See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/224154186

Antenna array enhancement using mushroom-like electromagnetic band gap

(EBG)

Conference Paper · May 2010

Source: IEEE Xplore

CITATIONS READS

11 239

5 authors, including:

Tharek Abd Rahman Sharul kamal abdul rahim

Universiti Teknologi Malaysia Universiti Teknologi Malaysia
352 PUBLICATIONS   2,786 CITATIONS    249 PUBLICATIONS   1,993 CITATIONS   

SEE PROFILE SEE PROFILE

Mohd Tarmizi Ali Mohd Faizal Jamlos

Universiti Teknologi MARA Universiti Malaysia Perlis
209 PUBLICATIONS   974 CITATIONS    255 PUBLICATIONS   1,500 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

3dB Branch-Line Coupler with Improved Bandwidth Using PDMS and Zoflex Conductor View project

wireless antenna system for wearable application View project

All content following this page was uploaded by Tharek Abd Rahman on 03 May 2014.

The user has requested enhancement of the downloaded file.

 ANTENNA ARRAY ENHANCEMENT USING
MUSHROOM-LIKE ELECTROMAGNETIC
BAND GAP (EBG)
M.N.Md.Tan1, T.A.Rahman2, S.K.A.Rahim2, M.T.Ali1 and M.F.Jamlos2
1
Faculty of Electrical Engineering, Universiti Teknologi Mara (UiTM), Malaysia
2
Wireless Communication Center (WCC), Universiti Teknologi Malaysia (UTM), Malaysia

Email: mnor1408@yahoo.com, sharulka@yahoo.com, mizi732002@yahoo.com, tharek@fke.utm.my.

ABSTRACT- A double elements micro-strip antenna with II. THEORY OF EBG

mushroom-like electromagnetic band gap (EBG) inserted between
elements have been investigated. The performances of the antenna The parametric study on mushroom-like EBG structure
before and after the EBG are inserted between the micro-strip is presented in [10]. It focused on four main parameters that
patches are analyzed. FR4 substrate with dielectric constant of 4.7
affecting the overall performance of the antenna design. The
and 1.6mm thickness was simulated at operating frequency of
5.8GHz. The results shown that the performance of the array parameters namely, patch width W, the spacing between
changes if EBG parameters such as EBG size, spacing between mushroom-like EBG s, substrate thickness h and substrate
EBG, number of column of the EBG inserted between the elements permittivity εr. In this paper, the study is focusing not only on
and also the thickness of the substrate are changed. The simulation W, s and h as in [10], but also on the spacing between patch
results shows that by inserting (3x3) mm2 mushroom-like EBG with element, g and the number of rows of the EBG inserted
spacing of 0.6mm between EBG elements, the side-lobe (SL) between the patch elements.
improved from -19.dB to -24dB with 0.3dB gain increment and Mushroom-like EBG consists of a ground plane, a
back radiation pattern reduced significantly. Meanwhile, by
dielectric substrate, metallic patches and vias that connecting
inserting (3x4.5) mm2 mushroom-like EBG the mutual coupling
improved from -30.87dB to -52.78dB. Overall, this paper provides the patches to the ground plane. The structure of this EBG and
significant guidelines in enhancing the performance of the antenna its equivalent lumped LC elements is shown in Fig. 1. The
array integrated with mushroom-like EBG. inductance and capacitance of the circuit are due to the
shorting vias and the spacing between the adjacent metal
patches [6]. The resonant frequency, capacitance and
I. INTRODUCTION inductance can be obtained by using the equation given in (1),
(2) and (3).
EBG structures are capable to enhance the performance
of micro-strip antenna in terms of gain, side lobe, back lobe
level and also mutual coupling. This is due to EBG exhibits Mushroom-like EBG
w s + C -
frequency band-pass and band-stop that can block surface
wave excitation in the operational frequency range of the
L
antenna arrays [1]. Various EBG structure have been used by h via 2r
εr
different researches for different applications. For example,
Dumbbell EBG structure is used in [2], Step-like EBG ground
structure in [3] and Mushroom-like EBG structure in [4]-[7]. (b)
Their objectives in using EBG were basically either to Fig. 1 (a) Mushroom-like EBG structure (b) Lumped LC model
increase the gain, to reduce the side lobe and back lobe level,
to reduce the mutual coupling (S21) or to produce dual band to
operate at different frequencies. As mentioned in [8], most Resonant frequency, w …(1)
√LC
researches have not focused on array antenna, their study
emphasized only on a single micro-strip element. This paper is
focusing on two elements micro-strip antenna with Capacitance, C
W
cosh
W
...(2)
Mushroom-like Electromagnetic band gap inserted between it.
Due to great potential unique properties and winning features
compactness [9], Mushroom-like EBG is chosen. The main
objective is study and observes which parameter is more Inductance, 2 10 0.5 0.75 ...(3)
dominant in enhancing the performance of the antenna.

