Engineering Science and Technology, an International Journal 20 (2017) 1260–1274

Contents lists available at ScienceDirect

Engineering Science and Technology,

an International Journal
journal homepage: www.elsevier.com/locate/jestch

Full Length Article

Parametric evaluation of microstrip log periodic dipole array antenna

using transmission line equivalent circuit
Shailendra Singh Pawar ⇑, Madhu Shandilya, Vijayshri Chaurasia
Department of Electronics and Communication Engineering, Maulana Azad National Institute of Technology, Bhopal P. O. Box 462001, India

a r t i c l e i n f o a b s t r a c t

Article history: This paper presents the design of a Microstrip Log Periodic Dipole Array (MLPDA) antenna for wireless
Received 27 August 2016 communication in C band. A simplest structure of the edge fed Microstrip Log Periodic Dipole Array is
Revised 2 August 2017 chosen as base design. The parameters of proposed antenna are identified by investigating the effect of
Accepted 12 September 2017
variation in geometrical structure on gain and bandwidth. A scale factor (s) and a spacing factor (r)
Available online 22 September 2017
are defined accordingly and imposed, in order to create the variations in these parameters. Complexity
of parametric variations involved has been observed through change in bandwidth and realized gain of
Keywords:
proposed MPLDA. A transmission line equivalent circuit of the proposed MLPDA is developed to give
Microstrip log periodic dipole array
Ultra wide band
physical insight of the structure and validation. Log periodic nature of the proposed antenna array is ver-
Wireless local area network ified by plotting input impedance with logarithm of frequency over the specified range. To validate the
Equivalent circuit results, prototype of proposed antenna is realized in the high frequency laboratory and results are veri-
fied with few experimentally measured values.
Ó 2017 Karabuk University. Publishing services by Elsevier B.V. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction by using tree dipole structure. An improved front to back ratio

The conventional Log Periodic Dipole Array antenna design was
proposed by R.H. Du Hamel and D.E Isbell in 1957 for broadband
applications [1–4]. The Printed Log Periodic Dipole Array (PLPDA)
antenna was presented by Campbell et al. in 1977 [5]. A microstrip
based structure of log periodic dipole array was introduced first
time using modified Carell’s method [3]. Later on the basic geom-
etry of a strip line Log periodic dipole antenna introduced by
Campbell et al. has been modified sequentially by several research-
ers to achieve better antenna performance in terms of large band-
width, reduced size, improved front to back ratio, low cross
polarization level and better gain [6–8].
A wide variety of techniques have been proposed for size reduc-
tion, bandwidth enhancement, gain improvement, feeding tech-
niques and cross polarization of PLPDA [9–22]. Gheethan et al. in
[9] and Anagnostou et al. in [10] has reduced the size of PLPDA
by using Meander shaped and Koch shaped dipoles respectively.
A reconfigurable PLPDA with switching elements which operates
in two modes to cover two bandwidths has been presented in
[11]. Self-similarity property of fractals is used by authors in [12]
⇑ Corresponding author.
E-mail addresses: pawar_shail@yahoo.co.in (S.S. Pawar), madhu_shandilya@
yahoo.in (M. Shandilya), vijayshree21@gmail.com (V. Chaurasia).
Peer review under responsibility of Karabuk University.
design of PLPDA for RFID application has been achieved in 2002
[13]. Casula et al. have proposed UWB PLPDA, in C, X and Ku bands
with improved feeding structure of mirror coaxial cable to obtain
stable phase center [14]. This structure was not easily realizable.
Hence same authors have suggested CPW (Co planer waveguide)
fed PLPDA with wide working band of 3–6 GHz with relatively
large axial dimension [15]. An UWB low cross polarization PLPDA
with a stable gain around 8 dB for 0.8–7 GHz range using FR4 sub-
strate is given in [16]. A multiband antenna in S-C band for
whether a radar application is proposed in [17]. Zhai et al. have
designed a super high gain PLPDA antenna Using SIW (Substrate
Integrated Waveguide) technology in Ka band (25–40 GHz) [18].
A non-cross feeding structure for PLPDA to avoid complexity in
feeding with long coaxial line was suggested by Kang et al. [19].
But this antenna design is less efficient along with lagging of
180° additional phase shift in conventional cross feed structure,
which eventually results in less bandwidth. A compact wideband
Printed LPDA antenna for WLAN applications in 5 GHz band have
been proposed in [20].
The research is comparatively limited in parametric analysis of
PLPDA to understand its basic phenomenon of working. Pantoja
et al. have given deep insight of feeding technique and effect of
dielectric substrate in the basic design of PLPDA [7]. It explains
the basic feeding mechanism by using a balun (balance to unbal-
ance transformer) analysis that relies on a quasi static model of

https://doi.org/10.1016/j.jestch.2017.09.002
2215-0986/Ó 2017 Karabuk University. Publishing services by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274 1261

