. Stoffelen, and S. de Haan, "An improved

C‐band scatterometer ocean geophysical model function:

CMOD5," Journal of Geophysical Research: Oceans, vol. 112,

article no. C03006, 2007.

[10] K. S. Chen, A. K. Fung, and F. Amar, "An empirical

bispectrum model for sea surface scattering," IEEE Trans-

actions on Geoscience and Remote Sensing, vol. 31, no. 4, pp.

830-835, 1993.

[11] A. K. Fung, Backscattering from Multiscale Roughness Surfac-

es with Application to Wind Scatterometry. London: Artech

House, 2015.

[12] T. Jin and Y. Oh, "An improved semi-empirical model for

radar backscattering from rough sea surfaces at X-

band," Journal of Electromagnetic Engineering and Sci-

ence, vol. 18, no. 2, pp. 136-140, 2018.

[13] A. K. Fung, Z. Li, and K. S. Chen, "Backscattering from a

randomly rough dielectric surface," IEEE Transactions on

Geoscience and Remote Sensing, vol. 30, no. 2, pp. 356-369,

1992.

[14] W. Jones, L. Schroeder, and J. Mitchell, "Aircraft meas-

urements of the microwave scattering signature of the

ocean," IEEE Journal of Oceanic Engineering, vol. 2, no. 1,

pp. 52-61, 1977.

back

scat

terin

g co

effic

ient

, dB


209

Taekyeong Jin received the B.S. and M.S. degrees in electronic and

electrical engineering from Hongik University, Seoul,

Korea, in 2017 and 2019, respectively. He worked at

the Hongik University Research Institute of Science

and Technology, Korea, from March to June 2019.

He joined Hanwha Systems in July 2019 as a Re-

searcher. His research interests include microwave

remote sensing, array antennas, and microwave scat-

tering from sea surfaces.

Yisok Oh received the B.E. degree in electrical engineer-

ing from Yonsei University, Seoul, Korea, in 1982,

the M.S. degree in electrical engineering from the

University of Missouri, Rolla, MO, USA, in 1988,

and the Ph.D. degree in electrical engineering from

the University of Michigan, Ann Arbor, MI, USA,

in 1993. In 1994, he joined the faculty of Hongik

University, Seoul, where he is currently a Professor

in the School of Electronic and Electrical Engineering. His current re-

search interests include polarimetric radar backscattering from various

Earth surfaces and microwave remote sensing of soil moisture and surface

roughness.




210

I. INTRODUCTION

An aircraft signal intelligence (SIGINT) system is used stra-

tegically and tactically to gather signals from radars and com-

munications systems. SIGINT systems are commonly divided

into communications intelligence and electronic intelligence

systems. As SIGINT systems require the ability to locate the

source of enemy signals [1], direction finding (DF) to support

this capability is an important requirement, and SIGINT sys-

tems generally operate in a wide frequency range of 20 MHz–40

GHz [2].

For electronic warfare (EW) requirements covering the Ka-

band spectrum, a channelized receiver is highly desirable. The

Future Combat Systems (FCS) radar and communications allo-

cations exist in this band of the spectrum [3].

Broadband receivers with wide millimeter-wave bands are es-

sentially required for electronic support systems. Millimeter-

wave systems with at least 30 GHz millimeter-wave bandwidths

with broad dynamic ranges, improved sensitivity, and front-end

complexity are commonly required in the fields of communica-

tion, radar, and wireless communication. To satisfy such high

requirements, general millimeter-wave systems include local

oscillator (LO) and downconverter functions in their front-end

receiver connected to an antenna. Moreover, the mixer, amplifi-

er, and the many necessary components generally require multi-

channels in which the frequencies have been converted into sub-

octave bands to satisfy the system requirements because of their

bandwidth and dynamic range limits. Channelized receivers

divided into multiple channels can be optimized to satisfy the

Design of a Ka Band Multi-Channel (25CH) Millimeter-

Wave Downconverter for a Signal Intelligence System Yuseok Jeon* · Jaejin Koo · Hyunkyu Kim

Abstract

In this study, we propose an approach for the design and satisfy the requirements of the fabrication of a reliable and stable high-frequency

downconverter for the millimeter-wave (Ka band) and detail the contents of the approach. We design and fabricate a stable downcon-

verter with a low noise figure, flat gain characteristics, and multi-channel characteristics suitable for millimeter-wave bands. The method

uses the chip-and-wire process for the assembly and operation of a bare MMIC device into the RF path. To compensate for the mis-

match among the many components used in the module, W/G transition, an image rejection mixer, a switch, and an amplifier suitable for

millimeter-wave frequency characteristics are designed and applied to the downconverter. To reject the spurious signals generated from

the complex local oscillation signals, the downconverter is designed to not affect the RF path. In the Ka-band downconverter, the gain is

measured from 41.89 dB to 42.83 dB at 33–35 GHz with flatness of about 0.94 dB. The measured value of the noise figure at CH1 is

4.936 dB with a maximum value in the 0.75–1.25 GHz intermediate frequency. The third intermodulation measurement result is 61.83

dBc under a -50 dBm input power and above gain, and the switching to select a channel takes about 622 μs.

Key Words: Downconverter, Front-End, Ka Band, Low Noise Figure, Millimeter-Wave, SIGINT System.

