SSC15-IV-8

X-band CubeSat Communication System Demonstration

Serhat Altunc, Obadiah Kegege, Steve Bundick, Harry Shaw, Scott Schaire, George Bussey, Gary Crum, Jacob C.
Burke
NASA Goddard Space Flight Center (GSFC)
8800 Greenbelt Road, Greenbelt, MD 20771; Phone: (301) 286-2933
[email protected]

Scott Palo, Darren O’Conor

University of Colorado Boulder, College of Engineering and Applied Science
Boulder, CO 80309; Phone: (303) 492‐2658
[email protected]

ABSTRACT
Today’s CubeSats mostly operate their communications at UHF- and S-band frequencies. UHF band is presently
crowded, thus downlink communications are at lower data rates due to bandwidth limitations and are unreliable due
to interference. This research presents an end-to-end robust, innovative, compact, efficient and low cost S-band uplink
and X-band downlink CubeSat communication system demonstration between a balloon and a Near Earth Network
(NEN) ground system. Since communication systems serve as umbilical cords for space missions, demonstration of
this X-band communication system is critical for successfully supporting current and future CubeSat communication
needs.

This research has three main objectives. The first objective is to design, simulate, and test a CubeSat S- and X-band
communication system. Satellite Tool Kit (STK) dynamic link budget calculations and HFSS Simulations and
modeling results have been used to trade the merit of various designs for small satellite applications. S- and X-band
antennas have been tested in the compact antenna test range at Goddard Space Flight Center (GSFC) to gather radiation
pattern data. The second objective is simulate and test a CubeSat compatible X-band communication system at
12.5Mbps including S-band antennas, X-band antennas, Laboratory for Atmospheric and Space Physics (LASP)
/GSFC transmitter and an S-band receiver from TRL-5 to TRL-8 by the end of this effort. Different X-band
communication system components (antennas, diplexers, etc.) from GSFC, other NASA centers, universities, and
private companies have been investigated and traded, and a complete component list for the communication system
baseline has been developed by performing analytical and numerical analysis. This objective also includes running
simulations and performing trades between different X-band antenna systems to optimize communication system
performance. The final objective is to perform an end-to-end X-band CubeSat communication system demonstration
between a balloon and/or a sounding rocket and a Near Earth Network (NEN) ground system.

This paper presents CubeSat communication systems simulation results, analysis of X-band and S-band antennas and
RF front-end components, transceiver design, analysis and optimization of space-to-ground communication
performance, subsystem development, as well as the test results for an end-to-end X-band CubeSat communication
system demonstration. The outcome of this work will be used to pave the way for next generation NEN-compatible
X-band CubeSat communication systems to support higher data rates with more advanced modulation and forward
error correction (FEC) coding schemes, and to support and attract new science missions at lower cost. It also includes
an abbreviated concept of operations for CubeSat users to utilize the NEN, starting from first contact with NASA’s
communication network and continuing through on-orbit operations.

Altunc 1 29th Annual AIAA/USU

Conference on Small Satellites
INTRODUCTION performance for a system utilizing the LASP S- and
This paper discusses SmallSat/CubeSat S- and X-band X-band transceiver. The orbital/link parameters used
in this analysis are shown in Table 1. In order to
communication system simulations and analysis,
transceiver development, testing, and demonstration. achieve the desired science data downlink of 12.5
The analysis covers transceiver requirements, space- Mbps in X-band, the communication link will need a
Carrier to Noise (C/N0) power level of at least 83.4dB
to-ground communication system
specifications/performance, and systems with the standard 3dB link margin. Due to bandwidth
demonstration utilizing a balloon platform. The limitations, a 5Mbps rate is desired in S-band, which
will require a C/N0 power level of 79.4dB.
LASP S- and X-band transceiver development section
will address the functional components, design
Table 1: Space to Ground Link Parameters
specifications, waveform description/parameters, and
applications, as well as compliance with NASA’s Orbit LEO (sun-synchronous)
NEN waveform-specific performance requirements.
The transceiver system level test results are presented Altitude 705 km
addressing: S-Band receiver, S-Band transmitter and S-band Frequency 2310 MHz
X-Band receiver. Also, the antennas that will be used
with the LASP S- and X-band transceiver have been X-band Frequency 8210MHz
characterized. S-band Antenna Gain 0 dBi
The SmallSat/CubeSat S- and X-band communication X-band Antenna Gain 9 dBi
system is a very promising technology that will S- and X-band
increase science data downlink performance for future Transmit Power 1 watt
missions. Other aspects of this paper include radiation
tolerance, and technology infusion plan. Reed Solomon
Encoding 6.4 dB required Eb/No