the transmission lines involved [7]. Abri et al. have proposed the
transmission line model of a series fed log periodic antennas array
over a band of frequencies for satellite communications [21]. A 14
element monolayer multi octave log periodic microstrip antenna
have suggested by Lei et al. [22]. A series fed structure is analyzed
using an equivalent circuit of patch antenna elements, for gap feed
and direct feed of patches [22].
The performance of Microstrip Log Periodic Dipole Array
(MLPDA) mainly determined by length, width and spacing between
dipoles, interleave factor r, and geometry constant s. Antenna
bandwidth, gain and radiation pattern are governed by these
parameters. In this work, an effort is made to analyze the effect
of parameters on reflection coefficient (S11), realized gain and radi-
ation pattern of MLPDA. An equivalent circuit model has also been
presented to give more physical insight of the resonance behavior
of antenna. Present analysis shows complexity of parametric vari-
ations in the design of MLPDA and can be helpful in design of opti-
mized antenna in terms of size, bandwidth and gain.
In rest of the paper, Section 2 gives the design of proposed
MLPDA antenna followed by its detailed analysis in Section 3. In
Section 4, experimental outcomes of realized antenna are pre-
sented. Finally conclusions are made in Section 5.
2. Proposed microstrip log periodic dipole array (MLPDA)
antenna
This work presents a detailed analysis of eight element MLPDA
antenna using FR4 substrate (er = 4.4 and width h = 1.6 mm) [20].
The structure is very close to a standard (wire) LPDA and therefore
the standard strategy of designing LPDA can be used [3], along with
some modifications (Fig. 1). The property of antenna is mainly
determined by 3 parameters; the scale factor s, spacing factor r
and number of diploes N. All the adjacent dipoles elements are
printed on two sides of micro strip substrate, in an alternate way
and are fed with coaxial cable at lower end using 50 X SMA
connector.
All the elements of antenna are fed by a paired microstrip to
match the resistance of 50 X. Fig. 1 completely describe the geom-
etry of MLPDA, where L1, L2. . ..L8 are half lengths of dipole element,
W1, W2,. . ...W8 are widths of dipole element, S1, S2. . .. . .S8 are center
to center spacing between dipoles, Ws is width of strip line and K is
feed length. For proper analysis of the antenna layout, microstrip
dipole elements of upper layer are denoted by odd numbers 1, 3,
5, 7. . .0.15, while the lower layer dipole elements are denoted by
even numbers 2, 4, 6. . .16.One complete dipole is formed by con-
sidering upper and lower dipole elements as one unit, which is dis-
tributed asymmetrically over the strip length. For calculating the
width of parallel strip feed line to match the required impedance
of 50 X, a 25 X standard micro strip with substrate height of h/2
is designed using calculations given in [14]. The resulting width
of parallel strip obtained is 4.185 mm (=Ws). The effective dielectric
constant is calculated by using relation;
er þ 1 er  1 1
eeff ¼ þ  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð1Þ
2 2 1 þ 10h Ws
The calculated value of eeff = 3.6962, from Eq. (1). Now, starting
with required bandwidth of (fu  fl) GHz, where fu is upper cut off
frequency and fl is lower cutoff frequency and following the rules
given in [23], the length of largest dipole Lmax is determined as:
Lmax ¼ K 1 kmax ð2Þ
where kmax is the largest effective wavelength at the lowest operat-
ing frequency fl, and is given as;
c
kmax ¼ pffiffiffiffiffiffiffi ð3Þ
eeff  f l
K1 is upper truncation constant that depends on scaling factor s and
is calculated using Eq. (4) [23],

13
16
9 12
5
1 8
Feed 4

2
6
3 10
7
14
11
15

(a)

(b) (c)

Fig. 1. Geometry of 8 element printed log periodic dipole array antenna (a) Complete schematic layout, (b) upper layer (solid lines), (c) Lower layer (blanked lines).
1262 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274

K 1 ¼ 1:01  0:519  s ð4Þ Table 2

Basic Geometry of MLPDA.
Then length of shortest dipole element can be calculated by
Ln (mm) Wn (mm) Sn (mm)
using Eq. (5);
L8 = 8.46 W8 = 1.83 S7 = 5.08
Lmin ¼ K 2 kmin ð5Þ L7 = 7.62 W7 = 1.65 S6 = 4.57
L6 = 6.85 W6 = 1.49 S5 = 4.11
kmin is the shortest effective wavelength at highest operating fre- L5 = 6.17 W5 = 1.34 S4 = 3.70
quency fu and K2 is lower truncation constant calculated using for- L4 = 5.55 W4 = 1.20 S3 = 3.33
mula given in [23].Values of kmin and K2 has been calculated using L3 = 5.00 W3 = 1.08 S2 = 3.00
L2 = 4.5 W2 = 0.97 S1 = 2.70
Eqs. (6) and (7) respectively; L1 = 4.05 W1 = 0.88 –
c
kmin ¼ pffiffiffiffiffiffiffi ð6Þ
eeff  f u
and listed in Table 2. As the antenna performance is greatly
K 2 ¼ 7:08  s3  21:3  s2 þ 21:98  s  7:30 ð7Þ depends on these parameters, proposed MLPDA antenna is then
analyzed for getting desired resonance characteristics.
Now to cover the desired frequency range, number of dipoles
can be calculated from Eq. (8);
    3. Analysis of microstrip log periodic dipole array antenna
log KK 21 þ log ff l
N ¼1þ ð8Þ
u
Performance of this antenna is greatly influenced by some sen-
log s sitive geometrical parameters like length (Ln) and width of element
The width of dipole element can be determined by using the (Wn), spacing between elements (Sn) and feed length (K). Two
expression of characteristic impedance Z0 of cylindrical dipole [2] design parameters s and r as discussed in previous section, deter-
as given by Eq. (9); mine length and width of elements and spacing between them. In
    order to analyze the structure, different values of s and r are cho-
g0 Ln sen [24].
Z0 ¼  ln  2:25 ð9Þ
p an Detailed antenna analysis is performed by three ways; trans-
mission line equivalent circuit, impact of geometric parameter
where an = radius of equivalent cylindrical dipole, and Ln = half
variation and feeding mechanism. Following subsections empha-
length (Fig. 1). The width of printed microstrip dipole elements
sizes the analysis of antenna in these three ways.
are calculated by using formula [5] in Eq. (10);