Manuscript received March 17, 2019 ; Revised April 25, 2019 ; Accepted June 12, 2019. (ID No. 20190317-015J)

Research and Development Department, Broadern Inc., Hwaseong, Korea. *Corresponding Author: Yuseok Jeon (e-mail: [email protected])




JEON et al.: DESIGN OF A Ka BAND MULTI-CHANNEL (25CH) MILLIMETER-WAVE DOWNCONVERTER FOR A SIGNAL INTELLIGENCE SYSTEM

211

requirements of ESM receivers. The problems with this archi-

tecture include the fact that the multiplied local oscillation sig-

nals are fed back into the antenna path because of the low isola-

tion between the RF and LO ports of the mixer and the re-

duced dynamic range due to the cascade structure that down–

converts the frequencies [4–7].

Although numerous studies have been conducted to design

millimeter-wave ultra-wideband receivers suitable for electronic

support systems, many constraints remain on the design and

manufacturing of such broadband downconverters. To make the

receivers have low noise figures, flat gain characteristics in the

band, wide dynamic ranges, and band-specific frequencies that

can be selected with low-loss transmission lines, a local oscillator

circuit necessary for frequency conversion should be incorpo-

rated in the module, and a closed structure is needed to avoid

signal interference from other devices (IFF transponder, Dop-

pler radar, communication system, etc.) in the system environ-

ment where the receivers are installed [8, 9].

In this paper, a design approach is proposed not only to im-

prove the signal acquisition accuracy by maximizing the number

of receiving input ports (25 input channels) but also to convert

the signal frequencies of 33–35 GHz in the millimeter-wave

Ka-band into the intermediate frequency (IF) 1 GHz (instanta-

neous frequency: 500 MHz) before the signals are delivered to

the signal processing unit. The content of the design approach

is as follows.

First, as the 33–35 GHz band received by an antenna th-

rough the free space medium has a waveguide structure, a tran-

sition structure should be designed to convert the channels into

micro-strip transmission lines. To implement 25 channels in the

form of grids in a limited space, the channels are designed using

the Rogers RT/duroid 5880 5-mil PCB material for small in-

sertion losses and better impedance matching.

Second, the signals are downconverted into IF signals with a

center frequency of 1 GHz and a frequency band of 500 MHz

through a low-noise amplifier (LNA) and a frequency converter

(packaging type) including an image suppression function. At

this time, the local oscillator signals necessary for frequency con-

version are phase locked with the internal voltage-controlled

oscillator (VCO) with an 8 GHz band using the external refer-

ence frequency of 1 GHz supplied by the system. As the voltage

turning (VT) voltage of the VCO is high, the structure of the

loop filter connected to the phase-locked loop (PLL) output

terminal was designed as an active type instead of a passive type.

Third, as the RF lines must be designed considering the am-

plitude matching among multiple channels (25 channels), the

RF lines of the individual channels are designed not only to be

electrically the same but also to have the same component

mounting positions and wire bonding length.

Fourth, to secure the degree of isolation between the input

and output channels, the mechanism is designed to prevent sig-

nals from flowing into other channels using valley form equip-

ment.

II. MILLIMETER-WAVE DOWNCONVERTER STRUCTURE

In this paper, a Ka-band multi-channel (25) millimeter-wave

downconverter module mounted on the front-end terminal of

an antenna is designed. This module has a complicated struc-

ture with 25 RF input (33–35 GHz) channels and 16 IF output

(0.75–1.25 GHz) channels. The module is divided into module

#1 and module #2 when it is mounted because of space re-

strictions for the entire system and to secure excellent noise fig-

ures. Module #1 is mounted near the antenna. A conceptual

diagram of the entire system is shown in Fig. 1.

Fig. 2 is a detailed block diagram of the multi-channel down-

Fig. 1. Configuration of the entire millimeter-wave downconverter.


212

converter necessary for the SIGINT system. The multi-channel

downconverter amplifies signals with a center frequency of 34

GHz (bandwidth 2 GHz) received from the antenna using an

LNA and converts the amplified signals into IF signals with a

center frequency of 1 GHz (RF bandwidth 2 GHz) through a

frequency converter (MIXER), which includes an image sup-

pression function. At this time, the local oscillator signals,

which are necessary for frequency conversion having a frequency

sweeping function with a 1 GHz bandwidth, are phase locked

with the internal VCO with an 8 GHz band using the external

reference frequency of 1 GHz supplied by the system, converted

to have a 16 GHz band through a frequency doubler, and input

into the LO port of the MIXER.

The frequency converter (MIXER) is a component that con-

tains a doubler inside, and the 16 GHz band is converted into a

32 GHz bandwidth inside the MIXER. In this case, the fre-

quency converter has a frequency variable function, so that the

local oscillator signals have a bandwidth of about 1.5 GHz (after

multiplying).

III. DESIGN AND SIMULATION

1. W/G to M/S Line Transition Simulation

In most millimeter systems, the waveguide and the micro-

strip are the two commonly used transmission lines. The wave-

guide is usually employed to connect the antenna and the mil-

limeter receiver or transmitter because of its low insertion loss

[10]. To transmit signals in the super high 34 GHz frequency

band from the waveguide (W/G) form to the micro-strip line

structure, the electric field mode must be converted through a

transition design.