1
 III. SIMULATION OF LINEAR ARRAY Table 1
Comparisons details between the results obtained with and without EBG for
ANTENNA INTEGRATED WITH EBG
g = 10.8mm, s = 0.5mm, vary EBG size
Simulation of the linear array antenna integrated with Gain SL S11 S12 Radiation HPBW
mushroom-like EBG is divided into two different cases. Since (dB) ratio (dB) (dB) Efficiency (o)
FR4 has low surface wave, we are expecting that no (dB) (%)
improvement in term of mutual coupling before and after wo EBG 6.4 -15.4 -38.2 -21.44 55.16 50.1
wEBG
EBG inserted into array antenna. In order to verify it, the first 2.5x2.5 6.3 -15.1 -18.6 -20.9 55.51 50.6
simulation starts with single layer of FR4 substrate having 3x3 6.2 -15.1 -15.5 -20.6 55.54 51.2
1.6mm thickness and dielectric constant of 4.6 as shown in 4x4 6.2 -14.9 -18.9 -17.7 57.27 53.2
Fig. 2. The second case is by doubling the substrate the FR4
layer. As a result, the thickness becomes 3.2mm instead of
1.6mm. (B) g = 23.8mm, s = 0.5mm, vary EBG size

The same process in part (A) is repeated in this part. The

only different is that the spacing between patch is wider, that
is 23.8mm (three quarter wavelength) from edge to edge.
Table 2 shows the simulation result. Referring to Table 2, the
results obtained is very similar to Table 1. The performance of
Fig. 2 Single layer substrate of microstrip patch antenna separated by the antenna without EBG is about the same as the antenna
mushroom-like EBG structure with EBG except that the return loss is dropped from
-38.23dB to about -18dB. Mutual coupling obtained by
23.8mm spacing is better than 10.8mm spacing as shown in
(A) g = 10.8mm, s = 0.5mm, vary EBG size Fig. 4. This improvement is due to the spacing between the
patch itself not due to the EBG effect. The conclusion from
Simulation starts with a single layer FR4 substrate with these two simulation results, for single layer of 1.6mm FR4
10.8mm (half wavelength) spacing between patch with one substrate, EBG is not the suitable to improve the antenna
column of mushroom-like EBG with 0.5mm spacing between performance. In fact, the return loss produced after the EBG is
EBG. Three different EBG size were simulated and the results inserted is worse than the antenna without EBG
are shown in Table 1. The results in Table 1 clearly shown
that although different size of EBG inserted between patches,
the performances produced by the antenna with EBG (wEBG)
are about the same with the antenna without EBG (woEBG) 0

except that with EBG, the return loss dropped drastically. This
-10
is due to the effect of inserting the EBG with vias inserted in
S Parameters, dB

S11_EBG2.5x2.5
the middle of EBG structures in the substrate, increases the S12_EBG2.5x2.5
S11_EBG3x3
-20
permeability values and change the tangential loss of the S12_EBG3x3
S11_EBG4x4
substrate [1]. Fig. 3 shows that there is no band gap produced S12_EBG4x4
-30
by the EBG, hence, no reduction of mutual coupling achieved
in this simulation.
-40

