This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/LAWP.2018.2880926, IEEE
Antennas and Wireless Propagation Letters

SIW-Integrated Parasitic DRA Array: Analysis,

Design and Measurement
Wael M. Abdel-Wahab, Mona Abdallah, Jonathan Anderson, Student Member, IEEE, Ying Wang, Senior Member,
IEEE, Hussam Al-Saedi, Student Member, IEEE, and Safieddin Safavi-Naeini, Life Fellow, IEEE

Abstract— The modeling and design of a new substrate using parasitic elements include less spurious radiation, lower
integrated waveguide (SIW) series fed dielectric resonator loss, simpler fabrication, and lower cost.
antenna (DRA) parasitic array are presented. Parasitic DRAs In this paper, a new design of SIW series fed DRA parasitic
are added on both sides of each active element of a DRA fed by array for MM-wave applications is presented. In this design,
longitudinal slot on SIW. The E-coupling mechanism between a
additional parasitic DRAs are added on both sides of each
driven DRA and two parasitic DRA elements with low dielectric
constant is discussed. Also, the impact of the coupling gap on active element of a DRA fed by longitudinal slot on SIW to
reflection coefficient, impedance bandwidth, and gain and replace the lossy planar feeding network, such as the one in
radiation pattern is presented. A four-element antenna array [11]. Both the driven and parasitic DRA elements are made of
operating at the millimeter-wave band (36GHz – 39GHz) is lower dielectric constant compared to the ones used in [9] and
designed and fabricated. The measured radiation pattern [10]. Although adding these parasitic elements to the original
demonstrates a broadside beam with a maximum gain of 12 dBi array antenna design results in improved antenna gain, the
over an impedance bandwidth (with |S11| < -10 dB) of 3.3 GHz. effects on DRA’s resonance frequency, impedance bandwidth,
and reflection coefficient have not been discussed before. A
Index Terms — Substrate integrated waveguide (SIW), four-element antenna array operating at the MM-wave band
radiation pattern, dielectric resonator antenna (DRA), and (36GHz – 39GHz) is designed and fabricated. The
coupling. measurement results show an impedance bandwidth of 3.3
I. INTRODUCTION GHz and a maximum gain of 12dBi, which is 1.4dB higher
than the array without parasitic elements in [6].
Dielectric resonator antennas (DRAs) have been studied
extensively due to their many advantages [1, 2]. They are
compact, lightweight, low cost, and compatible with various
feeding methods. Additionally, DRAs can achieve high
radiation efficiency, due to the absence of conductor loss, and
wide bandwidth. Recently substrate integrated waveguide
(SIW) as the feeding structure has been explored, which
further minimizes the conductor loss compared to other
traditional low-cost feeding networks [3, 4]. The SIW-fed
DRAs and arrays have been reported for millimeter-wave
(MM-wave) applications, demonstrating high gain and high
radiation efficiency [4-6].
Antenna performances can be improved by using parasitic
elements of an antenna, which are not directly connected to a
feeder, but coupled to the antenna, only by the fields [7-10].
Parasitic elements for DRAs are used to increase the
impedance bandwidth of the antenna in [7] and [8]. The Fig. 1.SIW-DRA parasitic sub-array fed by the SIW- longitudinal slot
impedance bandwidth can be enhanced when parasitic
II. SIW-INTEGRATED PARASTIC ARRAY:
dielectric resonators of different dielectric constants and sizes
PARAMETRIC STUDY
are positioned next to the active dielectric resonator. Also, by
placing parasitic elements very close to the active element, As shown in Fig. 1, a rectangular DRA sub-array consists
high boresight gain with fixed broadside or van beam patterns of a slot-excited (driven) element and two parasitic elements.
is achieved [9]. Passive phased array based on parasitic DRA For simplicity and ease of fabrication, all DRAs are made of
elements can be formed by using reactive loadings on the the same dielectric material, Rogers RT/6010 with εr_DRA =
parasitic DRAs [10, 11]. By adjusting the relative phase 10.20 and have the same dimensions to support the same
among the elements, the main beam can be scanned over ± 30˚ resonance frequency. Longitudinal slot is etched on top of the
[10]. In large scale parasitic phased arrays, the perforation SIW to couple energy to the driven DRA element as in [12] to
technique [11] was employed to integrate both driven and excite the DRA’s fundamental mode, TE11δ. Investigating the
parasitic DRA elements to avoid the alignment and bonding dominant electric field component Ey of the fundamental
challenges of DRA elements to the feeding board. Compared mode inside and in a close vicinity of the isolated DRA (no
to directly feeding each dielectric resonator, the advantages of parasitic elements exist) along the E-plane (yz-plane), it is