W n ¼ p  an ð10Þ 3.1. Equivalent circuit of proposed MLPDA antenna

After determining the length and width of first dipole element
from Eq. (10), other lengths, widths and spacing’s can be calculated A transmission line equivalent circuit of proposed MLPDA
from well known equation of scale factor [24] as given by Eq. (11). antenna is presented to explain the resonance behavior of antenna.
The main purpose of equivalent circuit is not to exactly predict the
Ln Wn Sn input impedance but to give a physical insight into the behavior of
s¼ ¼ ¼ ð11Þ
Lnþ1 W nþ1 Snþ1 the structure. It consists of two parallel microstrip lines fed at the
edge with inner conductor of the SMA connector connected at
where n = 1, 2, 3. . .N. Another important geometrical parameter of
upper conducting microstrip, while the outer conductor connected
MLPDA is spacing factor, r which can be calculated by using Eq.
at the lower conducting microstrip as ground. So its equivalent can
(12).
be derived from the microstrip line geometry, with series induc-
Sn tance and parallel capacitance, for loss free medium. For lossy
r¼ ð12Þ structures, a serial resistance may be added to the inductance
4  Ln
and a parallel resistance may be added to the capacitance. The
Now based on above design equations from Eq. (1) to Eq. (12),
resulting model of the MLPDA is shown in Fig. 2. Where R, L, G
initial design of proposed MLPDA antenna is developed for fre-
and C represent the parameters of parallel microstrip line dis-
quency range of 5–6 GHz, with s = 0.9, r = 0.166 and feed length
tributed over the strip length. Since each conductor has significant
K = 6.5 mm (kmin/4) to match the 50 ohm input impedance of
length and width therefore it is having resistance and inductance.
SMA connector. Key design parameters of basic geometry used
Since there are two microstrip conductors close to each other there
for analysis are listed in Tables 1 and 2, respectively. Where L1 is
is a capacitance formed between them, filled with er = 4.4, material.
the half length of first dipole element, W1 is corresponding width
Since filled dielectric material is not perfect, therefore there is a
and S1 is the spacing between first and second dipoles. Other dipole
leakage through it. Hence it is to be represented by a shunt conduc-
lengths, widths and spacing’s can be calculated by using Eq. (11)
tance. Now to model dipole elements, first look at the geometry of
MLPDA antenna in which each dipole is formed by connecting two
Table 1 microstrip elements of specified length and width, to upper and
Key Design Parameters. lower conducting strips at the edge. Further to provide 180° phase
shift between each dipole element, they are connected asymmetri-
Parameters Value
cally, which realizes the criss-cross feed structure of conventional
Width of parallel strip line, WS 4.185 mm
LPDA. The dipole elements are modeled by equivalent circuit of
Feed length, K 6.5 mm
Scaling factor, s 0.90 microstrip patch which consists of a parallel resonant circuit as
Spacing factor, r 0.166 described in [25,26]. Further in order to realize electromagnetic
Effective dielectric constant, eeff 3.6962 coupling between microstrip elements of the same face, connected
Length of largest dipole, Lmax = 2 L8 16.92 mm in parallel with microstrip feed line, a coupling capacitor is intro-
Length of smallest dipole, Lmin = 2 L1 8.10 mm
Number of dipole elements, N 8
duced between them. So finally, dipole element is modeled as
two microstrip patches connected asymmetrically at edge with
 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274 1263

GND GND GND GND

Equivalent of microstrip
element (r, l, c)
1 3 C3-7 5 15
C1-5 C11- 15

C2-6 C12- 16
Equivalent of Strip Line 2 4 6 16
(R, L, G, C) C4-8

GND GND GND GND

Fig. 2. Equivalent circuit of microstrip log periodic dipole array Antenna.

upper and lower microstrip feed line. A parallel resonant circuit jxaMN X amn
Z in ¼ þ jx ð13Þ
consists of r, l, and c accounts for the resonance frequency of each x2MN  x2 ð1  jtandeff Þ mn–MN
x2  x2
mn MN
dipole element as shown in Fig. 2. The values of r, l, and c vary for
each dipole element according to their physical dimensions as 1
listed in Table 2. Capacitor C1–5, C3–7. . ..C11–15 accounts for coupling ¼ h i þ jxl0 ð14Þ
g MN þ j xcMN  xl1MN
between dipole elements of upper layer, while capacitor C2–6, C4–
8. . ..C12–16 represents coupling between dipole elements of lower
where
layer [Figs. 1(a) and 2]. These capacitors have been used to show
X amn
the electromagnetic coupling between the dipole elements of same 0
l ¼
x 2  x2
phases i.e. dipole elements 1, 5, 7. . .15 for upper layer and 2, 4, mn–MN mn MN
6,. . .. . .16 for lower layer. Capacitor C1–5, C3–7. . ..C11–15 and C2–6,
C4–8. . ..C12–16 are effective only when the spacing between dipole and
elements (Sn) decreases, but as the spacing (Sn) increases these 1 P
capacitors can be omitted from the equivalent circuit. tandeff ¼ ¼
Q 2xW e

3.2. Geometry of microstrip patch element and parallel strip line where aMN is attenuation constant in particular mode, Q is quality
factor, P is total loss power and We is average electric energy stored
Microstrip patch antenna is the basic element of the proposed in the cavity. For TM01 mode due to large values of (x2mn  x2MN ), l’ is
MLPDA. Hence its resonant behavior is explained by assuming it small and can be neglected. It shows that the equivalent is nothing
to be a magnetic wall cavity [27]. Each dipole with particular but the simple r, l, c parallel resonant circuit of Fig. 3.
width, W and length L can be represented by a parallel resonant Hence input impedance of the patch element at its resonant fre-
circuit as shown in Fig. 3.The input impedance of antenna patch quency has been defined as;
element Zin for frequencies near resonance of a mode (M, N) but 1 1
sufficiently away from all other resonances is given as [27] (Eqs. Z in ¼ Rin þ jX in ¼ 1 ¼ ð15Þ
r
þ jxc þ jx1 l Y in
(13) and (14));
where Yin is input admittance at the edge of the patch element. In
r proposed design, dipole elements are printed on two sides of the
substrate, so for calculation of feed width and patch dimensions a
dummy ground plane is assumed at substrate height of h/2 as
l shown in Fig. 4(a) and (b) [14]. As stated earlier, feed width
l’
(Ws = 4.185 mm) has been calculated by assuming 25 ohm micro-
strip with substrate height of h/2 to match the input impedance
Z in c of 50 ohm of SMA connector (Fig. 4(b)). The proposed MLPDA can
be considered as 16 patches connected asymmetrically with two
parallel strips on the both sides of substrate assuming a dummy
ground plane at the middle of substrate. The lengths and widths
of patches are determined by basic design equations of microstrip
LPDA as L2 = s-1 L1, L3 = s2 L1. . . and W2 = s1 W1, W3 = s2 W1. . .
as given by Eq. (11). Due to log periodic nature each patch element
Fig. 3. Equivalent of patch element [27]. resonates at frequency scaled by s i.e. f2 = s f1, f3 = s 2 f1,. . . fn = sn1f1
1264 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274