-50
0 4.5 5.0 5.5 6.0 6.5 7.0 7.5

Frequency,GHz
-10
Fig. 4 Simulated S11 and S12 for different size of EBG with 0.5mm spacing
S parameters,dB

-20
Table 2
-30 Comparisons details between the results obtained with and without EBG for
g = 23.8mm, s = 0.5mm, vary EBG size
-40
S11_EBG2.5x2.5 Gain SL S11 S12 Radiation HPBW
-50 S12_EBG2.5x2.5
S11_EBG3x3 (dB) ratio (dB) (dB) Efficiency (o)
S12_EBG3x3 (dB) (%)
-60 S11_EBG4x4
S12_EBG4x4
wo EBG 7.3 -14.5 -38.9 -30.1 50.9 35.4
-70 wEBG
4.5 5.0 5.5 6.0 6.5 7.0 7.5 2.5x2.5 7.2 -14.4 -18.9 -30.4 50.4 35.5
Frequency,GHz 3x3 7.2 -14.5 -15.8 -30.3 50.1 35.5
Fig. 3 Simulated S11 and S12 for different size of EBG with 0.5mm spacing 4x4 7.3 -14.9 -20.5 -29.3 50.9 35.6

2
(C) g = 9.8mm, EBG size = (2.5mmx2.5mm), vary s spacing clearly shown that this spacing produced the best back
lobe. Another interesting point to talk about is that, the mutual
In this part, the substrate thickness is increased by coupling with 0.6 spacing shows a lot of improvement
doubling the layer with the same value of g as in (A). EBG compares to other spacing that is -53.92dB. The only problem
size of (2.5mmx2.5mm) was used for these simulations with is that this value is not obtained at 5.8GHZ but at 6.06GHz
9.8mm (half wavelength) spacing between antenna elements. instead (not shown in this paper).
The spacing between EBG was varied from 0.5mm to 0.7mm.
The detail results obtained from these simulations is shown in 0

Table 3 and simulated S11 and S12 graph are shown in Fig. 5. 30
0.8
330

Comparing the performance between the antenna with EBG

0.6
and without EBG, with EBG the performance is dropped for 60 300
all parameters. EBG with 0.5mm spacing yield the worse 0.4

performance in terms of gain, SL ratio, S11, S12 and efficiency. 0.2

EBG with 0.6mm and 0.7mm produced about the same
performance except that 0.6mm spacing has good return loss 90
0.8 0.6 0.4 0.2
0.0
0.0 0.2 0.4 0.6 0.8
27

and efficiency. There are no improvements of mutual coupling 0.2

and gain for all three spacing of EBG.

0.4
120 240
0 woEBG
0.6 EBG3x3_s0.5
EBG3x3_s0.6
-5 EBG3x3_s0.7
0.8
150 210
-1 0
180
-1 5

-2 0 Fig. 6 Polar pattern for linear array with and without EBG.
-2 5 S 1 1 _ w o EB G
S 1 2 - w o E BG
S 1 1 _ 2 . 5 x2 . 5 _ s0 . 5
-3 0 S 1 2 _ 2 . 5 x2 . 5 _ s0 . 5
S 1 1 _ 2 . 5 x2 . 5 _ s0 . 6
Table 4
-3 5
S 1 2 _ 2 . 5 x2 . 5 _ s0 . 6 Comparisons details between the results obtained with and without EBG
S 1 1 _ 2 . 5 x2 . 5 _ s0 . 7
S 1 2 _ 2 . 5 x2 . 5 _ s0 . 7 for g = 9.8mm, EBG size of 3mmx3mm ,vary s.
-4 0
4 .5 5.0 5 .5 6.0 6 .5 7.0 7 .5