CUBESAT RADIO REQUIREMENTS

The proposed CubeSat transceiver is designed to be Three of NASA’s Near Earth Network (NEN) ground
compatible with NASA’s NEN and comply with stations were baselined in this analysis: Wallops
waveform-specific performance requirements. High Ground Station (MGS) in Virginia, Alaska’s Station at
level requirements include: compatibility with a 6U Fairbanks (ASF), and the McMurdo Ground Station
CubeSat, operation for 12 months in LEO, Tx to (MGS) in Antarctica shown in Figure 1.
transmit up to 12.5 Mbps, Tx to support OQPSK STK’s Communications module provides the dynamic
modulation, Tx to support forward error correction link analysis (path loss, atmospheric effects, etc.). The
coding, Tx to have sufficient power to close the link STK link results were also verified by an internal link
between LEO and NEN, Rx capable of closing the link analysis. Table 2 gives a summary of the minimum
between LEO and NEN, and operating temperature data rates achievable at the three NEN sites. Figure S-
between -20C and +50C. 3 shows the dynamic C/N0 power levels provided by
the S- and X-band payloads on orbit to the multiple
ground stations.
SIMULATIONS AND ANALYSIS

Space –to-Ground System Analysis and Simulations

Systems Tool Kit (STK) simulations were performed
to evaluate space-to-ground communication

Altunc 2 29th Annual AIAA/USU

Conference on Small Satellites
 Figure 1: Space-to-Ground System Simulations

Table 2: Achievable data rates at three selected NEN sites

Wallops (WGS) Fairbanks (ASF) McMurdo (MGS)

Ground Station WGS 11.3M ASF 11M MGS (10m)
Frequency S-band X-band S-band X-band S-band X-band
Elevation Angle (deg) 5 10 5 10 5 10
Min Data Rate (Mbps)
4.3 42.3 4 93 2.6 33.5
(from 705km Alt.)
Contact Time Per Day (hrs) 0.71 0.494 1.674 1.136 2.253 1.546
Latency (hrs) Average 4.556 2.032 1.983 4.599 1.416 1.829
Maximum 11.843 10.032 8.374 11.879 1.6 6.094

Results in Table 2 were generated using the assumptions for the S- and X-band communications payload systems
given in Table 1.

Figure 2: SmallSat S- and X-band Payloads to NEN Stations

Altunc 3 29th Annual AIAA/USU

Conference on Small Satellites
For the S-band payload communications link to
Wallops Ground Station (WGS) 11.3m antenna at a 5
degree elevation angle, the S-band payload does not
close the 5 Mbps link with 3dB margin. There is about
0.6dB needed in the link. In order to close the link, the
S-band payload would need a slight increase in
transmit power or antenna gain or the payload would
be need to be operated solely above 7 degrees
elevation angles. For the link to McMurdo Ground
Station’s (MGS) 10m antenna at 5 degree elevation,
there is about 2.8dB needed to close the link with 3dB
margin. This would require a 2-watt power amplifier, Figure 3: System Demo Using Balloon Simulation
or operation of the 1-watt payload above 15 degree
elevations to close the 5 Mbps link with the standard
3dB link margin.

For the X-band payload, utilizing a 9dBi antenna gain,

there is 7.3dB margin on the link to the 10m antenna
at MGS at 10 degree elevation, and 8.3dB on the link
to WGS’s 11.3 antenna. Figure 2 shows that there is at
least 4 additional dB, over the required 3dB margin on
any of the X-band links to the NEN ground stations.
The links will also close with about 5dB margin if at a
5 degree of elevation with an ACS system.

For the S-band payload, the data rate of 5Mbps is not

achievable without some improvements over the
assumed payload performance. These could be Figure 4: WFF to Balloon Slant Range
achieved with additional antenna gain, transmitter
power, or with a more advanced forward error
correction (FEC) scheme. The S-band payload could
also be operated only at higher ground-to-space
elevation angles to achieve the desired 5Mbps the
NEN sites. The X-band payload as defined is sufficient
to close the 12.5 Mbps link to the multiple NEN
stations from a 705km orbit.