1536-1225 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/LAWP.2018.2880926, IEEE
Antennas and Wireless Propagation Letters

found that some field leaks out from the DRA’s two sides, coherency between the electric fields, Ey components, as do
which can be used as excitation sources (E-coupling) as increases as explained in [10]. This effect was studied more
shown in Fig. 2(a). By inserting two parasitic DRAs in these closely by investigating both the magnitude and phase over
regions to form a sub-array antenna, their corresponding the antenna aperture for separation distances do =1.20 mm
modes can be excited for radiation with simple excitation (reference) and do =5.0 mm, respectively at the fo= 36.68 GHz,
scheme as shown in Fig. 2(b). as shown in Fig. 4 and Fig. 5, respectively. It is found that the
The SIW-DRA parasitic sub-array in Fig. 1 was simulated two parasitic DRA elements are excited almost in phase with
by the EM full-wave solver, HFSS, to study the impact of half-magnitude of driven DRA element when do =1.20 mm as
these parasitic elements on the overall far-field and near-field shown in Fig. 4(a) and Fig. 4(b), which accounts for the gain
characteristics, such as resonance frequency, reflection enhancement in Fig. 3(b). However, when do =5.0 mm, the
coefficients, impedance bandwidth, and gain. It is observed parasitic DRAs are less excited in magnitude, but with almost
that as the separation distance between the driven DRA and 180˚ out-of-phase w.r.t. the driven DRA, as shown in Fig. 5(a)
the two parasitic DRA elements do increases from 0.1 mm to and Fig. 5(b), which degrades the overall radiation, mainly
1.9 mm, the resonance frequency, fo (defined by the minimum from the driven DRA, causing gain degradation as shown in
reflection coefficient), shifts from 39.26 GHz to 35.95 GHz Fig. 3(b). To confirm, as the separation distance increases to
with |S11| less than -15 dB, as shown in Fig. 3(a). When do do = 8.0 mm (not shown in Fig. 4(b)), the effect of the
increases above 3.50 mm, fo becomes stable at 36.95 GHz, parasitic DRAs on the gain becomes negligible and the gain
which is close to the isolated DRA’s resonance frequency increases to 4.0 dB similar to the case without parasitic
(without parasitic elements). elements.
40 -10

Min. Reflection Coefficient, S11 (dB)

39 -15
Resonanace Freq., f (GHz)
o

38 -20

37 -25

36 -30
(a)

35 -35
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Seperation Distance, do (mm)

(a)