As shown in Fig. 3(a), a transmission line with a single-layer

structure was placed at a point of the height of the Lambda/4

(about 1.7 mm) of the center frequency from the bottom surface

of the instrument of the W/G structure. For fine impedance

matching, signals were simulated in the form of patches at the

end of the transmission line, where the electric field is propagat-

ed. As shown in Fig. 3(b), the design results indicated the inser-

tion loss was not larger than 0.1 dB and that the reflection coef-

ficient was not larger than -23 dB in the 32–36 GHz band.

To relieve the problems by comparing the design value and

the actual measured value before installing on the module, a jig

was fabricated to make back-to-back measurement possible

using an Rogers RT/duroid 5880 PCB with permittivity of 2.2,

as shown in Fig. 4(a). As a result of the measurement in Fig.

4(b), insertion losses not exceeding -1 dB and reflection coeffi-

cients not exceeding -13 dB were obtained. The reason why the

measurement graph was wider by 1 GHz upward and down-

ward from the band (33–35 GHz) was that the frequency drift

was considered because of the errors that could occur in the ac-

tual fabrication.

2. PLL Design for the LO Section

To design the PLL local oscillation that is phase-synch-

Fig. 2. Detailed block diagram of the millimeter-wave downconverter.


213

(a)

(b)

Fig. 3. (a) Transition structure to convert the waveguide form into

the micro-strip line. (b) Result of the transition simulation

to convert the waveguide form into the micro-strip line.

(a)

(b)

Fig. 4. (a) Photo of the evaluation jig to verify the transition structure.

(b) Measurement results obtained with the evaluation jig.

Fig. 5. Block diagram of the PLL for local oscillation.

ronized with the external reference frequency, the external REF

applied to the REF port of the PLLIC was injected after con-

verting it into 200 MHz using a five-divider, and the LO fre-

quency in the 16 GHz band was created in the VCO output

using a doubler.

As shown in Fig. 5, to synchronize the phase with the phase-

locked external reference frequency, the output frequency of the

VCO was N-divided in the PLLIC, the phase was compared

with the reference frequency in the phase detector (PD), the

error voltage corresponding to the phase difference between the

two signals was outputted, and the VT terminal of the VCO

was controlled through the secondary passive loop filter, which

has the properties of a low-pass filter [11]. In this case, as the

maximum frequency input into the RF port of the PLLIC is

generally not higher than several GHz, although it may differ by

PLLIC, the VCO output frequency should be made to be in-

putted after two or four divisions using a prescaler. Likewise, as

the maximum frequency allowed to the RF port of the PLLIC

is not higher than several hundred MHz even when the external

reference frequency provided is high, the external reference fre-

quency should be made to be inputted after five divisions using

a prescaler [12].

The equation for the VCO frequency phase synchronized

with the external reference frequency is expressed in Eq. (1).

𝑓 𝑃 𝐵 𝐴 , (1)

where

𝑓 : Output frequency of the VCO,

𝑃: Dual modulus prescaler (8/9 or 16/17),

𝐵: Division ratio of the 13-bit counter,

𝐴: Division ratio of the 6-bit counter,

𝑓 : External reference frequency (here, frequency after

the prescaler),

R: Division ratio of the external reference frequency.

The final analyzed phase noise values in the 16 GHz band

were expected to be -78.5 dBc/Hz at 1 kHz offset and -81.5

dBc/Hz at 10 kHz offset, and the value of the reference phase

noise was a five-divided value from a 1 GHz external reference.

The overall phase noise of the analog PLL (Table 1) is as fol-

lows:

N

REXT.INPUT

1 GHz

4.03125 ~ 4.21875GHz

PLLICHMC704LP4EFractional N TypeFrequency : 8000 MHzNoise Floor : -230dBc/HzDC : 5V / 60mA

VCOHMC509LP5EFrequency : 7.8~8.8 GHzP/N : -115dBc/Hz@100KHzDC : 5V / 250mA

/ 28.0625~8.4375

GHz

X2

16.125~16.875GHz

/5


214

𝑇𝑜𝑡𝑎𝑙 𝑃𝑁 10 ∗ 𝐿𝑂𝐺 10 10 10

10 . (2)

The phase noise of the final output was calculated by Eq. (2)

[13].

The local oscillation signals in the 16 GHz band generated as

such were inputted into the frequency mixer for frequency con-

version. As the mixer included a doubler, the actual local oscilla-

tion signals were in the 32 GHz band.

3. Analysis of Noise Source in Low Dropout Regulators

The difference between insignificant noises and significant

noises is the degree to which the noises affect the operation of

the circuit in question. For example, a switching power supply

has a significant amount of output voltage ripple at 3 MHz. If

the circuit power by it has a bandwidth of only a few hertz, such

as a temperature sensor, this ripple may be of no consequence.

Conversely, if the same switching power supply powers an

RF PLL, the result can be quite different [14].

A low dropout (LDO) that shows excellent power supply re-

jection ratios (PSRR) in PLL circuits using VCO was applied.

Fig. 6 shows the PSRR characteristics of LDOs with improved

PSRR characteristics. The phase noise characteristic value ex-

pected at 10 kHz offset points was -81.5 dBc/Hz. Selecting and

applying LDOs with lower characteristic values than the previ-

ous is important.