circuit to check the validity of model. Value of total admittance

z YT [Eq. (17)] has been utilized to simulate the model using ADS
software. The reflection coefficient S11 (dB) obtained from the
equivalent circuit is compared with that from the CST microwave
studio in the Fig. 5. The patterns of both curves are almost similar,
Conducng with very few discrepancies. Reason behind this discrepancy is due
Patch W to the fact that equivalent circuit is based on the assumption that
each antenna element operates in its dominant mode. But log peri-
y odic antenna array operates over broad band; hence it will encom-
pass some portion of higher order modes of antenna elements in
h/2 the array. It is observed that discrepancy is more prominent at
L higher frequencies. This is due to the fact that as one move towards
the higher frequency, effects of lower frequencies cannot be
accounted properly; hence bandwidth obtained from equivalent
circuit is less as it does not include the effect of mode coupling
x due to higher order modes.
(a)
3.4. Effect of varying scale factor (s)
Ws Dummy Ground
Plane
Proposed MLPDA antenna has been analyzed for varying scale
2C h/2 factor s in the range; 0.88–0.96 with the steps of 0.20, while keep-
ing other parameters fixed i.e. spacing factor, r = 0.166; feed
εr 2C h/2 length, K = 6.5 mm; spacing between first two element,
S1 = 2.70 mm; length of first element, L1 = 4.05 mm; and width of
first element, W1 = 0.88 mm. The basic geometry of antenna as
Ws given in Table 2 is used for this purpose. The resulting other geo-
metrical parameters of antenna for each value of s is listed in
(b) Table 3.The length L1 and W1 of first microstrip dipole elements
determines the upper cutoff frequency of antenna. The other dipole
Fig. 4. (a) Geometry of microstrip patch element [27] (b) Equivalent circuit for elements lengths and widths are calculated accordingly with vary-
calculation of feed width Ws [14]. ing values of s from Eq. (11). Effect of the variation of these geo-
metrical parameters has been observed through reflection
coefficient |S11| (dB) and realized gain (dB) as shown in Figs. 6(a)
with corresponding wavelength k2 = s1k1, k3 = s2 k1,. . . kn = sn+1
and (b) and 7 respectively.
k1 and phase constant b2 = s b1, b3 = s2 b1,. . . bn = sn1 b1 [28]. A sin-
It is observed that for increasing values of s, length (L) and
gle microstrip antenna is typically a narrowband device but as all
width (W) of other dipoles decreases. Also the spacing between
patches resonate at different frequencies according to their dimen-
the dipoles, from S2 to S8 decreases. For s = 0.96, length (and width)
sions (L and W), coupling between all patches results a broader
of the largest dipole is lowest in comparison of the values at other
bandwidth. It is very difficult to obtain the closed form equation
values of tau (L8 = 5.38 mm, and W8 = 1.17 mm). For this case lower
of exact input impedance (Zin) of the whole structure, so the equiv-
cutoff frequency (fl = 6.49 GHz, frl = 7.26 GHz) is obtained. Since
alent circuit representation of MPLDA gives more physical about
increments in the tau value decreases spacing between elements
behavior of antenna.
drastically, therefore the amount of coupling between the ele-
An approximation of total input impedance can be found by cal-
ments become over i.e. called over coupling. This situation leads
culating the input admittance of nth dipole element and adding all
to the reduction in the overall bandwidth. This fact can be evalu-
those values by assuming that each element resonates with log
ated from the numerical value of the data listed in Table 3.This
periodically in terms of frequency, wavelength and phase constant.
increased coupling shifts the active region of the antenna towards
The input admittance of nth dipole element could be given as, Eq.
smaller elements that leads towards slight increment in the upper
(16);
cutoff frequency. Also the resonant frequency of the antenna
1 1 increases with increasing s, because of the shift in active region.
Y in;n ¼ þ jxn c n þ ð16Þ As the spacing between the antenna elements controls the loop
rn jxn ln
size, impedance plot on smith chart for s = 0.96 makes largest loop,
where rn, cn and ln are corresponding values of r, l, c resonant circuit while impedance plots for the other lower values of tau (s = 0.92
for particular dipole element resonating at frequency fn. The total and s = 0.88) make smaller sized large No. of loops, those lie within
admittance can be calculated as, Eq. (17); VSWR = 2 circle. These show the largest value of the overall impe-
dance bandwidth. This fact can be verified by the Table 3 numerical
X
N
data.
YT ¼ Y in;n ð17Þ
n¼1
Variation in gain of MLPDA with varying values of s has been
shown in Fig. 7. It is observed that peak gain of MLPDA antenna
reduces with increased values of s with shifting of gain values
3.3. Validation through equivalent circuit towards right. For the small values of s = 0.88 L8 is largest and
equals to 9.91 mm. The antenna starts resonating at fl = 4.398 GHz
Proposed MLPDA antenna with complete geometrical descrip- as listed in Table 3. The active region of antenna lies near this lar-
tion of Tables 1 and 2 are now simulated using CST microwave stu- gest element, hence antenna radiates more effectively in this
dio for getting desired resonance curves. Antenna operates in the region and gain obtained is maximum (=6.16 dB at f = 4.7 GHz)
frequency range 4.835–7.44 GHz, with resonant frequency of (Fig. 7). Now the increased values of s reduce the length of largest
6.34 GHz. The proposed antenna also simulated using equivalent element with increased coupling between elements. This increases
 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274 1265

Fig. 5. Simulated reflection coefficient of MLPDA using CST and equivalent circuit.