Realized Boresight Gain (dB) at min. S11

16 10

14
Impedance Bandwidth (%)

8
12

10
6

4
6

4
2
2
(b)
Fig. 2. (a) Normalized electric field distribution, Ey/Eymax along the E- 0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
0
plane at a distance 2 mm from the DRA’s top radiating surface at 36.78 Seperation Distance (mm)
GHz, (b) Electric field distribution (V/m) (b)
Investigations about the impedance bandwidth and gain are Fig. 3. Variation of (a) the resonance frequency and the minimum reflection
coefficients, |S11| (dB), and (b) impedance bandwidth (%) and the realized
shown in Fig. 3(b). A high realized boresight gain (θ=0˚) of 7
gain (dB) at minimum reflection coefficient of the antenna with do.
- 9.7 dB can be achieved when 0.1 mm ≥ do ≥ 2.7 mm with
shift in the fo. When do increases above this range, the gain To complete this analysis, the radiation patterns of both the
decreases significantly. However, over the same distance driven DRA element (without parasitic DRAs) and the sub-
range, the bandwidth level varies between 5.2% -13.10%, as array (with parasitic DRAs) are studied (not shown for brevity)
shown in Fig. 3(b). The degradation of gain in Fig. 3(b) is at a separation distance do=1.2 mm to provide the highest gain
attributed to the deterioration of the magnitude and phase of 9.30 dB with reasonable impedance bandwidth of 6.15% at
36.68 GHz. It is found that the radiation pattern becomes

1536-1225 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/LAWP.2018.2880926, IEEE
Antennas and Wireless Propagation Letters

narrower (more directive), especially in E-plane (yz-plane) at The SIW series-fed parasitic DRA array was designed and
the boresight direction compared to the isolated DRA due to RT/Duroid 6002 substrate material from Rogers (ɛr=2.87, and
the increase of antenna aperture’s effective area [10]. tanδ=0.002) with 20 mil thickness was used to build the SIW
feeding structure. RT/Duroid 6002 was recommended by the
manufacturer over RT/Duroid 5870 used in [6]. The design
parameters D and Xsc are optimized to achieve a wide
bandwidth for the reflection coefficient. The impact of
introducing parasitic DRA elements was studied numerically
by HFSS and the results are shown in Fig. 6 and TABLE I.

Fig. 4. Electric field, Ey, distribution over the antenna aperture for
separation distances do=1.20 mm, at 36.68 GHz, (a) Magnitude (v/m), (b)
Phase (Degrees).

Fig. 6. Reflection coefficient (|S11|) of the SIW-DRA array with and without
parasitic DRA elements. Both designs use RT/Duroid 6002 as SIW substrate.

TABLE I IMPACT OF ADDING PARASITIC ELEMENTS.

Array w/o Array with
Parameters parasitic parasitic
elements elements
Two freq. @ ( Min. S11) GHz 36.7, 37.8 37.3 , 38
Center freq., fo (GHz) 37.3 37.7
10 dB Impedance Bandwidth
2.50 2.50
(GHz)
Gain (dB)at fo 11.0 13.0
Efficiency (%) @ fo 91 91

It was found that adding parasitic elements to the array

shifts the |S11| response up in frequency by approximately
Fig. 5. Electric field, Ey, distribution over the antenna aperture for do=5.0 mm 400MHz as shown in Fig. 6. However, the gain is increased
separation distance at 36.68 GHz, (a) Magnitude (v/m), (b) Phase (Degrees). by ~ 2.0 dB with almost no impact on the radiation efficiency
Another possible parasitic sub-array could be formed by and impedance bandwidth as shown in Table I. Ideally, gain
placing the parasitic DRA elements along the driven DRA’s of 15-16 dB was expected, but it drops by ~ 2-3 dB due to
width. In this case, the coupling between the driven DRA’s reduction in the total effective aperture area as the individual
mode and parasitic element is attributed to the H-coupling. It effective areas of subarrays overlap at D=2.60mm.
is found that the H-coupling is not as efficient as the E- Afterwards, it was fabricated using a multi-layer PCB
coupling, especially when SIW longitudinal coupling slot is fabrication process, similar to the one described in [13]. The
employed; therefore, only E-coupled SIW-integrated parasitic fabricated structure consists of two layers, namely the SIW-
DRA array is considered in this paper. fed layer and the DRA elements. The DRAs are cut to the
correct sizes, and then bonded to one side of the SIW using
III. SIMULATION AND MEASURED RESULTS low loss thin adhesive bonding material. The fabricated
Once properly designed, the sub-array element can be antenna prototype is shown in Fig. 7. For testing of the array,
treated as a regular single antenna element in an array design. a coaxial line to SIW transition is designed and added to the
It is therefore important to study how the performance of the array as can be seen from the front and back views in Fig. 7.
sub-array changes as the E-coupling changes, which have The fabricated prototype is measured to determine the
been thoroughly investigated in the previous section. On the reflection coefficient, radiation pattern, and gain.
other hand, the H-coupling between sub-arrays was found to The EM simulated and measured reflection coefficient of
be not as strong as E-coupling. the fabricated antenna array is shown in Fig. 8. The measured