F re q u e n cy ,G H z Gain SL S11 S12 Radiation HPBW

Fig. 5 Simulated S11 and S12 for (2.5mmx2.5mm) EBG size with variation of (dB) ratio (dB) (dB) Efficiency (o)
spacing between EBG (dB) (%)
wo EBG 6.6 -19.1 -38.1 -29.8 60.9 52
Table 3 EBG
Comparisons details between the results obtained with and without EBG spacing
for g = 9.8mm, EBG of (2.5mmx2.5mm), vary s 0.5mm 6.0 -12.3 -16.3 -14.6 47.4 40.6
0.6mm 6.9 -24.3 -16.9 -14.3 55.3 44.1
Gain SL S11 S12 Radiation HPBW 0.7mm 6.0 -18.5 -18.4 -14.6 51.9 50.9
(dB) ratio (dB) (dB) Efficiency (o)
(dB) (%)
wo EBG 6.6 -19.1 -38.1 -29.8 60.9 52 (E) g = 9.8mm, s = 0.6mm, vary EBG width
EBG
spacing
0.5mm 4.7 -14.0 -16.7 -13.6 49.9 60.1 As mentioned in part (D), 0.6mm spacing produced the
0.6mm 5.6 -16.1 -22.9 -14.9 54.6 57.4 best mutual coupling. The only problem is that, this value is
0.7mm 5.7 -16.4 -15.9 -17.4 54.1 56.7 not obtained at the required frequency. The main target in this
part is to investigate the parameter of EBG that can adjust the
frequency of the best the mutual coupling to the required
(D) g = 9.8mm, EBG size (3mmx3mm), vary s frequency. 0.6mm spacing between EBG is used for all
simulation. The EBG size is varied until the best value of
The simulation in (C) is repeated in this part. mutual coupling is centered to 5.8GHz. The simulations result
(2.5mmx2.5mm) EBG is replaced by (3mmx3mm) EBG. The is shown in Table 5. The simulation starts with EBG size of
spacing range varied from 0.5mm to 0.7mm and the (3mmx4mm). As usual, the performances with EBG inserted
simulation results obtained is shown in Table 4. Overall, the in the linear array antenna are dropped. The best point of S12 is
performances of the antenna with EBG are dropped except the -36.38dB at 6.08GHz. This EBG is replaced with
antenna with 0.6mm spacing EBG. EBG with 0.6mm spacing (3mmx5mm) and the deepest point is moved to 5.988GHz
yielded the highest gain, efficiency and SL ratio. The gain frequency with S12 of -33.83dB. It is followed by using EBG
increased from 6.6dB to 6.9dB with very good SL ratio of size of (3mmx6mm) and the best point is 5.85GHz with S12 of
-24.3dB compared to only -19.1dB without EBG. The -40.19dB. Finally, the EBG size of (3mmx6.3mm) is inserted
efficiency for all three spacing is dropped as well but 0.6mm into the array and now the best point of S12 is centered to
spacing has the highest value. Refer Fig. 6, EBG with 0.6mm almost 5.8GHz with S12 of -33.72dB. This is shown clearly in

3
Fig. 7. Refer to Table 5, every time the EBG size is increased, EBG size of (3mmx4.5mm) is selected to get the required
the chances to get better performance is higher. The gain in frequency. Mutual coupling produced by this EBG is
increased from 5.2dB to 6.8dB when the size of EBG -52.78GHz at 5.81GHz which is much better than the previous
increased from (3mmx4mm) to (3mmx6) mm. The same trend result. The changes of mutual coupling are clearly shown in
also goes to S11, S12 and efficiency. Fig. 8. Both table 5 and 6 proved that in order to adjust the
frequency, the width of EBG is the most dominant parameter.
The only problem is that, the side-lobe ratios are dropped
0
drastically.

