XXX-X-XXXX-XXXX-X/XX/$XX.00 ©20XX IEEE
A CubeSat Train for Radar Sounding and Imaging of
Antarctic Ice Sheet
Prasad Gogineni(1), Christopher R. Simpson(2), Jie-BangYan(1), Charles R. O’Neill(2), Rohan Sood(2), Sevgi Z. Gurbuz(1), Ali C. Gurbuz(1)
(1)Department of Electrical and Computer Engineering, (2) Department of Aerospace Engineering and Mechanics
The University of Alabama Tuscaloosa, AL 35401, USA
Abstract—In spite of more than 50 years of airborne radar
soundings of Antarctic ice by the international community, there are
still large gaps in ice thickness data. We propose a CubeSat satellite
mission for complete sounding and imaging of Antarctica with 50
CubeSats integrated with a VHF radar system to sound the ice and
image the ice-bed. One of the major challenges in orbital sounding of
ice is off-vertical surface clutter that masks weak ice-bed echoes. We
must obtain fine resolution both in the along track and cross track
directions to reduce surface clutter. We can obtain fine resolution in
the along track direction by synthesizing a large aperture by taking
advantage of the forward motion of a satellite. However, we need a
large antenna-array to obtain fine resolution in the cross track
direction. We propose a train of 50 CubeSats with optimized offset
position to obtain a 500-m long aperture and also coherently combine
data from multiple passes of the train to obtain a very large aperture
of 1-2 km in the cross track direction. Our initial analysis shows that
we can obtain measurements with horizontal resolution of about 200
m and vertical resolution of about 20 m. The CubeSat will carry a
transmitter and receiver with peak transmit power of about 50 W. We
will synchronize all transmitters and receivers with a Ka-band system
that serves as a communication link between the earth and Cubesats
to downlink data and as command and control for the CubeSats.
Keywords—CubeSat, Satellite formation, cryosphere, sounding,
Extensive satellite and airborne observations of large ice
sheets in Greenland and Antarctic are showing both ice sheets
are retreating and their contribution to sea level has been
increasing over the last decade [1-2]. IPCC reported that sea
level would increase between 26 and 96 cm over the next
century under different warming scenarios . Rahmstorf 
and Jevrejeva et al.,  used semi-empirical models and
paleoclimate records to generate sea level rise estimates and
reported that it could be as large as 2 m. Even modest sea level
rise in heavily populated regions is a major problem
particularly during storm surges as illustrated during recent
super-storms Harvey and Irma. We need a more accurate
estimate of sea level rise projections to develop coastal
projections in a warming climate. In the next 3-4 years NASA
will launch two satellite missions, ICESat-2 and NI-SAR to
monitor polar regions. ICESat-2 will measure changes in
surface elevations and determine the current mass balance of
each ice sheet. The changes in surface elevation provide
information to document mass loss or gain. One of the major
objectives of the NI-SAR mission is to map surface velocities
of glaciers to provide information on the speed-up or slow-
down of outlet glaciers. Additional information on ice-bed
topography and basal conditions is required to develop ice-
sheet models to generate more accurate estimates of sea level
rise in a warming climate. Bed topography controls ice flow
and basal conditions determine how fast ice flows.
Large gaps in ice thickness measurements still exist over
Antarctica despite many years of airborne measurements.
Fretwell et al.  generated a new bed map by compiling all
data collected over Antarctica. In some areas, data points are
separated by as much as 100-200 km. Also, ice thickness
errors of up to a few hundred meters, greater than ten percent,
are common in areas with significant topographic changes.
Figure 1 shows a comparison of data collected by the Center
for Remote Sensing of Ice Sheets (CReSIS) as a part of the
NASA Operation IceBridge Mission (OIB) after the bed map
was generated. The results clearly show that interpolation used
to generate the new bedmap with very sparse sampling results
in poor representation of peaks and troughs in bed topography.
A satellite mission is the only way to effectively obtain
complete ice thickness data for Antarctica. Bed topography
and basal conditions generated from a satellite radar
sounder/imager data would substantially enhance the scientific
value of ICESat-2 and NI-SAR.
Airborne radars operating at frequencies between 60 and
300 MHz have been very effective in sounding and imaging
ice [8-11]. A formation of 50 CubeSats with a radar sounder
operating at 150 MHz create a large synthetic aperture to
obtain spatial resolutions of about 200 m both in along and
cross track directions. The CubeSat formation will synthesize
a large cross track aperture to reduce surface clutter that can
mask bed returns and obtain the high sensitivity needed to
Figure 1: Comparison of interpolated bedmap with
measurements in areas with significant topography .
