ICSSC

7-10 September 2015, Gold Coast, QLD, Australia

33rd AIAA International Communications Satellite Systems Conference AIAA 2015-4331

A Ring-Focus Antenna Design for Simultaneous X and Ka

Band with Monopulse Tracking on Both Bands

Chris Leat1 and John Ness2

EM Solutions Pty Ltd, 55 Curzon St, Tennyson, QLD, 4105

Monopulse tracking offers excellent tracking performance and acquisition speed for
Communications On The Move (COTM) satellite terminals. Furthermore, it is clear that the
use of a single antenna for multiple bands is desirable for the terminal user, providing
increased data rates and improved reliability without needing additional space and with the
minimum increase of weight. We describe the design and performance of what we believe to
be the first fully monopulse, simultaneous X and Ka band 1 metre diameter antenna. Our
feed design uses a tapered polystyrene (Rexolite) rod for the Ka band within a circular
waveguide for the X band. Ka band TE 11 and TE21 signals are fed via circular metallic Ka
band multi-mode waveguide into the dielectric rod through a tapered adapter. A mode
converter is included within the Rexolite-filled waveguide run to improve E- and H-plane
symmetry. The X band waveguide has a TE 11 section for coupling of the data and a TE 21
section of increased diameter for coupling of the X band monopulse information using an
array of electric monopole probes fed via a suspended substrate network. Our reflectors use
a conventional, shaped Axially Displaced Elliptical design with a rimmed 1 metre main
reflector. Using a FEKO hybrid modelling approach in which the feed was modelled using
MoM and the resulting aperture fields used to illuminate the reflectors, aperture efficiencies
of 58% at 7.6GHz, 56% at 20GHz and 69% at 30GHz are predicted for the 1 metre
monopulse antenna while WGS antenna sidelobe and polarisation specifications are still met.

Nomenclature
MoM = Method of Moments
MLFMM = Multi Level Fast Multipole Method
LEPO = Large Element Physical Optics
COTM = (satcom) Communications On The Move
ADE = Axially Displaced Elliptical (reflector antenna geometry)
Rexolite = a proprietary polystyrene dielectric material of low loss

I. Introduction

The reflector components of reflector antennas possess very wide intrinsic bandwidths. However, the same can’t
be said for feed antennas: For example, a single, corrugated feed horn does not have the required bandwidth to cover
both X and Ka bands which, extending from around 7.25GHz to around 31GHz, represent a two octave bandwidth.
Typically, an approximate maximum bandwidth of half an octave may be achieved with a corrugated horn, due
mainly to the need to have the quarter-wave corrugations near the mouth appear an approximate open circuit while
those corrugations near the throat appear an approximate short circuit. Thus, a more complex design is required to
cover both X and Ka band. For small COTM antennas, dichroic surfaces are probably not practical either, given the
small electrical size of subreflectors, the small number of resonant elements therefore in a dichroic surface, and also
the limited space in which feeds can be located without causing feed blockage. Any 1 metre design must also meet
sidelobe requirements at Ka band, further constraining design options. Polarisation purity is also a strict
requirement: If dual circular polarisation handedness is required, then frequency independent designs like spirals
become inappropriate.
1
Antenna Designer, EM Solutions Pty Ltd, chris.leat@emsolutions.com.au
2
Chief Technology Officer, EM Solutions Pty Ltd, john.ness@emsolutions.com.au
1
American Institute of Aeronautics and Astronautics

Copyright © 2015 by EM Solutions Pty Ltd. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission.
ICSSC
7-10 September 2015, Gold Coast, QLD, Australia
33rd AIAA International Communications Satellite Systems Conference AIAA 2015-4331

Another dual band design option is

a coaxial feed. In such designs, the
higher frequency band is conveyed
within the inner waveguide and the
lower frequency band is conveyed
between the inner and outer
waveguides. The inner waveguide
may be either a dielectric rod or a
metallic waveguide.