Table 3
Geometrical description of MLPDA antenna with varying scale factor (s).

Parameter Values of tau symbol Bandwidth(GHz) Antenna Geometrical Dimensions Resonant frequency (GHz) BW (%)
(mm)
Ln
s ¼ Lnþ1 ¼ WWnþ1
n
¼ SSnþ1
n 0.88 4.398–7.354 = 2.965 L8 = 9.91 W8 = 2.15 S7 = 5.81 6.08 50.45
L7 = 8.72 W7 = 1.89 S6 = 5.11
L6 = 7.67 W6 = 1.66 S5 = 4.50
L5 = 6.75 W5 = 1.46 S4 = 3.96
L4 = 5.94 W4 = 1.29 S3 = 3.48
L3 = 5.22 W3 = 1.13 S2 = 3.06
L2 = 4.60 W2 = 1.00 S1 = 2.70
L1 = 4.05 W1 = 0.88 –
0.90 4.835–7.44 = 2.605 L8 = 8.46 W8 = 1.83 S7 = 5.08 6.32 42.44
L7 = 7.62 W7 = 1.65 S6 = 4.57
L6 = 6.85 W6 = 1.49 S5 = 4.11
L5 = 6.17 W5 = 1.34 S4 = 3.70
L4 = 5.55 W4 = 1.20 S3 = 3.33
L3 = 5.00 W3 = 1.08 S2 = 3.00
L2 = 4.5 W2 = 0.97 S1 = 2.70
L1 = 4.05 W1 = 0.88 –
0.92 5.375–7.529 = 2.154 L8 = 7.26 W8 = 1.57 S7 = 4.45 6.62 33.38
L7 = 6.67 W7 = 1.45 S6 = 4.09
L6 = 6.14 W6 = 1.33 S5 = 3.76
L5 = 5.65 W5 = 1.22 S4 = 3.46
L4 = 5.20 W4 = 1.13 S3 = 3.18
L3 = 4.78 W3 = 1.03 S2 = 2.93
L2 = 4.40 W2 = 0.95 S1 = 2.70
L1 = 4.05 W1 = 0.88 –
0.94 5.97–7.65 = 1.68 L8 = 6.24 W8 = 1.35 S7 = 3.91 6.92 24.66
L7 = 5.87 W7 = 1.27 S6 = 3.67
L6 = 5.51 W6 = 1.19 S5 = 3.45
L5 = 5.18 W5 = 1.12 S4 = 3.25
L4 = 4.87 W4 = 1.05 S3 = 3.05
L3 = 4.58 W3 = 0.99 S2 = 2.87
L2 = 4.30 W2 = 0.93 S1 = 2.70
L1 = 4.05 W1 = 0.88 –
0.96 6.49–7.79 = 1.3 L8 = 5.38 W8 = 1.17 S7 = 3.44 7.26 18.20
L7 = 5.17 W7 = 1.12 S6 = 3.31
L6 = 4.96 W6 = 1.07 S5 = 3.17
L5 = 4.76 W5 = 1.03 S4 = 3.05
L4 = 4.57 W4 = 0.99 S3 = 2.92
L3 = 4.39 W3 = 0.95 S2 = 2.81
L2 = 4.21 W2 = 0.91 S1 = 2.70
L1 = 4.05 W1 = 0.88 –
1266 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274

Fig. 6. Effect of varying scale factor, s (a) Reflection coefficient |S11| (dB) (b) Smith chart impedance plot.

the lower cut off frequency of antenna and shifts the active region
accordingly. Hence the peak values of gain obtained, shifted
towards right with decreased value. The decrement in the gain is
because of the fact that the overall current distribution of elements
becomes poor due to increased coupling between elements. As the
upper cut off frequency remains constant in all cases determined
from first element (L1 = 4.05 mm), radiation from this element is
weakest because it does not lie within active region of MLPDA. This
reduces the gain in this region as shown in Fig. 7 and gain curves
form negative slope as the frequency increases.
3.5. Effect of varying spacing factor (r)
In this case the spacing factor r has been varied in the range
0.14–0.22 with step size of 0.02, while other parameters was fixed
with scaling factor, s = 0.90 and feed length, K = 6.5 mm; length of
first element, L1 = 4.05 mm; and width of first element,
W1 = 0.88 mm. In this case length, width of dipole elements and
spacing between them were calculated according to a fixed value
of s = 0.90 (Eq. (11)), and their values are listed in Table 4. With
r varies from 0.14 to 0.22, spacing between elements increases
 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274 1267

Fig. 7. Effect of varying scale factor s on Realized gain of MLPDA.

Table 4
Geometrical description of MLPDA with varying spacing factor (r).