4. Design for the Synchronized Main Path versus Monitoring

Port

A switch (SP25T) was designed so that when the signals re-

ceived through 25 channels were converted into IF frequencies

and transmitted to the signal processing area of system #1, a

power divider was applied to each main path output to sequen-

tially select the 25 paths. A switch was provided to a separate

system #2 to design a monitoring port that could determine the

signal intensity and the channels where the signals come in Figs.

7 and 8 illustrate the block diagram and simulation results.

In addition, a circuit was designed to supply the synchronized

control signals that fit the monitoring port. As timing is im-

portant when selecting channels 1–25, the switching time

should be considered when selecting the RF switch.

Fig. 6. Characteristic graphs by the frequency offset of LDOs with

excellent PSRR characteristics.

Table 1. Analyzed values of the phase noises in the 16 GHz band for local oscillation

BW (kHz) PFD (MHz) FO (GHz) N

Analog PLL (VCO) 10 10 8.05 805

Frequency (Hz) @ OFFSET 10 100 1,000 1.0E + 04 1.0E + 05 1.0E + 06

VCO phase noise −30.0 −30.0 −65.0 −88.0 −103.0 −125.0

Reference phase noise −127 −137 −155 −165 −165 −165

Prescaler phase noise 0.0 0.0 0.0 0.0 0.0 0.0

PD phase noise −97.9 −97.9 −97.9 −97.9 −97.9 −97.9

High pass attenuation (HF) 100.0 60.0 20.0 0.0 0.0 0.0

Low pass attenuation (LF) 0.0 0.0 0.0 0.0 20.0 40.0

Reference multiply −68.9 −78.9 −96.9 −106.9 −106.9 −106.9

No. 1 VCO output −130.0 −90.0 −85.0 −88.0 −103.0 −125.0

No. 2 Reference output −68.9 −78.9 −96.9 −106.9 −126.9 −146.9

No. 3 Prescaler output 0.0 0.0 0.0 0.0 −20.0 −40.0

No. 4 Phase detector output −97.9 −97.9 −97.9 −97.9 −117.9 −137.9

Total PLVCO's phase nose −68.9 −78.5 −84.5 −87.5 −102.8 −124.8

16.1 GHz PLVCO PN (X2) −62.9 −72.5 −78.5 −81.5 −96.8 −118.8


215

Fig. 8. Simulation results of the synchronizing signals and output

signals when channel 1 is selected.

IV. FABRICATION AND MEASUREMENT

A millimeter-wave Ka downconverter, which has 25 RF in-

put ports (33–35 GHz), 16 output ports (0.75–1.25 GHz), and

one external reference frequency (1 GHz) input port, consisted

of module #1 (millimeter-wave unit) and module #2 (control

and switching unit). Module #2 was designed and fabricated in

the form of four submodules. Fig. 9(a) illustrates the entire set-

up in which the two modules are connected to each together.

Fig. 9(b) is a photograph of the inside of module #1, which is

an area where the RF frequency is mixed with the LO frequen-

cy in the mixer (Hittite's HMC1065LP4E), which includes an

image rejection function to output IF signals in the 1 GHz

band. The PCB was designed to be installed using the epoxy on

the entire bottom surface of the equipment, so that the frequen-

cy in the Ka-band could be transmitted through the transmis-

sion line without any loss or distortion.

(a)

(b)

(c)

Fig. 9. (a) Photograph of the complete setup (from 33–35 GHz to

1 GHz). (b) Photograph of module #1 (RF section). (c)

Photograph of module #1 (DC section).

Fig. 7. Block diagram of the systems where one output trigger is designed for 25 channels.

IF AMP (1)IF AMP (2)LPFHPFP/D

SP25T

LIMITERLPF Thermal Pad Atten.

IF AMP (1)IF AMP (2)LPFHPFP/D LIMITERLPF Thermal Pad Atten.

IF #1

IF #25

IF Section

SYSTEM #1

SYSTEM #1

SYSTEM #2

OUTPUTTRIGGER

FROM RF #1

FROM RF #25


216

RT/duroid 5880 5-mil PCB was used up to the mixer stage

where the Ka band exists, and Rogers RO4003 20-mil PCB

(a)

(b)

(c)

(d)

Fig. 10. (a) Photograph of module #2. (b) The PLL board includ-

ing the PLLIC and VCO. (c) The IF board including the

AMP and filter. (d) The switch module including the dis-

tribution board.

(a)

(b)

(c)

(d)

Fig. 11. Results of the measurement of gain using a Scalar Net

Work Analyzer: (a) 33.0–33.5 GHz band, (b) 33.5–34.0

GHz band, (c) 34.0–34.5 GHz band, and (d) 34.5–35.0

GHz band.

was used at low frequencies in the L band converted into IF to

make SMT possible.

Fig. 10(a) shows the inside of module #2, Fig. 10(b) the PLL


217

board that generates the LO signal needed to convert the RF

frequency into the IF frequency, Fig. 10(c) the IF board for am-

plifying and filtering the IF signal, and Fig. 10(d) the distribu-

tion board in which 25 paths are distributed and switched to 16

paths.

1. GAIN Measurement Results

Fig. 11(a)–(d) show the total gain of channel 1 from the RF

input port to the IF output port. As the bandwidth of the IF

frequency was about 500 MHz, the total gain was measured

four times after changing the RF and LO frequencies to 500

MHz.