However, monopulse antennas

introduce an additional layer of
requirements: If monopulse is to be
achieved for the higher band using
metallic waveguide, then either an
increased diameter is required to
support the propagation of higher
modes, or four complete waveguides
are required. Both options increase
the diameter of the inner waveguide
section, making the reflection Figure 1. Modeled E and H plane symmetry of a dielectric rod radiator
matching and pattern of the annulus- with and without the corrugated mode-converter section. Using a mode-
shaped aperture available for theconverter, the E/H plane asymmetry is kept below 0.4dB out to 35 degrees
lower band more challenging. from axis, whereas without it, the asymmetry is 1.4dB.
Furthermore, the increased size of the
internal waveguide may increase its outer circumference sufficiently that it may approach resonance in the lower
frequency band. The intrusion of the inner waveguide into the outer annulus is made worse if corrugations are
introduced inside the inner waveguide.

Whilst radius variations, like steps, notches and irises can be used to improve the matching in the annulus, it is
likely to be challenging to simultaneously match the monopulse information in higher modes, for example TE 21 or
TM01. We describe the design and modelling of a 1m COTM monopulse antenna with the following criteria:
Table 1 Specifications of the COTM antenna
Receive frequencies 7.25GHz-7.75GHz, 19.2GHz-21.2GHz
Transmit frequencies 7.9GHz-8.4GHz, 29GHz-31GHz
Polarization Simultaneous RHCP Tx and LHCP Rx or
configurations LHCP Tx and RHCP Rx (software selectable)
Sidelobe masks Meets MIL-STD-188-164B at Ka band
G/T X band : 16dBi/K, Ka band: 21dbi/K
Polarization purity AR < 1dB on axis.
Beam alignment < 0.50

II. Feed Design and Modelling

The challenges described above become less severe if a dielectric rod radiator is used for the inner waveguide:
Firstly, the beamwidth and phase centre for the dielectric rod is a function of its length and rate of taper, so that an
extra two degrees of design freedom are obtained; Secondly, though the dielectric rod partially supports the lower
(X) band propagation, its progressive taper to zero diameter causes much less reflection of the X band wave at the
feed mouth than does a metallic waveguide and it’s small diameter at the throat of the X band horn causes negligible
effects on the pattern; Thirdly, mode-converting corrugations for the higher (Ka) band can be moved out of the way
of the X band waveguide annulus and horn.

2
American Institute of Aeronautics and Astronautics

Copyright © 2015 by EM Solutions Pty Ltd. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission.
ICSSC
7-10 September 2015, Gold Coast, QLD, Australia
33rd AIAA International Communications Satellite Systems Conference AIAA 2015-4331

It is best if the dielectric

rod is considered first in the
design. A number of
constraints become
immediately apparent for the
rod: It was chosen to be a
body of revolution for
manufacturing reasons and
also in order to make the
polarisation specifications
more readily achievable.
Furthermore, when the
dielectric rod is in metallic
waveguide, it must allow
propagation of the TE21 modes
as well as the TE11 modes. The
rod design consists of 4
sections: 1) A cylindrical rod
section surrounded by
conducting surface where the
fields are guided into the feed; Figure 2. The metallic housing for the dielectric rod, shown in cross-section. The
2) A mode converter section to rod itself is omitted for clarity. On the left are tapers to transition the Ka band
convert the TE11 modes inwaves from air-filled waveguide to dielectric-filled waveguide. To the right of
section 1 into a hybrid mode this are seen a succession of mode-converting corrugations. Two rows of X band
for improved E- and H-plane probes are seen inside the larger waveguide, one for TE 11 and one for TE21. On
symmetry; 3) A constantthe right, the X band axial horn is seen.
radius cylindrical rod section
where the rod is surrounded by an annulus of free space between it and the X band waveguide wall in order to adjust
the position of the Ka band phase centre relative to the X band horn later included; and 4) A conical tapering section
from which the radiating fields will be launched, as the diameter transitions from one in which the Poynting flux is
almost exclusively conveyed inside the rod to a diameter in which the great majority of Poynting flux occurs in the
free space surrounding the rod. The benefits of the mode-converter section in improving the E- and H-plane
symmetry are shown in Fig. 1.