Parameter Values of sigma symbol Bandwidth(GHz) Antenna Geometrical Dimensions (mm) Resonant Frequency (GHz) BW (%)
Sn
r ¼ 4L 0.14 4.969–7.634 = 2.665 L8 = 8.46 W8 = 1.83 S7 = 4.27 6.58 42.29
n
L7 = 7.62 W7 = 1.65 S6 = 3.84
L6 = 6.85 W6 = 1.49 S5 = 3.45
L5 = 6.17 W5 = 1.34 S4 = 3.11
L4 = 5.55 W4 = 1.20 S3 = 2.80
L3 = 5.00 W3 = 1.08 S2 = 2.52
L2 = 4.5 W2 = 0.97 S1 = 2.27
L1 = 4.05 W1 = 0.88 –
0.16 4.864–7.488 = 2.624 L8 = 8.46 W8 = 1.83 S7 = 4.87 6.4 42.48
L7 = 7.62 W7 = 1.65 S6 = 4.38
L6 = 6.85 W6 = 1.49 S5 = 3.95
L5 = 6.17 W5 = 1.34 S4 = 3.55
L4 = 5.55 W4 = 1.20 S3 = 3.20
L3 = 5.00 W3 = 1.08 S2 = 2.88
L2 = 4.5 W2 = 0.97 S1 = 2.59
L1 = 4.05 W1 = 0.88 –
0.18 4.758–7.318 = 2.56 L8 = 8.46 W8 = 1.83 S7 = 5.49 6.2 42.39
L7 = 7.62 W7 = 1.65 S6 = 4.94
L6 = 6.85 W6 = 1.49 S5 = 4.45
L5 = 6.17 W5 = 1.34 S4 = 4.00
L4 = 5.55 W4 = 1.20 S3 = 3.60
L3 = 5.00 W3 = 1.08 S2 = 3.24
L2 = 4.5 W2 = 0.97 S1 = 2.92
L1 = 4.05 W1 = 0.88 –
0.20 4.649–7.160 = 2.511 L8 = 8.46 W8 = 1.83 S7 = 6.09 6 42.52
L7 = 7.62 W7 = 1.65 S6 = 5.48
L6 = 6.85 W6 = 1.49 S5 = 4.93
L5 = 6.17 W5 = 1.34 S4 = 4.44
L4 = 5.55 W4 = 1.20 S3 = 4.00
L3 = 5.00 W3 = 1.08 S2 = 3.6
L2 = 4.5 W2 = 0.97 S1 = 3.24
L1 = 4.05 W1 = 0.88 –
0.22 4.519–7.038 = 2.519 L8 = 8.46 W8 = 1.83 S7 = 6.69 5.8 43.59
L7 = 7.62 W7 = 1.65 S6 = 6.02
L6 = 6.85 W6 = 1.49 S5 = 5.42
L5 = 6.17 W5 = 1.34 S4 = 4.88
L4 = 5.55 W4 = 1.20 S3 = 4.39
L3 = 5.00 W3 = 1.08 S2 = 3.95
L2 = 4.5 W2 = 0.97 S1 = 3.56
L1 = 4.05 W1 = 0.88 –
1268 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274

while lengths and widths of the elements remained fixed, this will
decreases the coupling between elements and hence the active
region shifted towards longer elements. This results in shifting of
resonant frequency towards left side, Fig. 8(a). As the length and
width of dipole remains constant (with s = 0.90), bandwidth of
antenna does not change significantly, as in previous case of vary-
ing s, but there is a shift in the lower and upper cutoff frequencies
towards left side because of the shifting of the active region due to
decreased coupling. The result of decreased coupling between ele-
ments results in smaller loop area on impedance smith chart, plot-
ted for few values of r, Fig. 8(b). As can be seen that, impedance
curve for r = 0.22 covers smaller loop area than r = 0.18 and
r = 0.14.
Variation in the gain of antenna with different values of spacing
constant has been shown in Fig. 9. The length (Ln) and width (Wn)
of all dipole elements remain constant in this case as it is not

Fig. 8. Effect of varying spacing factor r (a) Reflection coefficient |S11| (dB) (b) Smith chart impedance plot.
 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274 1269

Fig. 9. Effect of varying spacing factor r on realized gain of MLPDA.

dependent on r (Table 4). Hence upper and lower cutoff frequen-
cies do not varies significantly as in the case of varying s. This will
cause maximum radiation at 5 GHz because the active region of
MLPDA lies near longest element L8 (=8.46 mm) which resonates
near this frequency for all values of r. There is increment in the
overall gain is observed for increasing values of r. This is because
of the fact that increased spacing between elements makes proper
current distribution among elements of antenna. This will cause
stronger radiation towards short element of MLPDA. Hence the
gain curves show overall increment throughout the bandwidth as
the sigma increases from 0.14 to 0.22 as shown in Fig. 9.
is soldered to coaxial pin of SMA connector while lower strip line is
connected with ground of connector. Since the feed is given to the
smallest dipole of length L1 and width W1, therefore feed length K
should be such that it matches with the input impedance of this
dipole. The feed length K is varied from kmin/8 to kmin/4 (3.25–
6.5 mm), while keeping other parameters fixed i.e. s = 0.90 and
r = 0.166 and simulation has been done with various values of feed
length with the step size 0.8125 mm.
It is observed that resonant pattern dip is best in the case
K = 5.6875 mm, as it is showing maximum value, 44.46 dB at

Table 5
3.6. Effect of varying feed length (K) Effect of feed length K on resonance characteristic.

Feed Length, K (mm) Resonant Frequency (GHz) |S11| dB

A Method of feeding any antenna is a very important. It is a very 3.25 6.48 26.7356
very crucial task because performance of antenna is affected sensi- 4.0625 6.44 29.6108
bly by this. In this work feed length has been optimized by ‘hit and 4.875 6.38 35.2447
trial method’. Proposed antenna has been fed by coaxial SMA con- 5.6875 6.36 44.4757
6.5 6.32 31.3353
nector through the edge. A strip line connected with upper dipoles

Fig. 10. Effect of varying feed length, K.

1270 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274

Fig. 11. Log periodic nature of MLPDA Antenna.

Fig. 12. Hard ware of proposed MLPDA (a) Upper Layer (b) Lower Layer.

Fig. 13. Measured and simulated results of |S11 | (dB).

 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274 1271

6.36 GHz. It means, in this case antenna is reflecting minimum
power and allowing maximum amount of power to propagates in
forward direction (Fig. 10). Other values of resonant frequencies
dips for different values of K are listed in Table 5.
3.7. Log Periodic behavior analysis of proposed antenna
The design of this MLPDA is based on frequency independent
antenna principle, which when applied to a periodic structure;
result in scaling of the dimensions from period to period so that
the performance is periodic with the logarithm of frequency [24].
This principle is used to design each microstrip dipole element of
this MLPDA. As discussed earlier that length, width and spacing
between two adjacent elements are related initially to the scale
factor s (Eq. (11)), so if one multiples the all dimensions of the
array by with s the element n become n+1 and the element n+1
become n+2.Consequently the array will have the same radiation
properties at all the frequencies which are connected by the scale
factor s. For example the resonant frequency of the first dipole ele-
ment is f1, then resonant frequencies of other dipoles are given by,
f2 = s f1, f3 = s 2 f1, f4 = s3 f1 and so on. Where
f2 f
ln ¼ ln s; ln 3 ¼ 2  ln s ð18Þ
f1 f1
To verify the log periodic nature of the antenna as described by
Eq. (18), input impedance of the antenna is plotted with respect to
logarithm of frequency for s = 0.88, and r = 0.166 and K = 6.5 mm
(Fig. 11). It is observed that antenna impedance varies periodically