Balloon Simulation
A balloon was simulated in STK by assuming a
constant rise rate to a float altitude of 37km. A direct
eastward bound drift rate of 5km/hr was used. Wallops
Flight Facility (WFF) was chosen as the launch facility
for the balloon flight analysis (see Figure 3).
Figure 5: Balloon to WGS 11.3m Antenna
As can be seen in Figure 4, at 24 hours of simulation, Simulation
the balloon reaches a slant range of 580km from the For this balloon trajectory, the simulation shows
facility at Wallops Island. As can be seen in Figure 4, significant margin on the links. For both the S- and X-
the elevation at this time and distance is about 1 band payloads, the simulations show nearly 24 hours
degree. Dynamic link margins from the simulation are with achievable data rate goals of 5Mbps and
shown in Figure 5 for both the S-band and X-band 12.5Mbps, respectively. At 5 degree elevation, there is
links to the Wallops Ground Station (WGS) 11.3m approximately 21dB and 23dB margin on the S- and
antenna. X-band links, respectively. For the X-band link, the
3dB margin threshold on a 12.5Mbps link is only
crossed when the elevation angle decreases to about 1
degree, which is when the multipath/scintillation
effects have significantly degraded the link. In

Altunc 4 29th Annual AIAA/USU

Conference on Small Satellites
practice, communications would not be performed control signal feeding the NCO input. There is an
below 3-5 degree elevation. Thus, the additional finite impulse response loop filter, F3, to
SmallSat/Balloon payload radiated power remove unwanted frequency components that are not
performance assumed in this analysis will not be an fully attenuated by filters F1 and F2.
issue for the proposed S- and X-band communication
links from a balloon trajectory. Instead, the links will
likely be bandwidth limited when above very low
elevation angles (<2 degrees). The X-band payload
will provide more link margin than the S-band link for
elevation angles above 2.5 degrees. The S-band link is
more efficient through the atmosphere and at very low
elevation angles, but with the greater antenna gain, the
X-band payload would provide better performance
(more link margin) at 12.5 Mbps than the S-band
payload at 5Mbps for elevation angles above 2
degrees. Ultimately, both the S- and X-band payloads
are expected to provide adequate performance margin
on their respective 5Mbps and 12.5Mbps
communication links for a balloon test flight.

TRANSCEIVER DEVELOPMENT Figure 6: Block diagram of AD9364 and FPGA

The University of Colorado, Boulder, is continuing based digital Costas Loop for BPSK reception
advancement of a transceiver consisting of an X-Band As with the transmitter design process, we have
transmitter and S-Band receiver under a SmallSat implemented a model of the system using Simulink.
Technologoy Partnership award (NNX13AR01A). The model is shown in Figure 7. As of this time, the
Now in year two, the X-Band transmitter design has model has shown that the loop achieves lock at
achieved the goal of TRL5, and the team, including frequencies within 2% of the NCO start up frequency.
NASA GSFC engineers and management, continues to The loop then maintains lock as the carrier frequency
work on moving the receiver to this technology is swept on the transmitter side (not shown in Figure
readiness level. 7). The entire range of Doppler-shifted frequencies has
not been simulated; however, a drift of 1 MHz / second
The receiver design is based on the Analog Devices
showed lock over a 100 KHz capture range once lock
AD9364 integrated RF agile transceiver IC. As shown
was achieved. This drift rate far exceeds the
in Figure 6, the AD9364 baseband and down-shifted
approximately 83 Hz / second expected drift rate,
subcarrier will be fed to an FPGA. The digital circuits
showing that the Costas loop can maintain lock with a
in the FPGA implement a Costas loop, which
much more challenging frequency shift/drift rate than
accommodates the Doppler shift in frequencies
expected in flight by four orders of magnitude.
experienced by the receiver during a pass over a
Simulation times make verification of the expected,
ground station. The Costas loop consists of a
slower drift frequency rate too slow, but loop
numerically controlled oscillator (NCO), raised cosine
performance will only increase with the decreased
filters matched to that used in filtering the transmitted
frequency drift rates. The slower drift rate will be
uplink data (F1 and F2 in Figure 6), as well as
tested in the final hardware implementation modeled
multipliers to act as frequency and phase detectors,
in Figure 7.
and to combine the two filtered paths to generate a