-10
Table 6
Comparisons details between the results obtained with and without EBG
for s = 22.8mm, g = 0.5mm with variation of EBG width
S parameters, dB

-20
Gain SL S11 S12 Radiation HPBW
(dB) level (dB) (dB) Efficiency (o)
-30 S11_3x4_s0.6
(dB) (%)
S12_3x4_s0.6 wo EBG 6.6 -10.9 -30.9 -29.8 60.2 52
S11_3x5_s0.6 EBG size
S12_3x5_s0.6
-40 S11_3x6_s0.6 3x4 7.2 -7.1 -21.4 -33.0 62.0 31.1
S12_3x6_s0.6 3x5 6.5 -6.1 -19.6 -34.1 59.8 33.5
S11_3x6.3_s0.6
S12_3x6.3_s0.6 3x4.5 6.4 -5.7 -19.3 -40.6 58.8 31.2
-50
4.5 5.0 5.5 6.0 6.5 7.0 7.5

Frequency, GHz 0

Fig. 7 Simulated S11 and S12 spacing between EBG of 0.6mm with variation
of EBG width.
S Parameters, dB -10

Table 5 -20
Comparisons details between the results obtained with and without
EBG for g = 9.8mm, s = 0.6mm, vary EBG width
-30 S11_woEBG
S12_woEBG
Gain SL S11 S12 Radiation HPBW S11_3x4_s0.5
(dB) level (dB) (dB) Efficiency (o) S12_3x4_s0.5
(dB) (%) -40 S11_3x5_s0.5
S12_3x5_s0.5
wo EBG 6.6 -19.1 -38.1 -29.8 60.9 52 S11_3x4.5_s0.5
EBG size S12_3x4.5_s0.5
3x4 5.2 -6.2 -19.3 -13.7 53.9 39.5 -50
4.5 5.0 5.5 6.0 6.5 7.0 7.5
3x5 6.6 -8.0 -16.8 -17.1 56.4 35.6
3x6 6.8 -9.7 -27.2 -29.6 58.93 39.8 Frequency,GHz
3x6.3 6.5 -9.2 -32.6 -33.3 58.2 40.3 Fig. 8 Simulated S11 and S12 spacing between EBG of 0.5mm with variation of
EBG width.

(F) g = 22.8mm, s = 0.5mm, vary EBG width (G) G = 22.8mm, s = 0.5mm, EBG size of (3mmx3mm),
vary column
In this part, the simulation is very similar to the
simulation done in part (E). The spacing between patch is fix All the simulation done previously is using only one
to 0.5mm and the spacing between patch is bigger than in (E) column of EBG between the patch. In this part, the number of
that is 22.8mm. The EBG size is varied in order to achieve the column is varied from one column to 3 columns with fix
best mutual coupling value produced after the EBG is inserted spacing of 0.5mm between EBG and (3mmx3mm) EBG size.
between the patch. The simulation results are shown in Table Table 7 shows the simulation results and Fig. 9 shows the
6. The mutual coupling before the EBG is inserted between simulated S11 and S12. The results show that at frequency of
the patch is -30.21dB. After the (3mmx4mm) is inserted, the 5.8GHz, the performance is dropped when the number of
mutual coupling is increased to -33dB at 5.8GHz frequency column is increased. One column of EBG produced the best
but the best point with this EBG size is at 5.87GHz with value performance compared to two and three columns. In fact, the
of -39.97dB. Result obtained from part (C) shows that in order gain and efficiency of one column EBG is better than the
to reduce the frequency of the deepest point, the width of the antenna without EBG. The gain and efficiency improved
EBG need to be increased. Based on that, EBG with about 0.2dB and 4.12% respectively. Referring to Fig. 9, by
(3mmx5mm) is placed between the patch resulting mutual increasing the number of column, S12 shows a lot of
coupling of -38.29dB at 5.74GHz which is lower than the improvement. S12 for 1 column EBG is -29.86 dB at .854GHz,
required frequency. Hence, the width of EBG to be selected 2 columns EBG S12 is -43.27dB at 5.934GHz and for 3
should be lower than 5mm but higher than 4mm. In this case, columns S12 is -52.78dB at 5.96GHz.

4
 Table 7: REFERENCES
Comparisons details between the results obtained with and without EBG
for g = 22.8mm, s = 0.5mm with variation number of column. [1] Z.Illuz, R.Shavit and R. Bauer, “Micro-strip Antenna Phased Array
with Electromagnetic Band-Gap Substrate”, IEEE Transactions on
Gain SL S11 S12 Radiation HPBW Antenna and Propagation, Vol. 52, No 6, June 2004.
(dB) level (dB) (dB) Efficiency (o) [2] A.Yu and X.Zhang, “A Novel to Improve the Performance of Micro-
(dB) (%) strip Antenna Arrays Using a Dumbbell EBG Structure”, IEEE
wo EBG 6.6 -10.9 -30.1 -30.2 60.1 35.7 Antennas and Propagation Letters, Vol.2, 2003.
EBG [3] F.Yang, C.S.Kee and Y.Rahmat-Samii, “Step-like and EBG Structure
column to Improve the Performance of Patch Antennas on High Dielectric
1 8.0 -10.3 -23.7 -29.5 64.3 34.0 Substrate”, Antennas and Propagation Society International
2 7.3 -7.0 -22.3 -27.9 63.5 31.1 Symposium, Vol 2, 2001.
3 6.7 -5.4 -20.8 -25.0 62.0 29.8 [4] L-J. Zhang, C-H.Liang, L,Liang and L.Chen, “A Novel Design
Approach for Dual-Band Electromagnetic Band-Gap Structure”,
Progress in Electromagnetic Research M, Vol 4, 2008.
0
[5] F.Yang and Y.Rahmat-Samii, “Applications of Electromagnetic
Band-Gap (EBG) Structures in Microwave Antenna Design”, 3rd
International Conference on Microwave and Millimeter Wave
-10
Technology Proceedings, 2003.
[6] M.Fallah and L.Shafai, “Enhanced Performance of a Micro-strip
-20 Patch Antenna using a High Impedance EBG Structure”, Antennas
S parameters,dB

and Propagation Society International Symposium, Vol 3, 2003.

-30 [7] T.Masri, M.K.A.Rahim, F. Zubir, O.Ayop, M.N.A.Karim, and
H.A.Majid, “Micro-strip Antenna Array's Gain Enhancement via
-40
Electromagnetic Band Gap Structures”, IEEE International RF and
S11_3x3_1col
S12_3x3_1col Microwave Conference. RFM2008, 2008.
S11_3x3_2col [8] F.Yang and Y.Rahmat-Samii, “Micro-strip Antennas Integrated with
-50 S12_3x3_2col
S11_3x3_3col Electromagnetic Band Gap (EBG) Structures: A Low Mutual
S12_3x3_3col Coupling Design for Array Applications”, IEEE Transactions on
-60 Antenna and Propagation, Vol. 51, No 10, Oct 2003.
4.5 5.0 5.5 6.0 6.5 7.0 7.5
[9] L.Li, X-J.Dang, B.Li and C-H.Liang, “Analysis and Design of
Frequency,GHz Waveguide Slot Antenna Array Integrated with Electromagnetic
Fig. 9 Simulated S11 and S12 spacing between EBG of 0.5mm, EBG size of Band Gap (EBG) Structures”, IEEE Antennas and Wireless
(3mmx3)mm with variation number of column. Propagation Letters, Vol 5, 2006.
[10] F.Yang and Y.Rahmat-Samii, Electromagnetic Band Gap Structures
in Antenna Engineering, 1st Edition, Cambridge University Press,
2009.
IV. CONCLUSION

This paper is basically study the parameters of the

mushroom-like EBG. The results shows that not all the EBG
structure suitable to improve the performance of the antenna.
In fact there are some case showed that the performance of
antenna become worse when the EBG inserted between the
micro-strip patch. The selection of the size of EBG, the
number of column and the spacing/gap between the patch play
an important role in order to improve the performance of the
antenna.

ACKNOWLEDGEMENT

The authors would like to thank to everyone for their

helps and supports in completing this project especially to
University Technology Malaysia (UTM) and also to Wireless
Communication Center (WCC). Great thanks also go to
University Technology Mara (UiTM) for the financial
supports.

View publication stats