4138978-1-5386-7150-4/18/$31.00 ©2018 IEEE IGARSS 2018
overcome high ice attenuation and spreading losses. A Ka-
band system for each CubeSat will function as radar for
precise measurement of distance and velocity between the
CubeSats for accurate orbit determination, as a
communication link to synchronize the distributed transmitter
array, and as a down-link to transmit sounding and imaging
data to a ground station.
II. ORBITAL RADAR SOUNDING AND SURFACE CLUTTER
Orbital radar must have high sensitivity to overcome large
ice attenuation and spreading loss of operation at an orbit of
400 km or higher and minimize off-vertical surface clutter that
masks weak ice-bed returns. Radar must have a large power-
aperture product to provide the necessary sensitivity and spatial
resolution. A peak transmit power a few kilowatts or higher is
needed to combat both the spreading and ice loss to sound the
ice-bed with a reasonable signal-to-noise ratio (SNR). A large
antenna aperture is also essential to provide a narrow
beamwidth for fine spatial resolution and suppressing surface
clutter. Signals scattered by the rough ice surface, referred to as
surface clutter, can mask weak echoes from the ice-bed
interface. Surface clutter is a major issue in high-altitude and
orbital sub-surface sounding because the ice surface is
illuminated at small incidence angles. While SAR processing
can be used to reduce the along track antenna beamwidth to
filter out surface clutter coming along the flight direction,
advanced array processing techniques must be used to reduce
clutter in the cross track direction. An orbital radar operating at
an altitude of 400 km; the surface scattering from incidence
angles of 7.5o and 12o can mask returns from ice-beds covered
with 2 and 5 km thick ice, respectively. If the two-way
attenuation rate is assumed to be 15 dB/km, clutter must be
reduced by ~60 dB to obtain discernable echoes from the ice-
bed interface covered with 5 km of ice. A link budget for the
radar system in given in Table 1.
Table 1: Link Budget
Antenna/Array Processing Gain
Wavelength Effect, (λ = 2cm)
Two way power transmission coeff.
Loss for 5 km thick ice (15 dB/km)
Pulse compression gain, CI
Total Spreading loss term
Noise power (Bandwidth = 10MHz)
Reflection coefficient at the bottom
Li et al.  developed coherent and incoherent clutter
reduction, as well as a technique to reduce cross track clutter,
and applied the coherent clutter reduction technique to data
collected with a radar on the NASA DC-8 aircraft from a
height of about 10,000 m above the surface . Figure 2
shows SAR- and array-processed results. The left echograms
show results after SAR processing the data combined with a
sum-delay beamformer with Hanning weights. The ice-bed
echoes are masked by the cross track surface clutter. The right
shows the results from a Minimum Variance Distortionless
Response (MVDR) beamformer; the bed echoes are clearly
discernable. To adapt the technique to a spaceborne ice
sounder we estimate that a cross track antenna array of 40-50
m long is required. A 50 m long antenna aperture at 150 MHz
provides an antenna beamwidth of about 4o after weighting,
which is required to reduce antenna-array sidelobes to
suppress clutter. With advanced signal processing using
multiple passes, we can obtain ice thickness measurements
with a vertical resolution of 20 m and horizontal resolution of
200 m to satisfy most of the major glaciology and geology
III. SATELLITE FORMATION
50 CubeSats spaced 50 m in the along track and 1 m
spacing in the cross track direction will synthesize a large cross
track array using the Earth’s rotation . The cross track array
sample spacing is determined by the offset distance between
adjacent CubeSats for a single pass. We can combine multiple
passes to synthesize a very large aperture to obtain fine spatial
resolution in the cross track direction.
Creating the synthetic aperture requires coordinated
autonomous operation using the Ka-band device for phase-
locked operation of the VHF radar. A highly inclined orbit
using the argument of perigee to create phasing was chosen.
The five satellites in Figure 3 are an example of the PICS
(Polar Inclined CubeSats) formation.
Over the Antarctic coordinated autonomous operation of
the VHF radar is required. Each CubeSat carrying a 150 MHz
radar can be considered a single transmit and receive center.
Operation of the VHF radar will be phase-locked by using the
Ka-band device. The device will provide the range and range-
rate of each spacecraft’s neighbors. GPS differenced
measurements will provide additional positional and velocity
knowledge. SAR formed by multiple small spacecraft has not
been operationally achieved . The best formation flying
spacecraft mission to-date is the CanX-4&5 mission;
successfully demonstrating autonomous formation flight with
sub-meter control error and centimeter-level relative position
knowledge with the closest controlled range of 50 m during
formation flight . For perspective, the closest controlled
range a CubeSat in PICS will experience is 49 m. The 1 m
cross track distance will drive the propulsion choices and the
return from the Ka-band prototype will drive the ADCS
A. PICS Formation and Operations
PICS varies the argument of perigee to separate the
CubeSats in the along track direction. The inclination of each
orbit is varied slightly to achieve the 1 m cross track separation
Figure 2. Bed returns recovered after surface clutter suppression
. Top and bottom echograms are before and after array
over Antarctica. The phasing distance, relative position
information from the Ka-band device, and absolute position
information from differenced GPS when equipped with a cold-
gas thruster will provide sufficient capability to achieve the
desired offset position and maintain a safe separation. Each
CubeSat will have an identifier to distinguish it from its
compatriots as it changes position in the formation.