Altunc 5 29th Annual AIAA/USU

Conference on Small Satellites
 Figure 7: Simulink diagram of Costas loop implemented for verification of performance
the transmitter output from the AD9364 to test the
receiver during the receiver characterization.
Figure 7 shows the Simulink diagram of Costas loop
implemented for verification of performance. Figure 8
shows the digital data (blue) sent to the BPSK modeled
transmitter in the Simulink system, and the received data
(green) resulting from the Costas loop BPSK
demodulation. While the effects of filtering through RCF
filters in the modulation and demodulation path have
eliminated higher frequency components in the received
signal, the received signal has enough information to
reproduce the original data time series.

The 12 layer board design is now nearly complete and

will go out for manufacture while the FPGA
implementation is completed. Also during the interim,
the configuration parameters for the AD9364 will be
determined from Analog Devices’ Linux based test
platform for the chip. While using such an integrated
device has simplified much of the RF development than
would have been required designing with lower level Figure 8: Transmitted (blue) and received (green)
components of a receiver, such as was done with the X- data from the Simulink model shown in Figure 7
Band transmitter, understanding the configuration of the
device and its performance remain not-insignificant
tasks with this device. Because the AD9364 also
contains a transmitter chain, future possibilities include
the design of an S-Band transmitter and receiver on a
single CubeSat form factor board. In fact, we plan to use

Altunc 6 29th Annual AIAA/USU

Conference on Small Satellites
 Figure 10 shows the I,Q FPGA outputs with the
repeating 127 bit, PRBS7 pattern. The convolution and
differential encoding were turned off for these
measurements to make it easier to visually see the pattern
which starts with 7 consecutive ones. This test
produced data at 12.5 Mbits/s that is split 50/50 on the I
and Q data paths.

Additional Bit Error Rate (BER) testing is planned using

a Cortex XXL receiver in the lab, and also with the
Wallops 11 meter antenna. Shown in Figure 11 is the
BER test block diagram.

X-band Noise Cortex

D/C XXL
Xmtr Source
Var Atten Rcvr

Test Inject

Figure 9: S-Band Receiver PCB Layout NEN 11m Antenna

Figure 11: Transmitter BER Test Setup

SYSTEM LEVEL TESTS AND RESULTS
Both the I and Q FPGA LVDS outputs were verified with
a pseudo random pattern defined by ANTENNA TEST AND CHARACTERIZATION
PRBS7 = . In this test mode, the Actel
Antenna pattern testing is being conducted at the
Proasic FPGA generates this pattern which is then sent Wallops dual mode far-field/compact antenna test range.
outside the PCB through a test port with LVDS. Antennas will be mounted on a Cubesat model for
pattern measurements. Figure 13 shows the 4 element
A loopback test was performed with cabling such that the
patch antenna array which was designed and built by Ant
output of this test port was put back into the PCB/FPGA Dev Corporation. It is compact, effective, low cost, and
input to represent user data. efficient with stable high gain characteristics.
Internally to the FPGA, the user data is split into separate
I,Q data paths then optionally differential and
convolutional encoded.

Figure 12: WFF Far-Field Compact pattern

measurement range

Figure 10: Receiver I and Q FPGA outputs

Altunc 7 29th Annual AIAA/USU

Conference on Small Satellites
 Figure 13: X-Band 4-Patch Array Antenna
(AntDevCo), fc 8080 MHz, 9 dB gain

Figure 14 and 15 presents X-Band 4-Patch Array

Antenna Gain Linear, Theta Cut (0 to 360 deg) at theta
phi = 0 degrees and theta phi = 90 degrees respectively.
Figure 16 presents X-Band 4-Patch Array Antenna Figure 15: X-Band 4-Patch Array Antenna Gain
VSWR. Plot - Etheta Linear, Theta Cut (0 to 360 deg) at
theta phi = 90 degrees. Plot Max 12 dBi with 2 dB
per division.