A dielectric rod of this kind was then co-axially incorporated into a conducting X band circular waveguide and
horn assembly as shown in Fig. 2. Waveguide tapers are used to reduce the air-filled Ka band waveguide dimensions
to the scaled dielectric-filled (Rexolite) waveguide, whilst the rod itself is tapered up in diameter to finally fill its
waveguide (section 1 of the rod as described earlier). The dielectric-filled Ka band waveguide then has a corrugated
mode-converter section (section 2 of the rod) before the dielectric rod emerges (section 3 and 4 of the rod) into a
larger circular waveguide suitable for TE11 propagation at X band frequencies, where 4 monopole probes introduce
the data channels at X band. The four probes are phased to transmit and receive X band right and left hand circular
polarizations for transmit and receive simultaneously. This waveguide then tapers to a larger diameter in which 8
monopole probes extract the TE21 X band signal. These probes are suitably phased to receive either circular
polarization of the TE21 signal. Finally, the X band waves are radiated from an axial corrugated horn. The section 3
of the dielectric rod (where its diameter is constant) was then adjusted to cause the Ka band phase center to coincide
with the X band phase center. A suspended substrate network (not shown) surrounds the X band waveguide and is
used to couple and phase the probes to provide suitable polarization and modal isolation. The modeled, linear-
polarized, TE11, electric near-fields of the feed are shown at Ka band and X band in Figs. 3 and 4. Aspects of this
feed form part of a patent application1.

Monopulse operation results from the ability of the feed to detect the lateral displacement of the Airy disc at the
focal point. Although there are several alternatives, in this design, TE 21 mode excitation is used to convey this
information deeper within the feed waveguides. In the case of the X band, it is extracted by the 8 monopole probes
around the X band waveguide. In the case of the Ka band, the TE 21 energy enters the dielectric rod, passes back
through the mode-converter and it coupled off from the multimode waveguide.

3
American Institute of Aeronautics and Astronautics

Copyright © 2015 by EM Solutions Pty Ltd. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission.
ICSSC
7-10 September 2015, Gold Coast, QLD, Australia
33rd AIAA International Communications Satellite Systems Conference AIAA 2015-4331

Figure 3. FEKO model of the feed showing near electric fields propagating inside the rod before being
radiated from near the tip of the dielectric rod at 30GHz.

Figure 4. FEKO model of the feed showing the near electric fields when the feed X band TE11 probes are fed
in linear polarization at 7.6GHz.

III. Reflector Design and Modeling

We selected the Axially Displaced Elliptical (ADE) dual reflector design as the basis of a shaped reflector
antenna. The ADE is desirable for small monopulse satellite antennas, since it can accommodate a wide beam-width
feed pattern due to the absence of feed blockage. This is advantageous since, typically, the feed is designed with the
data-carrying, sidelobe- and G/T-determining TE 11 modes as the highest priority whereas the patterns of the
pointing-error-sensing TE21 or TM01 modes are of lower priority and therefore have less controlled and usually wide
beam widths. Thus, in order to capture as much as possible of the higher mode feed pattern, a sub-reflector
subtending a wide angle is desirable. In a conventional Cassegrain design, this could cause unacceptable feed
blockage of the rays traveling between the sub-reflector and main-reflector, or unacceptable aperture blockage.

4
American Institute of Aeronautics and Astronautics

Copyright © 2015 by EM Solutions Pty Ltd. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission.
ICSSC
7-10 September 2015, Gold Coast, QLD, Australia
33rd AIAA International Communications Satellite Systems Conference AIAA 2015-4331