The total gains were as follows:

Up to 42.83 dB in the 33.0–33.5 GHz band




2. NF Measurement Results

As for the results of the measurement of noise figures at the

IF frequency in the L band that was changed from the Ka band,

which is a receiving frequency band, the value measured at the

0.75–1.25 GHz band was 4.936 dB with a maximum value.

This result is mainly due to the noise figure characteristics from

the W/G transition structure to the mixer in which the LNA is

included. Therefore, the characteristics vary greatly depending

on how much the insertion loss is minimized, as shown in Fig.

12.

3. Third Intermodulation Measurement Results

In channel #1, which is representative, the third-order IMD

characteristics that show the linearity of the module measured at

the center frequency of 1 GHz of the IF frequency was meas-

ured as 61.83 dBc (Fig. 13).

4. Switching Speed Measurement Results

Switching speed is important at the system level because the

speed of the switch used to select the RF path determines how

Fig. 12. Results of the measurement of noise figure (IF band).

Fig. 13. Results of the measurement of IM3 (IF band).

fast the signal can be received. It took 622 μs to switch to an-

other channel and select the channel through which the signals

came in Fig. 14.

V. CONCLUSION

In this paper, we designed and fabricated a Ka-band down-

converter module with a 25-channel RF input port with a low

noise figure, flat gain characteristics, and reliability by applying

the chip-and-wire process for assembly into the RF path and

operation of a bare-type MMIC device. To compensate for the

mismatch among the many components used in the module,

W/G transition, an image rejection mixer, a switch, and an am-

plifier suitable for millimeter-wave frequency characteristics

were designed and applied to the downconverter.

The main RF line was a dielectric substrate (RT/duroid

5880), which had a relative dielectric constant of 2.2 and a die-

lectric thickness of 0.127 mm, and AI203 (ceramic, ATC Co.),

which had a relative dielectric constant of 9.8 and a dielectric

thickness of 0.254 mm. In the Ka-band downconverter module,

the gain was 41.89–42.83 dB at 33.0–35 GHz, with flatness of

about 0.94 dB. The measured value of the noise figure at CH1

was 4.936 dB, with a maximum value in the 0.75–1.25 GHz IF

frequency. The third intermodulation measurement result was

61.83 dBc. The switch to selecting a channel took 622 μs.

Fig. 14. Results of the measurement of the switching speed.


218

The millimeter-wave (Ka band, 33–35 GHz) 25CH down-

converter module proposed in this paper is considered applica-

ble to the modules that are installed at the rear end of the an-

tenna of the SIGINT system in the field of EW to collect sig-

nals precisely. However, further studies on how to satisfy and

increase the dynamic range still have to be conducted.

This work was funded by the technology development

business fund of the Ministry of SMEs and Startups in 2019.

REFERENCES

[1] D. Adamy, EW 102: A Second Course in Electronic Warfare.

Norwood, MA: Artech House, 2004.

[2] T. Macnamara, Introduction to Antenna Placement and Instal-

lation. Chichester, UK: John Wiley & Sons, 2010.

[3] J. T. Clark, A. K. Witcher, and E. D. Adler, "Ka band

channelized receiver," US Army Research Laboratory, Re-

port No. ARL-TR-7446, 2015.

[4] D. E. Allen, "Channelised receiver: a viable solution for EW

and ESM systems," IEE Proceedings F (Communications, Ra-

dar and Signal Processing), vol. 129, no. 3, pp. 172-179, 1982.

[5] R. T. Logan and R. D. Bynum, "Millimeter wave optical

link/frequency converter system," Tracor Aerospace Elec-

tronic Systems Inc., Lansdale, PA, 1998.

[6] S. T. Winnall and D. B. Hunter, "A fibre Bragg grating

based scanning receiver for electronic warfare applications,"

in Proceedings of 2001 International Topical Meeting on Mi-

crowave Photonics Technical Digest (Cat. No. 01EX476),

Long Beach, CA, 2002, pp. 211-214.

[7] I. B. Yun and C. M. Jeong, "Creation of search areas con-

Yuseok Jeon obtained his B.S. in electronic communication engi-

neering from Chungju University, Chungju, Korea,

in 2000, his M.S. in electronic engineering from

Chungbuk National University, Cheongju, Korea, in

2004, and his Ph.D. in electronic and electric engi-

neering at Dankook University, Yongin, Korea, in

2018. He worked as director in R&D division at

Broadern Inc., Hwaseong, Korea, from 1999 to

2019. His research interests include the development of transceivers, receiv-

ers, and synthesizers using chip-and-wire processes for EW and radar sys-

tem applications. He received the Korean Engineer Award in 2018.

sidering the receiver of the electronic warfare ES system,"

Journal of the Korea Institute of Military Science and Technology,

vol. 14, no. 6, pp. 1091–1096, 2011.

[8] M. S. Rao, "EMI/EMC effects on EW receiver systems of

military aircraft," in Proceedings of 2008 10th International

Conference on Electromagnetic Interference & Compatibility,

Bangalore, India, 2008, pp. 63-67.

[9] W. R. Brinlee, A. M. Pavio, C. L. Goldsmith, and W. J.

Thompson, "A monolithic multifunction EW broadband

receiver converter," in Proceedings of the 15th Annual GaAs IC

Symposium, San Jose, CA, 1993, pp. 207-210.

[10] S. Sun and Y. Huang, "A wideband waveguide-to-micro-

strip transition via a substrate-integrated waveguide trans-

former," in Proceedings of 2014 XXXI URSI General Assem-

bly and Scientific Symposium (URSI GASS), Beijing, China,

2014.

[11] X. Gai, G. Liu, S. Chartier, A. Trasser, and H. Schu-

macher, "A PLL with ultra low phase noise for millimeter

wave applications," in Proceedings of the 40th European Mi-

crowave Conference, Paris, France, 2010, pp. 69-72.

[12] Analog Device Inc., "PLL Frequency Synthesizer (ADF-

4106) data sheet," 2015; https://www.analog.com/media/

en/technical-documentation/data-sheets/ADF4106.pdf.

[13] Y. Jeon and S. Bang, "New configuration of a PLDRO

with an interconnected dual PLL structure for K-band ap-

plication," Journal of Electromagnetic Engineering and Sci-

ence, vol. 17, no. 3, pp. 138-146, 2017.

[14] G. Morita, "Noise sources in low dropout (LDO) regula-

tors (AN-1120 Application Note)," 2011; https://www.

analog.com/media/en/technical-documentation/application-

notes/AN-1120.pdf.

Jaejin Koo obtained his B.S. in electronic engineering from

Soonchunhyang University, Asan, Korea, in 2006

and his M.S. from the Department of Electrical and

Communication Engineering, Soonchunhyang Uni-

versity, Asan, Korea, in 2008. He worked as general

manager in R&D division at Broadern Inc.,

Hwaseong, Korea, from 2008 to 2019. His research

interests include the development of transceivers and

receivers using chip-and-wire processes for EW and radar system applica-

tions.


219

Hyunkyu Kim obtained his B.S. in electronic engineering from

Sunmoon University, Asan, Korea, in 2009 and his

M.S. in electronic and electric engineering, Dankook

University, Yongin, Korea, in 2012. He worked as

general manager in R&D division at Broadern Inc.,

Hwaseong, Korea, from 2009 to 2019. His research

interests include the development of transceivers,

receivers, RF amp, and GaN transistors using chip-

and-wire processes for EW and radar system applications.

JEES Journal of Electromagnetic Engineering and Science

Instructions for Authors

Journal of Electromagnetic Engineering and Science (J. Electromagn. Eng. Sci.; JEES) is an official English journal of the Korean Institute of Electromagnetic and Engineering Science (KIEES). The objective of JEES is to publish academic as well as industrial research results and foundings on electromagnetic engineering and science. The journal covers all aspects of researches and technology related to electromagnetics: Electromagnetic Compatibility/Electromagnetic Interference, Microwave and Millimeter-Wave Engineering, Antenna and Propagation, Electromagnetic Theory, Wireless Communication, Lightwave and Electro-Optics, Materials and Components, Software Defined Radar, Radar, Bioelectromagnetics, and etc. I. Copyright and Creative Commons Attribution Licensing The copyright and the transfer rights of the digital content of published papers and the journal is owned by the KIEES. All published materials are also assigned a Creative Commons Attribution License (http://creativecommons.org/licebses/by-nc/4.0/). The journal accepts manuscripts for consideration with the understanding that the manuscript has not been published previously and is not currently under consideration for publication elsewhere, and in addition, that the authors (or their employer, if it holds the copyright) are authorizing the transfer of the copyright to the Institute when the manuscript is submitted. Author should check the copyright transfer conditions and forms at http://www.jees.kr. II. Research and Publication Ethics Research published in JEES must have followed institutional, national, and international guidelines. For policies on research and publication ethics that are not stated in these instructions, please refer to the Committee on Publication Ethics (COPE) Guidelines (http://publicationethics.org/resources/code-conduct).1. Originality and Duplicate Publication A manuscript submitted for publication in JEES should be an original work with technical values. It must not have been previously published and is not under consideration for publication elsewhere. JEES assumes that the materials submitted for its publications are properly available for general dissemination for the readership of those publications. It is the responsibility of the authors, not JEES, to determine whether disclosure of their materials requires prior consent of other parties and, if so, to obtain it. If an author uses charts, photographs, or other graphics from any previously printed materials, he/she is responsible for obtaining written permissions from the publisher to use them in his/her manuscript. Responsibility for the contents of published papers rests upon the authors, not JEES. 2. Conflict of Interest Authors are required to complete a declaration of conflict of interests. All conflict of interests that are declared will be listed at the end of published articles.3. Authorship Authorship should be restricted to those individuals who have met each of the following three criteria: 1) made a significant contribution to the conception and design of the project, or the analysis and interpretation of the data, or other substantial scholarly effort; 2) participated in drafting, reviewing and/or revising the work; and 3) approved the final version for publication. After the initial submission of a manuscript, any changes whatsoever in authorship (adding author(s), deleting author(s), or re-arranging the order of authors) must be explained by a letter to the editor from the authors concerned. The content of this

letter must be acknowledged and agreed upon by all of the authors of the paper. If you accept the policies on research and publication issued by KIEES, authors must click the response of “The Code of Research Ethics” through online. III. Submission of Manuscripts Authors are expected to be members of the KIEES except for some special cases approved by the Editorial Board of KIEES. All manuscripts should be submitted electronically through the online submission and review site (https://mc03.manuscriptcentral.com/jees). For the first submission, you may be required to create an account on the submission site. A manuscript can be submitted at any time of the year. When submitting a manuscript, authors need to make sure that their manuscripts do not provide any of their identities such as authors’ names and affiliations as the review is double-blinded. More detailed submission instruction is available in the upper right corner of the submission site. All manuscripts submitted to the Journal must comply with the instruction and the standard format of the Journal. Otherwise, it will result in return of the manuscript and possible delay in publication. For assistance, please contact us via e-mail ([email protected]). IV. Peer Review Process The manuscript will be forwarded to three reviewers selected for their expertise in the field of the submitted manuscript. The acceptance criteria for all papers are based on the quality and originality of the research and its clinical and scientific significance. During the review process, the author is often asked to expand, rewrite, or clearly explain the specific contents of his/her paper. It is not uncommon that an author is asked to provide another draft with the suggested changes for further review. A revised manuscript should be submitted to the homepage within a month from the date on which any change of the manuscript is requested to the author. Once a manuscript has received the final approval of the reviewers and Editor-in-Chief, the author will be notified and asked to prepare the manuscript for final publication and to possibly complete an additional information form. V. Publication Type The papers are classified into five categories.Regular Paper should be an original work that contributes to the academic interests of the KIEES members with technical values. The paper should also be written within 12 pages with A4 size including figures, charts, and tables (The main body of text consists of two columns). Letter consists of reports of preliminary results or short reports on completed work that are of current interest to many researchers in the field. Comments on material previously published in the journal, suggestions for new directions, and errata could be included. Their length must be less than 3 pages (with a two-column format) including paper title, author affiliation, reference, etc. Review Paper will be published by direct submission as well as from invited experts. In both cases, the work will be subject to editorial review. Review papers should critically review topics not only to inform the reader of the background, but also to communicate the state of the art and outstanding research problems. Technical Report is on innovative technical achievements of interest to the community and usually a report of an extensive series of measurements. Report is often involving display in the form of tables or graphs, with text describing the conditions and procedures of measurement.Editorial is a brief report of research findings adequate for the journal’s scope and of particular interest to the community. VI. Manuscript Preparation All manuscripts must be written in MS-Word and adhere to the following guidelines:

1. A cover of each paper manuscript should include a title, authors’ names (main author and co-authors), author’s organizations, contact information (e-mail and phone number), and the author’s area of expertise. 2. The first page of a main text should only contain title, abstract with a length of about 150 words, and key words with around five words.3. The contents of the manuscript should be arranged in the order of abstract, main text, acknowledgments, references, and appendix.4. The numbers corresponding to chapters in the manuscript should be written in Roman numerals (I, II, III, IV...) and the numbers corresponding to sections should be written in Arabic numerals (1, 2, 3, 4...).5. Equation numbers should be given in Arabic numerals enclosed in parentheses on the right-hand margin. They should be cited in the text as, for example Eq. (1) or Eqs. (1)-(3).6. All tables should be numbered consecutively with Arabic numerals. They should be referred to in the text and should be numbered according to their order of mention in the text. In addition, all tables should, not only list all abbreviations in the table in footnotes at the end, but also have a title that is concise and describes the table’s contents. Vertical lines are not used. The table should be self-explanatory and supplement, not duplicate, the text. If the table or any data therein have been published, a footnote to the table must give permission information to the original source. The structure should be clear, with simple column headings giving all units. A table should not exceed one page when printed. Use lowercase letters in superscripts a,b,c... for special remarks.7. All figures should be of high quality meeting with the publishing requirement with legible symbols and legends. In preparing the figures, authors should consider a size reduction during the printing process to have acceptable line clarity and character sizes. Use only figures that are necessary to illustrate the meaning of the text. Figures must be black and white of high contrast. All figures should be referred to in the text as, for example, Fig. 1, Fig. 2(a), or Figs. 1-3.8. Only those references cited in the text should be listed in the references. Authors are responsible for the accuracy and completeness of their references and the correct text citations. In the text the reference should be numbered in bracket in ascending order (e.g., [1, 3], or [4-6]; Lee [2] and Kim and Park [5]; Jang et al. [7]). In case of the paper title, only the first letter is to be capitalized. However, in case of journal and book titles, the first letter of each word should be capitalized and all of the letters should be italicized. See the example below. Books[1] F. Giannini and G. Leuzzi, Nonlinear Microwave Circuit Design. NewYork, NY: John Wiley & Sons Inc.,

2004.Journals[2] H. Ahn and B. Kim, "Equivalent transmission-line sections for very high impedances and their application to

branch-line hybrids with very weak coupling power," Journal of Electromagnetic Engineering and Science, vol. 9, no. 2, pp. 85-97, Jun. 2009.

Report[3] E. E. Reber, R. L. Michell, and C. J. Carter, "Oxygen absorption in the earth’s atmosphere," Aerospace Corp.,

Los Angeles, CA, Tech. Rep. TR-0200 (4230-46)-3, Nov. 1988.Conference Proceedings[4] S. P. Bingulac, "On the compatibility of adaptive controllers," in Proceedings of the 4th Annual Allerton

Conference on Circuit and System Theory, NewYork, pp. 8-16, 1994.Papers Presented at Conferences[5] J. G. Kreifeldt, "An analysis of surface-detected EMG as an amplitude-modulated noise," presented at the 8th International Conference on Medical and Biological Engineering, Chicago, IL, 1969.[6] J. Arrillaga and B. Giessner, "Limitation of short-circuit levels by means of HVDC links," presented at the IEEE Summer Power Meeting, Los Angeles, CA, Jul. 1990.

Theses (M.S.) and Dissertations (Ph.D.)[7] N. Kawasaki, "Parametric study of thermal and chemical nonequilibrium nozzle flow," M.S. thesis, Department

of Electronic Engineering, Osaka University, Osaka, Japan, 1993.[8] J. O. Williams, "Narrow-band analyzer," Ph.D. dissertation, Department of Electronic Engineering, Harvard

University, Cambridge, MA, 1993.Standards[9] IEEE Criteria for Class IE Electric Systems, IEEE Standard 308, 1969.Online Sources[10] R. Bartle, "Early MUD History," Nov. 1990; www.ludd.luth.se/aber/mud-history.html.

9. When citing any paper published in JEES, it should be indicated the name of the journal as Journal of Electromagnetic Engineering and Science or J. Electromagn. Eng. Sci.10. Acknowledgment, if needed, appears before the reference. Sponsorship or financial support acknowledgment should be included here. 11. Unit and Abbreviation:If the authors describe length, height, weight, and volume, they should use standard metric units. Temperature should be given in degrees Celsius. All other units should follow the International System of Units (SI). All units must be preceded by one space except percentages (%) and temperatures (°C). Abbreviations must be used as an aid to the reader, rather than as a convenience of the author, and therefore their use should be limited. Generally, abbreviations that are used less than 3 times in the text, including tables and figure legends, should be avoided. Standard SI abbreviations are recommended. Other common abbreviations are as follows (the same abbreviations are used for plural forms): h (hour), min (minute), s (second), d (day), wk (week), mo (month), y (year), L (liter), mL (milliliter), μL (microliter), g (gram), kg (kilogram), mg (milligram), μg (microgram), ng (nanogram), pg (picogram), g (gravity; not g), nm (nanometer), μm (micrometer), mV (millivolt), mA (milliampere), mW (milliwatt), C (coulomb), μF (microfarad), mH (millihenry), n (samplesize), SD (standard deviation of the mean), and SE (standard error of the mean).

VII. Accepted Manuscript Once the review process has been completed with a decision of acceptance, the final manuscript accommodating all of the reviewers’ comments should be submitted along with photos of the authors and their brief biographies (including major research areas). The accepted papers will be published, in principle, in the order of initially submitted dates subject to decision of the Editorial Board.1. Page Proofs Authors will be given an opportunity to review the laser printed version of their manuscripts before printing. One set of page proofs in PDF format will be sent by e-mail to the corresponding author. The review should be solely dedicated to detecting typographical errors.2. Publishing Charge The publishing charge for general publishing is KRW 150,000 (US$150) for up to the first 6 pages. For 7 to 8 pages, an extra charge of KRW 30,000 (US$30) per page, and for 9 pages or more, an extra charge of KRW 40,000 (US$40) per page. Furthermore, the KIEES charges extra KRW 100,000 (US$100) for the paper acknowledging a financial support from an institution, in addition to the above mentioned page charge. Twenty reprints without a cover will be supplied without an additional charge.

Contact UsEditorial office of the Korean Institute of Electromagnetic Engineering and Science217, Saechang-ro, Yongsan-gu, Seoul, 04376, KoreaTel: +82-2-337-9666/332-9665 Fax: +82-2-6390-7550http://www.jees.kr, E-mail: [email protected]

AUTHOR CHECKLIST

Title :

General

□ This paper has not been and will not be published in any other journal.

□ This paper follows the format of the KIEES paper submission guideline, is less than 12 pages long, including figures

and tables (3 pages for a letter), and has a two-column layout.

□ This paper includes a cover page, an abstract, keywords, main text, appendix, and references in the correct order.

Each page has a consecutive page number.

Cover page

□ The cover page includes the title, authors, affiliation of all authors, identifier of the corresponding author, and provides

contact information for the corresponding author (phone number, e-mail address, and mobile number).

Abstract and key words

□ An abstract is provided and is less than 500 words long.

□ Fewer than 5 keywords are provided.

Main text

□ The reference number is written next to the quote.

□ The chapter numbers are written in Roman numerals (I, II, III, IV…) and subheading numbers are written in Arabic

numerals (1, 2, 3, 4…).

References

□ Sufficient Korean as well as international literature is referenced in the paper.

□ Only the references used in the main text are included in the reference list and these are numbered according to their

order of appearance.

□ All reference citations follow the correct format indicated in the submission regulations.

Figures and tables

□ For the figures and tables, the first letter of the very first word is capitalized.

□ All figure legends include both a title and a detailed explanation to help in understanding what the figure depicts.

□ All tables are self-explanatory and do not repeat the content from figures or sources included in the main text.

* Check each box after confirming that the statement is true.

The authors of this paper have confirmed the above items, following the paper submission regulations of the Korean

Institute of Electromagnetic Engineering and Science (KIEES), and are requesting publication of this paper.

/ / 2019

Representative author: ____________________ (signature)

jees-19-3.pdf - Journal of Electromagnetic Engineering and ...

Documents

Transcript of jees-19-3.pdf - Journal of Electromagnetic Engineering and ...