1536-1225 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/LAWP.2018.2880926, IEEE
Antennas and Wireless Propagation Letters

impedance bandwidth (with |S11| < -10 dB) is 3.30 GHz (35.60 IV. CONCLUSIONS
GHz-38.90 GHz), which is larger than EM result of 2.50 GHz A new design of parasitic SIW-DRA array is presented.
bandwidth. The difference between the two results is Parasitic DRAs are added on both sides of each driven DRA.
attributed to the SIW fabrication tolerances and the The coupling mechanism between the driven and parasitic
imperfection in shaping of the DRA blocks as shown in Fig. 7. DRA elements is studied. Results and discussions for the
However, both results show the same trend over the band with impact of the coupling on antenna performance are presented.
almost the same centre frequency. The experimental data are presented for a four-element
antenna array operating at the MM-wave band.
XZ Plane
0

-5

-10

-15

Power (dB)
-20

-25
COP-Sim. (HFSS)
-30
XP-Sim. (HFSS)
COP-Meas.
-35
XP-Meas.
Fig. 7. The fabricated prototype (front and back views) and the coaxial- -40
-80 -60 -40 -20 0 20 40 60 80
connector used for testing. The optimized dimensions (mm) are: D=2.60mm, Theta (Degrees)
XSC=4.8, slot length=3.2, and slot width=0.30.
YZ-Plane
0

-5

-10

-15
Power (dB)

-20

-25

COP-Sim.(HFSS)
-30
XP-Sim(HFSS)
-35 COP-Meas.
XP-Meas.
Fig. 8. Reflection coefficient (|S11|) of the fabricated four element SIW-DRA -40
-80 -60 -40 -20 0 20 40 60 80
parasitic array and the EM simulation of the structure. Theta(Degrees)
The radiation pattern of the array antenna was tested and it Fig. 9. EM simulated and measured radiation patterns of the array antenna on
shows a good consistency over the operating band. For brevity, H-plane (xz-plane) and E-plane (yz-plane) at 36.50 GHz.
the pattern in two orthogonal planes, H-plane (xz-plane) and 14
E-plane (yz-plane) at 36.50 GHz is shown in Fig. 9. The
13
measurement is consistent with the EM simulations in Fig. 9,
especially close to the main beam of the COP radiated field. 12

However, at higher angles from the boresight, the 11

Gain (dB)

difference increases, especially in the E-plane due to the wide Sim. (HFSS)
10
beam pattern which could not be captured accurately by the Meas.
planar NF scanner as shown in Fig. 9. Both radiation pattern 9
results demonstrate a broadside beam with maximum linear 8
polarized (LP) gain of 11.90 dB and COP to XP polarization
discrimination of more than 17 dB in the two planes. Fig. 10 7

depicts the measured gain variation with frequency showing a 6

35 35.5 36 36.5 37 37.5 38 38.5 39
good correlation and similar trend compared to the simulation Frequency (GHz)
result over the frequency 35-39 GHz. It shows an average Fig. 10. EM simulated and measured gain (dB) of the array antenna.
measured gain of 11.50 dBi and 12.0 dBi over the operating
frequency bands, 36.50-38.25 GHz, and 36.40-38.50 GHz, ACKNOWLEDGMENT
respectively. There is ~ 1.10 dB difference in gain attributed This work was supported by Natural Sciences and
to the fabrication tolerances and the measurement inaccuracies. Engineering Research Council of Canada (NSERC).