Each CubeSat will consist of a radar transmitter and
receiver operating in the Ka band to provide local navigation
and communication within the formation. The transmitter on
each CubeSat will generate a complementary coded chirp
digitally. The chirp will be amplified to the required power
level with a driver and power amplifier chain and coupled to
the antenna through a transmit/receive (T/R) switch. The
coding will allow us to isolate and process signals from each
CubeSat before combining signals from all CubeSats to
synthesize large apertures both in the along track and cross
track directions. Table 1 shows our preliminary link budget for
the radar. We can sound close to 5-km thick low-loss ice in
Antarctica with our CubeSats train.
Data quality critically depends on the antenna aperture
synthesized by the PICS formation. The Ka-band device will
provide precise ranging, velocity and high-speed downlink
capabilities for each CubeSat. We have chosen a FMCW radar
architecture for ranging and relative velocity measurements,
similar to commercially available automotive radars. The
proposed FMCW radar not only has the common advantage of
low peak power requirement, but the continuous-wave signal
can also be used to synchronize the local oscillators used in the
ice sounders on different CubeSats.
It is possible to optimize the along track separation by
varying the eccentricity vector to change the along track
distance as the formation passes over a targeted latitude using
the cold-gas thruster in a nadir orientation. Cross track
separation can be varied through inclination changes or taking
advantage of natural nodal regression. PICS, taking advantage
of the precession of the line of nodes, can vary the cross track
distance in less than 7 days. Orbital altitude will be maintained
by periodic drag make up maneuvers. A cold-gas thruster will
be sufficient for small attitude and up-keep maneuvers to fine-
tune the formation. Other formations have taken advantage of
specially designed orbits to maintain formation position .
Using differenced GPS, IMUs, and the Ka-band device we can
expect centimeter-level relative positional knowledge and
millimeter-level precision. The sub-meter control error will
need to be addressed in order to maintain 1 m cross track
B. Synthetic Aperture Acquisition
Our high resolution polar mapping leverages both the
natural synthetic aperture characteristics of the radar array
formation and the variation in orbital revisits. Figure 4
conceptually illustrates the cross track synthetic aperture width
for a single orbit and the track separation for subsequent
orbits. The cross track synthetic aperture width is length of the
total formation ground track. The track separation (spatial
resolution) takes advantage of the Earth’s natural rotation and
precession of the orbital plane, which creates a natural
variation in inbound and outbound orbit angles (i.e. apparent
crosshatching in the tracks in Figure 4). Element track
separation is achieved by the PICS formation cross track
separation. The proposed PICS formation design is driven by
the need to form a synthesized aperture with minimum cross
track grating lobe levels. The element track separation is 1 m
(seen in Figure 3 and 4). For the 50 CubeSat formation, the
synthetic aperture width will be about 500 m.
Our mission altitude, 400 km, gives a 92.6±1.67-5 min
orbit in which the Earth rotates by approximately 23.2±0.06˚.
Revisit coverage depends on the target location and mission
duration. The pole receives multiple revisits early on because
of the highly inclined orbits (85˚≤ i ≤ 95˚). Other locations
will receive no early revisits because of their latitude or
inclination. By day ten of the mission, all ground targets have
seen at least one revisit. The variation in inbound and
outbound ground paths create iso-latitude rings of at least 2
passes over the same ground target at day ten. Figure 5
illustrates the revisit coverage at 400 days. Coverage is not
uniformly spaced with bands in specific radii and angles.
Flight correction/maneuvers on-orbit have the potential to
provide more detailed scanning in targeted areas.
C. Effect of Ka-band Device PICS Operations
The 35 GHz device serves three purposes: ranging radar,
downlink, and phase-lock for VHF sounding radar. When
functioning as a ranging radar the device will provide range
and range-rate information to each CubeSat. The on-board
autonomous command will determine orbit correction
maneuvers to maintain the synthetic aperture. The SNR will
Figure 3: Five CubeSats creating a synthetic aperture through
separation of 1 m in the cross track and 10 m in the along track.
Figure 4: Ground track separation of array
drive the Attitude Determination and Control System (ADCS)
selection. Maryland Aerospace currently provides ADCS with
a 0.007˚ pointing capability at a technology research level
(TRL) 6 [18, 19]. There is a negative correlation between the
SNR level and the resolution required of the ADCS; see an
example in Figure 5.
When operating as a downlink, the spacecraft will slew as
it passes over the target ground station. The CubeSats will be
unable to rely on the ranging radar capability at this point in the
flight. Position and velocity updates will be provided by
differenced GPS measurements and an on-board IMU. The
ADCS and the CubeSat bus will have to be able to slew at the
rate required to keep pointing the highly directional beam at the
ground station. The required slew rate is proportional to the
inverse of the half beam width, θ. At 400 km for an infinitely-
thin beam, the spacecraft would need to slew at ~0.567˚/sec if
it does not use electronic steering of the beam.
In the paper we presented a preliminary concept for a CubeSat
train-base radar sounder/imager to map Antarctica completely.
A formation of 50 CubeSats with a radar sounder operating at
150 MHz create a large synthetic aperture to obtain spatial
resolutions of about 200 m both in along and cross track
directions. Both 144 MHz and 225 MHz are approved for
radar use. The CubeSat formation will synthesize a large cross
track aperture to reduce surface clutter that can mask bed
returns and obtain the high sensitivity needed to overcome
high ice attenuation and spreading losses. A Ka-band system
for each CubeSat will relative positions, serve as a
communication link to synchronize the distributed transmitter
array, and as a down-link to transmit sounding and imaging
data to a ground station.
 R. Thomas et al., “Accelerating ice loss from the fastest
Greenland and Antarctic glaciers,” Geophysical
Research Letters, vol. 38, no. 10, pp. L10502, 2011.
 Shepherd et al., “A Reconciled Estimate of Ice-Sheet
Mass Balance,” Science, vol. 338, no. 6111, pp. 1183-
 IPCC, Climate Change 2013: The Physical Science
Basis. Contribution of Working Group I to the Fifth
Assessment Report of the Intergovernmental Panel on
Climate Change. Cambridge, United Kingdom and New
York, NY, USA: Cambridge University Press, 2013.
 S. Rahmstorf, “A semi-empirical approach to projecting
future sea-level rise,” Science, vol. 315, 2007.
 S. Jevrejeva et al., “How will sea level respond to
changes in natural and anthropogenic forcings by 2100?”
Geophys. Res. Lett., vol. 37, pp. L07703, 2010.
 T. Stumpf. Private communication about results from the
 P. Fretwell et al., “Bedmap2: improved ice bed, surface
and thickness datasets for Antarctica,” The Cryosphere,
vol. 7, 2013.
 Evans S; and Smith BME, “A radio echo equipment for
depth sounding in polar ice sheets,” Journal of Scientific
Instruments (Journal of Physics E), Ser. 2, v. 2, 1969.
 Drewry DJ and Meldrum DT; “Antarctic airborne radio
echo sounding, 1977-78,” The Polar Record, 19(120),
pp. 267-273, 1978.
 Gogineni S; et al., “An improved coherent radar depth
sounder,” Journal of Glaciology, 44(148), 1998..
 Peters ME; Blankenship DD and Morse DL; Analysis
techniques for coherent airborne radar sounding:
Application to West Antarctic ice streams, Journal of
Geophysical Research, B Solid Earth, vol. 110, no. B6,
 Li, J., et al., (2012), “Coherent and Incoherent clutter
Reduction Techniques to Sound Ice from High
Altitudes,” EUSAR 2012, Nuremberg, Germany, 2012.
 J. Li et al. “High-Altitude Radar Measurements of Ice
Thickness Over the Antarctic and Greenland Ice Sheets
as a Part of Operation IceBridge,” IEEE Trans. Geosci.
Remote Sens., vol. 50, no. 12, 2012.
 Raney, K. “WITTEX: An innovative multi-satellite radar
altimeter constellation,” Internal JHU Document SRO-
01-08, March 2001.
 S. Chung et al., “Review of Formation Flying and
Constellation Missions Using Nanosatellites,” J.
Spacecraft and Rockets, vol. 53, no. 3, 2016.
 H. N. Roth et al., “Flight Results From the CANX-4 and
CANX-5 Formation Flying Mission,” ESA and CNES:
The 4S Symposium, May 30 – June 3, 2016.
 Krieger, G., Hajnsek, I., Papthanassiou, K. P., Younis,
M. Y., and Moreira, A., “Interferometric Synthetic
Aperture Radar (SAR) Missions Employing Formation
Flying,” Proceedings of the IEEE, Vol. 98, No. 5, 2010.
 Agasid, E. et al., State of the Art of Small Spacecraft
Technology, NASA, March 2017. <https://sst-
 “NASA TRL Definitions,” NASA NPR 7123.1B, Oct
 Jenn, D., “Radar Fundamentals,” Naval Postgraduate
School, 2017 [presentation].
Figure 5: RCS affects SNR adversely requiring a higher
resolution from the ADCS, .