(b) f = 6.2 GHz

(a) f = 5.2 GHz

(c) f = 7.4 GHz

E Plane

Fig. 14. Simulated E plane and H plane radiation patterns at (a) f = 5.2 GHz (b) f = 6.2 GHz (c) f = 7.4 GHz.
1272 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274

with logarithm of frequency for a specified bandwidth and satisfy
the Eq. (15), with each cycle being identical to the preceding one.
Other parameters that undergo similar variations are the pattern,
directivity, beam width etc., those can also be plotted and log peri-
odicity can be verified.
4. Results and discussions
To validate the results obtained from simulation, a prototype of
proposed MLPDA antenna is developed in a high frequency labora-
tory for a specific value of s = 0.90, r = 0.166 and K = 6.5 mm using
a simple FR4 substrate (er = 4.5 and width h = 1.6 mm) (Fig. 12) and
the results for reflection coefficient |S11| dB is observed through a
Vector Network Analyzer (VNA) – HP8720 B. Reflection coefficient
of simulated and measured MLPDA antenna is plotted in Fig. 13. It
is observed that antenna covers the useful frequency range of C
Band from (4.835–7.44 GHz) with resonant frequency of
6.34 GHz. Measured and a simulated result are matched satisfacto-
rily, validate the numerical simulation method, Finite Difference
Time Domain (FDTD) of CST and verifies the same through devel-
oped equivalent circuit. The differences may be due to effect of
SMA connector and mismatching tolerance. There is a slight shift-
ing of measured |S11| curve from 10 dB line in upper frequency
range (6.5–7 GHz), because of soldering effect of SMA connector
at lower end dipole element. As this dipole element decides upper

(b) f = 6.2 GHz

(a) f = 5.2 GHz

(c) f = 7.4 GHz

E Plane

H Plane

Fig. 15. Measured E plane and H plane radiation patterns at (a) f = 5.2 GHz (b) f = 6.2 GHz (c) f = 7.4 GHz.
 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274
Fig. 16. Measured and simulated results of gain (dB).
cutoff frequency of proposed MLPDA, there is more chance of devi-
ation of |S11| curve in this region due to soldering. Also the size of
this element is smallest one (L1 = 4.05 mm, W1 = 0.88 mm), causing
mismatch of |S11| curve due to manufacturing defects in high fre-
quency region.
The simulated and measured radiation pattern of the proposed
antenna is shown in Figs. 14 and 15, at three different frequencies
of 5.2 GHz, 6.2 GHz and 7.4 GHz. E and H field radiation pattern for
directivity (dBi), at these resonant frequencies shows stable radia-
tion pattern in end fire direction. There is slight deviation in simu-
lated and measured radiation pattern in forming main lobe,
because of the soldering of SMA connector near first microstrip
dipole element, affect high frequency region of proposed MLPDA.
There is a large variation in forming side lobes. The reason behind
this variation may be because of the dielectric substrate used is a
simple FR4 epoxy board with great loss. Realized gain of proposed
MLPDA has been plotted in Fig. 16 for simulated and measured val-
ues. Antenna shows positive gain throughout the bandwidth with
maximum value of 4.8 (dB) (measured value). Measured gain is
approximately 1 dB less than simulated value. This is because of
the fact that the substrate used is FR4 with loss tangent of the
order of 102 magnitude, which is greater than other Roger series
material (103 magnitude). The energy loss caused by it will make
a difference of reflection coefficient and gain. The influence from
soldering and fixture when measuring radiation pattern will affect
the measured result of side lobe level. Besides, the instability of FR-
4 material will also affect the measured result of antenna radiation
pattern.
5. Conclusion
This paper presents a Microstrip Log Periodic Dipole Array
antenna design based on geometrical parameter variation for
attainment of desired resonant frequency, operating bandwidth
and gain. Variations of scale factor, spacing constant and feed
length analyzed for different values, which significantly affects
bandwidth and gain of proposed MLPDA. A parametric analysis of
proposed MLPDA antenna has been performed by using FDTD
(Finite Difference Time Domain) based CST simulator software.
Log periodic nature of antenna is verified for one of these parame-
ter values. An equivalent circuit of the antenna is developed to give
physical insight of the structure. Geometry of microstrip patch ele-
1273
ment and parallel strip line has been explained with equivalent
lumped circuit elements. To validate the equivalent circuit of
MLPDA S11 parameter obtained from ADS which is further com-
pared with outcomes of CST microwave studio. Equivalent circuit
shows satisfactory approximation of S11 curve with reference to
CST simulation. Further, the validity of simulation has been
checked by developing a prototype of MLPDA and verified through
some measurements. Hence proposed MLPDA antenna is suitable
for wireless applications in C band, including WLAN applications
at 5 GHz band. Detailed analysis using parametric variations may
be helpful in optimization of antenna in terms of size, bandwidth
and gain.
References
[1] V.H. Rumsey, Frequency independent antennas, IRE Natl. Convention Rec. pt.1
(1957) 114–118.
[2] R.H. DuHamel, D.E. Isbell, Broadband logarithmically periodic antenna
structures, IRE Natl. Convention Rec. pt. 1 (1957) 119–128.
[3] R. Carrel, The Design of log-periodic dipole antennas, IRE Int. Conv. Rec., IX,
pt.1, 61–75, 1961.
[4] P.C. Buston, G.T. Thompson, A note on the calculation of the gain of log-
periodic dipole Antennas, IEEE Trans. Antennas Propag. 24 (1976) 105–106.
[5] C.K. Campbell, I. Traboulay, M.S. Suuthers, H. Kneve, Design of a stripline log-
periodic dipole antenna, IEEE Trans. Antennas Propag. AP-25 (5) (1977) 718–
721.
[6] P. Alakananda, G. Inderjeet, An analysis of log periodic antenna with printed
dipoles, IEEE Trans. Antennas Propag. AP-29 (2) (1981) 114–119.
[7] P.R. Pantoja, A.R. Sapienza, F.C. Medeiros, A microwave printed planar log-
periodic dipole array antenna, IEEE Trans. Antennas Propag. AP-35 (10) (1987)
1176–1178.
[8] N. Galchenko, Sergey Vartanyan, ‘‘Development of electromagnetic methods of
printed dipole antenna design,” in: International Conference on Mathematical
Methods in Electromagnetic Theory – MMET ’98, Vol. 2, 562–564, 1998.
[9] A.A. Gheethan, Dimitris E. Anagnostou, ‘‘Reduced size planar log-periodic
dipole arrays (LPDAs) using rectangular meander line elements,” in: Antennas
and Propagation Society International Symposium, IEEE, 1 – 4, San Diego, CA,
2008.
[10] D.E. Anagnostou, John Papapolymerou, Manos M. Tentzeris, Christos G.
Christodoulou, A printed log-periodic koch-dipole array (LPKDA), IEEE
Antennas Wireless Propag. Lett. 7 (2008) 456–460.
[11] A.A. Gheethan, Dimitris E. Anagnostou, ‘‘The design of reconfigurable planar
log-periodic dipole array (LPDA) using switching elements,” in: Antennas and
Propagation Society International Symposium, APSURSI ’09. IEEE,1-4, 2009.
[12] S. Lin, S. Luan, Y.D. Wang, X. Luo, X. Han, X.Q. Zhang, Y. Tian, X.Y. Zhang, A
printed log-periodic tree-dipole antenna (PLPTDA), Progr. Electromagn. Res. M
21 (2011) 19–32.
[13] T. Huang, Heng-Tung Hsu, ‘‘Log-periodic dipole array with improved front-to-
back ratio for universal UHF RFID handheld reader applications,” in: Antennas
and Propagation Society International Symposium (APSURSI) IEEE, 1 – 2, Jul.
2012.
1274 S.S. Pawar et al. / Engineering Science and Technology, an International Journal 20 (2017) 1260–1274
[14] G.A. Casula, Paolo Maxia, Giuseppe Mazzarella, Giorgio Montisci, Design of a
printed log-periodic dipole array for ultra-wideband applications, Progr.
Electromagn. Res. C 38 (2013) 15–26.
[15] G.A. Casula, Paolo Maxia, Giorgio Montisci, Giuseppe Mazzarella, Francesco
Gaudiomonte, A printed LPDA fed by a coplanar waveguide for broadband
applications, IEEE Antennas Wireless Propag. Lett. 12 (2013) 1232–1235.
[16] Y. Liu, Xiaozhong Shui, Hong-Da Lu, Li-Ming Si, Ya-Fen Ge, Xin Lv, ‘‘Design of
ultra-wideband printed log-periodic dipole antenna,” in: IEEE International
Conference on Microwave Technology & Computational Electromagnetics,
Aug. 2013, pp. 176–178.
[17] G.A. Casula, Paolo Maxia, ‘‘A multiband printed log-periodic dipole array for
wireless communications,” in: International Journal of Antennas and
Propagation, Hindawi Publishing Corporation, Article ID 646394, Volume
2014.
[18] G. Zhai, Yong Cheng, Qiuyan Yin, Leung Chiu, Shouzheng Zhu, Jianjun Gao,
Super high gain substrate integrated clamped-mode printed log-periodic
dipole array antenna, IEEE Trans. Antennas Propag. 61 (6) (2013) 3009–3016.
[19] C. Kang, Shu Lin, Zhi-Hua Zhao and Li-Wen Jing, ‘‘The simulation and
experiment of a non-cross-feeding printed log-periodic antenna,” Int. J.
Antennas Propag., Hindawi Publishing Corporation, Article ID 783405,
Volume 2015.
[20] S.S. Pawar, M. Shandilya, A compact wideband printed LPDA antenna for
WLAN applications in 5 GHz band, Int. J. Comput. Appl. 136 (10) (2016) 18–22.
[21] M. Abri, F.T. Bendimerad, N. Boukli-Hacene, M. Bousahla, A log periodic series-
fed antennas array design using a simple transmission line model, Int. J.
Electron. Commun. Eng. 2 (3) (2009) 161–169.
[22] F. Lei, Zeng Rui Li, Qing Xin Guo, Hui Zhang, Xue Qin Zhang, Jie Wang, Guo
Sheng Liu, Jun Hong Wang, Yaoqing Yang, A monolayer multi-octave
bandwidth log-periodic microstrip antenna, Progr. Electromagn. Res. Lett. 41
(2013) 97–104.
[23] T.A. Milligan, Modern Antenna Design, second ed., John Wiley & Sons Inc., IEEE
Press, New Jersey, 2005.
[24] C.A. Balanis, Antenna Theory Analysis and Design, third ed., John Wiley & Sons
Inc, New Jersey, 2005.
[25] J. Anguera, C. Punte, C. Borja, A procedure to design stacked microstrip patch
antennas based on a simple network model, Microwave Opt. Technol. Lett. 30
(3) (2001) 149–151.
[26] S.K. Jain, Shobha Jain, ‘‘Performance analysis of coaxial fed stacked patch
antennas,” Freq. J. RF Eng. Telecommun., (2014) 1–12. doi:10.1515/freq-2013-
0073.
[27] D.G. Fang, Antenna Theories and Microstrip antennas, Science Press, Taylor
and Francis Group, 2010.
[28] R.K. Mishra, ‘‘Transmission line model for log periodic microstrip antenna,” in:
Antennas and Propagation Society International Symposium (APSURSI), 2013
IEEE, Jul. 2013, pp. 168–169.