1536-1225 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/LAWP.2018.2880926, IEEE
Antennas and Wireless Propagation Letters

REFERENCES
[1] A. Petosa and A. Ittipiboon, “Dielectric Resonator Antennas: A
Historical Review and the current state of art,” IEEE Trans. Antennas
Propag., vol. 52, no.5, pp. 91-116, Oct. 2010.
[2] K. Gong, and Xue Hui Hu, “Low-Profile Substrate Integrated
Dielectric Resonator Antenna Implemented With PCB Process,” IEEE
Antennas and Wireless Propag. Lett., vol. 13, pp. 1023-1026, 2014.
[3] Y. Li, and K. Luk, “60-GHz Substrate Integrated Waveguide Fed
Cavity-Backed Aperture-Coupled Microstrip Patch Antenna Arrays,”
IEEE Trans. Antennas Propag., vol. 63, no. 3, pp. 1075-1085, March
2015.
[4] Q. Lai, C. Fumeaux, W. Hong, and R. Vahldieck, “60 GHz aperture-
coupled dielectric resonator antennas fed by a half-mode substrate
integrated waveguide,” IEEE Trans. Antennas Propag., vol. 58, no. 6,
pp. 1856-1864, 2010.
[5] W. M. Abdel Wahab, D. Busuioc and S. Safavi-Naeini, "Low Cost
Planar Waveguide Technology-Based Dielectric Resonator Antenna
(DRA) for Millimeter-Wave Applications: Analysis, Design, and
Fabrication," IEEE Trans. on Antennas and Propagation, vol. 58, no. 8,
pp. 2499-2507, Aug. 2010.
[6] W. M. Abdel-Wahab, S. S. Naeini, and D. Busuoic, “Millimeter-wave
high radiation efficiency planar waveguide series-fed dielectric
resonator antenna (DRA) array: analysis, design, and measurements,”
IEEE Trans. Antennas Propagation., vol. 59, no. 8, pp. 2834-2843,
2011.
[7] R. N. Simons, and R. Q. Lee, “Effect of parasitic dielectric resonators
on CPW/aperture-coupled dielectric resonator antennas,” Proc. IEE
Microw., Antennas, Propag.,vol. 140, no. 5, pp 336-338, 1993.
[8] Z. Fan, Y.M.M. Antar, A. Ittipiboon and A. Petosa, “Parasitic coplanar
three-element dielectric resonator antenna subarray,” Electronics Lett.,
vol. 32, no. 9, pp. 789-790, 1996.
[9] M. R. Nikkhah, J. R. Mohassel, and A. A. Kishk, “High-gain aperture
coupled rectangular dielectric resonator antenna array using parasitic
elements,” IEEE Trans. Antennas Propag.,vol. 61, no. 7, pp. 3905-
3908, 2013.
[10] M. Ranjbar Nikkhah, J. Rashed-Mohassel and A. A. Kishk, "Compact
Low-Cost Phased Array of Dielectric Resonator Antenna Using
Parasitic Elements and Capacitor Loading," IEEE Transactions on
Antennas and Propagation, vol. 61, no. 4, pp. 2318-2321, April 2013.
[11] Movahedinia, M. R. Chaharmir, A. R. Sebak, M. Ranjbar Nikkhah and
A. Kishk, "Realization of Large Dielectric resonator Antenna ESPAR,"
IEEE Transactions on Antennas and Propagation, vol. 65, no. 7, pp.
3744-3749, July 2017.
[12] W. M. Abdel-Wahab and S. Safavi-Naeini, "Improvement of aperture
coupling in SIW-fed DRA using embedded metallic posts," 2012 IEEE
International Symposium on Antennas and Propagation, Chicago, IL,
2012, pp. 1-2.
[13] W. M. Abdel-Wahab, Y. Wang and S. Safavi-Naeini, "SIW Hybrid
Feeding Network-Integrated 2-D DRA Array: Simulations and
Experiments," IEEE Antennas and Wireless Propagation Letters, vol.
15, pp. 548-551, 2016.

1536-1225 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.