Figure 14: X-Band 4-Patch Array Antenna Gain Figure 16: X-Band 4-Patch Array Antenna VSWR
Plot - Etheta Linear, Theta Cut (0 to 360 deg) at Plot
theta phi = 0 degrees. Plot Max 12 dBi with 2 dB per
division.
TRANSCEIVER NEN COMPATIBILITY
The transceiver was designed to be compatible with the
NASA NEN frequency allocations in S- and X-band.

Altunc 8 29th Annual AIAA/USU

Conference on Small Satellites
The transceiver will be tested for compatibility with the Technology infusion plan
Wallops Ground Station (WGS) 11.3-meter antenna. The results of this demonstration will be shared with
NEN S-band command characteristics of the WGS
scientists throughout NASA and industry, for potential
include: NEN S-band uplink band, RHC or LHC use on their future missions. This enabling capability
polarization, options of PM, FM, BPSK or QPSK / will increase CubeSat/SmallSat development and test
OQPSK and FSK carrier modulations, and options of
capabilities. Also, this S- and X-band communication
NRZ-L, and M, or S; or Bif-L, M, or S data formats. X- system will provide a test bed for new technology
band telemetry characteristics of the WGS include: demonstration missions that will enable new mission
NASA NEN X-band downlink, RHC or LHC
classes or reduce the cost, schedules and risk of current
polarization, options of QPSK, UQPSK, SQPSK, NASA’s mission design methodologies. This system
AQPSK modulations, NRZ-L or M data formats, and serves as a viable communications framework that can
ability to support data rates of 6 – 150 Mbps. be leveraged as part of a small technology demonstration
Also, the transceiver’s baseband data format will be satellite.
compatible with the NEN. WGS is compatible with
baseband data in any of the following formats: IP, serial
clock and data, and 4800-bit blocks encapsulated in IP CONCLUSION
packets.
This S- and X-band communication system will be NEN
compatible and will be used as a potential standard
baseline for CubeSats throughout NASA. The outcome
FUTURE WORK AND TECHNOLOGY INFUSION of this development and demonstration will pave the way
PLAN of a next generation NEN compatible X-band CubeSat
communication system which supports higher data rates
This S-band uplink and X-band downlink CubeSat
communication system development and demonstration with more advanced modulation and forward error
will provide a framework to build and launch a small correction (FEC) coding schemes as well as to support
and attract new science missions at lower cost.
technology demonstration satellite in response to the
NASA's HEOMD CubeSat Launch Initiative. Also, X-
Acknowledgments
band will be more attractive for high data rate CubeSat
science missions. In the near term, CubeSats are more The authors would like to give acknowledgments to
likely to move into X-band for higher data rates, and will those who provided support and data that was compared
utilize higher order modulation and encoding schemes with some of the measurements in the paper.
that will enhance link quality and overall system
capability. References
1. S. Palo, D. O’Connor, E. DeVito, R. Kohnert, G.
Crum and S. Altunc, “Expanding CubeSat
Radiation Tolerance Capabilities with a Low Cost Transceiver,” 28th
The transceiver uses an Actel Proasic FPGA. The Annual AIAA/USU Conference on Small
components in the design were carefully selected to Satellites, Logan, UT, Aug. 2014
operate in LEO or GEO radiation environments. 2. The Future of CubeSats:
http://www.nasa.gov/content/goddard/the-future-
Radiation tests can be performed on our current S-band of-cubesats/
and X-band hardware. Radiation testing will help to
provide part selection insights. All components shall be 3. NASA Team Set to Deliver Newfangled 6U
designed to avoid or tolerate errors due to non- CubeSat:
destructive Single Event Upsets (SEUs). No permanent https://www.nasa.gov/content/goddard/nasa-
loss of function shall result due to non-destructive SEUs. team-set-to-deliver-newfangled-6u-cubesat
Degradation of individual components shall be included
4. NASA, Near Earth Network (NEN) Users’ Guide:
in appropriate worst case analyses.
http://esc.gsfc.nasa.gov/assets/files/453-UG-
Several of the parts may be susceptible to 002905(2).pdf
destructive/catastrophic Single Event Effects. 5. NASA, Space Network Users' Guide (SNUG):
http://esc.gsfc.nasa.gov/assets/files/SN_UserGuid
e.pdf

Altunc 9 29th Annual AIAA/USU

Conference on Small Satellites