A major disadvantage of an ADE reflector pair, is that typically the aperture field strengths near the main reflector
center are down by around 10dB, due to their derivation from the weakest fields of the feed (originating from the
edges of the feed pattern) due to the ray inversion associated with the real image produced by the sub-reflector. This
causes a weakness in the aperture fields near the
optical axis which creates high first side-lobes.
Correcting this is a trade-off with exact ADE
reflectors, since in order to reduce the sub-reflector
spill-over and scattering side-lobes, very low fields
of the feed must be intercepted by the sub-reflector
edge (which at around -10dBi, are 25dB below the
feed maximum) and directed to the main reflector
central region, thus exaggerating the inner side-
lobes due to the low near-axial aperture fields.
Conversely, improving the first sidelobes will
increase the sub-reflector spill-over sidelobes.
Although this variation of field strength at the
aperture center is partly compensated by the ADE
geometry (since large angles of the feed field are
mapped to small radii of the main reflector
aperture), it is still problematic and leads to high
side-lobes. The sub-reflector curvature need only
deviate very slightly from the basic ellipse to
dramatically alter the field distribution at the main
reflector: For example a little less concave curvature
at the sub-reflector outer edge will bring the rays
closer together at the main reflector. This is the
Figure 5. Rays traced from the feed phase center after a
motivation for reflector shaping.
1 metre ADE reflector pair has been shaped to improve
aperture field uniformity for improved sidelobe levels.
A simple iterative Matlab script was written to
shape the reflectors with the aim of reducing side-lobes and improving bore-sight gain. This script assumes a feed
pattern of , which is a good fit to the feed pattern of the feed at Ka band. It starts with an ADE reflector
pair cross-sectioned in 2D, and maps the field strength across the main reflector aperture by discretising it into
around 40 rings. Energy conservation is used to equate the power in a spherical angle between adjacent ray angles
from the feed to the power in the corresponding annulus at the main reflector. The mapping is achieved using ray
tracing. Then, the shape of the sub-reflector is slightly adjusted to redistribute the power contained between the rays
over the main reflector to achieve a desired field distribution, including tapering. Now that the sub-reflector is no
longer an exact ellipse, the main reflector must be altered in shape slightly to restore the total path lengths of all the
rays to uniformity. This is done iteratively using a polynomial perturbation function added to the axially-displaced
parabola of the ADE design. A shaped reflector pair with a 150mm sub-reflector diameter and 1m main reflector
diameter is shown diagrammatically in Fig. 5 together with the ray paths from the feed phase center to the aperture.
Some bunching of rays is seen towards the inner and outer parts of the main reflector; This represents the extent of
aperture field upward adjustment resulting from the shaping aimed at improving the field uniformity across the
aperture and gradually tapering it at the outer edge. Also apparent, is the blurring of the intermediate ring focus into
a focal region.

IV Reflector Antenna 3D Modeling

The reflectors were modeled using Multilevel Fast Multipole Method (MLFMM) on the sub-reflector and Large
Element Physical Optics (LEPO) on the main reflector using the commercial EM modeling software FEKO. The
feed illumination was introduced to the reflector model via an aperture field computed by the MoM modeling of the
feed in isolation. After some iterations, it was noted that a rim added to the main reflector was required to reduce
main reflector spill-over lobes below the required masks.

5
American Institute of Aeronautics and Astronautics

Copyright © 2015 by EM Solutions Pty Ltd. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission.
ICSSC
7-10 September 2015, Gold Coast, QLD, Australia
33rd AIAA International Communications Satellite Systems Conference AIAA 2015-4331

The TE11 gain patterns are shown for the antenna in X band (7.6GHz) Ka band receive (20GHz) and Ka band
transmit (30GHz) in Fig. 6. The on-axis gains achieved correspond to aperture efficiencies of 58%, 56% and 69%
respectively.

Figure 6. The modeled far-field H-plane patterns of the 1metre, shaped, ADE, antenna fed with the
feed as described, shown at 7.6GHz, 20GHz and 30GHz. The inset image shows the reflector pair and
the square region over which the sampled pre-computed feed field is reconstituted as Huygen's
sources to illuminate the sub-reflector, instead of modeling the feed and reflectors is a single attempt.

Acknowledgments
This work was supported by the Australian Defence Department through a Concept Technology Demonstrator
grant. The authors would like to thank the Department and in particular the members of the Department who have
assisted with the conduct of the CTD for their insights and assistance. They would also like to thank the members of
EM Solutions who have assisted with mechanical design drafting, system design and discussion and project
management.

References
1
Leat,Christopher., EM Solutions Pty Ltd., Patent Application Nos. 2015901623 and 2015901709 “Improved Dielectric Rod
Antenna”, filed 11th and 15th May, 2015 with IP Australia.

6
American Institute of Aeronautics and Astronautics

Copyright © 2015 by EM Solutions Pty Ltd. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission.