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

net/publication/292707628

SDR-based system for satellite ranging measurements

Article in IEEE Aerospace and Electronic Systems Magazine · January 2016

CITATIONS READS

10 1,876

3 authors, including:

Alessandro Di Carlofelice Piero Tognolatti

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Alessandro Di Carlofelice on 13 January 2021.

The user has requested enhancement of the downloaded file.

Ilaria Lucresi, Alessandro Di Carlofelice, Alessandro Alilla(*), Marco Faccio, Piero Tognolatti
Dept. of Industrial and Information Engineering and Economics, University of L’Aquila
L’Aquila, Italy
(*)
now with Infineon Technologies, Padova, Italy
[email protected]
Introduction
Accurate ranging measurements from an Earth Station to a Satellite are of interest for many
reasons, including tracking operations, in-orbit measurements and prediction of ephemerides.
The collected data can be used for many different applications. One of the most recent
applications is the distribution of synchronization signals for terrestrial systems, as meticulous
time-delay measurements are necessary to accurately synchronize, to a given reference,
remote clocks located in different positions. The range measurement of a satellite is obtained
by measuring the round-trip transit time of a marker signal transmitted and received by an
Earth Station (E/S). The real-time determination of the range values can be obtained using
tone ranging or pseudonoise (PN) techniques that modulates a microwave carrier [1]-[4].
When a satellite is provided with retro-reflectors, which are able to efficiently reflect the
received photons, the ranging measurement can be even made using light pulses produced by
a LASER [5], [6]. In tone ranging, the transmitter sends tones of different frequencies, the
lowest frequency, having the longest wavelength, defines the unambiguous range and the
highest frequency defines the precision of the range [7]. The PN ranging techniques do not
require high transmission power since the receiver detects the PN sequence using a
correlation process. These PN methods are widely used in communication systems, such as
the Direct-Sequence Spread Spectrum (DS-SS), Code Division Multiple Access (CDMA),
and in the Global Satellite Navigation System (GNSS), which includes GPS, GLONASS and
GALILEO. In ranging measurements, the PN sequence length (i.e. number of so called
“chips”) determines the maximum unambiguous distance where the satellite can be located.
The measurement accuracy depends on the used chip-rate, which determines the chip-length
(minimum measurable distance). The higher the chip rate, the larger the PN sequence
spectrum. Ranging techniques based on PN sequence can operate with very low Signal-to-
Noise (S/N) ratios and with low signal power. For this reasons they can share the same
frequency band of an already existing signal, without producing significant impairment to the
latter. Nowadays, SATRE modems are among the most used instruments to realize ranging
measurements with the PN techniques [8].
In the present paper, the PN techniques were used to make ranging measurements on a
geostationary satellite, using the ANRITSU MS2830A vector signal analyzer and generator.
The originality of this approach is due to the use of a simple, general-purpose, laboratory
instrument, instead of dedicated expensive equipment. The concerned satellite was “Telstar
12” of Eutelsat, located on geostationary orbit at 15.0° W. The measurement campaign was
made with the support of Telespazio S.p.A., Fucino Space Centre, which provided satellite
transponder, uplink and downlink segments. The developed Software-Defined-Radio (SDR)
system performed real-time accurate ranging measurements. In the following sections, we
describe briefly the proposed approach, the measurement setup and then the obtained results.
The SDR approach
The developed system is based on one transmission and one receiving block. The
transmission system is shown in Fig. 1, where all blocks, except the ones within dashed box,
are software defined. Two pseudo-noise (PN) Gold sequences of different lengths are
generated at a given common chip-rate. Sequence lengths and chip-rate shall be accurately
selected to allow good resolution and appropriate measurement range.

Fig. 1 Block diagram of ranging transmitter. All blocks, except the ones within dashed box,
are software defined.

We refer to such sequences as short sequence and long sequence, made of 511 chips and
65408 chps, respectively. Notably, the long sequence contains 128 short sequences and is
obtained truncating a 216-1 Gold sequence. These PN sequences represent real and imaginary
parts of the baseband complex signal, respectively. This signal is filtered by square-root
raised cosine filter to limit bandwidth without producing intersymbol interference (ISI).
Filtered signal is then digitized by software, through “Anritsu IQ-producer”, and converted to
a Quadrature Phase Shift Keying (QPSK) signal at intermediate frequency (IF) by the Signal
Generator module of the ANRITSU MS2830A-021. The QPSK modulation is made up of two
Binary Phase Shift Keying (BPSK) modulations, one for the short sequence and one for the
long sequence, which are located on the in-phase and quadrature channels of a QPSK
modulation, respectively.
The choice of the PN sequences, and specifically of Gold sequences, is due to their
characteristics, such as autocorrelation and intercorrelation properties, which make them
particularly suitable for communication techniques based on the Code Division Multiple
Access (CDMA). In fact, the autocorrelation of a sequence gives a maximum value, and the
correlation among sequences gives values close to zero. Another useful feature of the PN
sequences is that, given their random, noise-like nature, they are particularly suited to spread
spectrum transmissions on channels already occupied by other kinds of signals.
In addition, the PN sequences are easily generated and this makes their selection particularly
desirable. Analytically, a characteristic polynomial of same degree corresponds to each
sequence of degree n. The sequence generation, started from a random seed, is expanded by
means of an algorithm to obtain the desired sequence. The expansion is achieved by specific
logic functions that can be easily implemented in programmable logic. If the introduced seed
has length n, then the maximum length of the sequence that can be generated will be equal to
(2n-1)·n bits. One of the simplest and most widely used architecture for PN sequences
generation is the linear feedback shift register (LFSR). It is a shift register with feedback that
is flanked by a simple network of exclusive-or logic gates. This is easy to implement both by
hardware and software. The produced sequence will have the following generator polynomial:
g(x)  g n x n  g n1 x n1    g1n1  g 0

where coefficients g0 and gn weigh 1, other gi can have a value of zero and one.
Chip rate and length are two important parameters to be selected when the PN sequences are
used for ranging applications. High chip rate allow to have great spatial resolution: minimum
round-trip measurable distance is, in fact, equal to
c
Lchip 
2  chiprate

where c is the speed of light in the vacuum. The Pseudo Noise sequence length determines the
maximum unambiguous distance interval where the satellite can be located. At a given chip
rate, long sequences allow to measure larger intervals with lower accuracy as a smaller
number of sequences is averaged during the observation time. Short sequences improve
accuracy, but make it impossible to determine the exact position of far objects, as they cannot
resolve ambiguity due to a number of integer sequences contained in the distance undergoing
the measurement. In our system, we decided to carry out an initial coarse identification of the
satellite position using the long sequence. The measurement was then refined through an
averaged readout from the short sequence. Transmitted sequences were generated at one
sample per chip, then an upsample operation was performed during filtering, to obtain 10
samples per chip at the modulator input, allowing a very low out-of-band emission.

Fig. 2 Block diagram of ranging receiver. All blocks, except the ones within dashed box, are
software defined.

After transmitting the signal, it transited through the satellite “bent-pipe” transponder, and
then it was received and processed.
The receiving system is illustrated in Fig. 2. On the receiving side, the IF signal coming from
E/S downconverter enter the signal analyzer, which provides to demodulate it. Real and
imaginary components of the received signal, sampled at 4 samples per chip, are then filtered
with square-root raised cosine filter, an operation that performs an upsample by a factor of
two. So the post-processing algorithm works with signals at 8 samples for symbol. The two I-
Q received sequences are then sent to correlators, which uses a local replica of the transmitted
sequences.
The phase of peak values at the output of the short-sequence correlator were used to calculate
the frequency shift due to the Doppler effect and make, if necessary, the appropriate
corrections. Since transmitting and receiving are synchronous processes, sharing a common
master clock located inside the MS2830A analyzer, the time position of correlation peaks can
be used to calculate the round-trip delay, allowing to compute satellite range in real time. The
positions of correlation peaks for the short sequence were properly averaged with a circular
mean and allowed to compute fine satellite range within an interval of length . Correlation
peaks at the output of the long-sequence correlator allow to compute the coarse satellite range,
with no ambiguity.

Measurement setup

To test the validity of the developed approach, a first set of measurements was made in the
laboratory, using the following setup:
 Anritsu MS2830A-021
 laptop computer
 Coaxial Cable (RG58) of different lengths.
The stability of Anritsu MS2830A 10 MHz internal reference oscillator ensures frequency
stability up to aging rate of ±1×10-5/year and temperature stability of ±2.5×10-6 (5°to 45°C).
As a preliminary laboratory test, fifteen measurements were performed for each cable to
evaluate its length. After that measurements values were averaged, a r.m.s. error of about 0.5
m was found, which demonstrated the correct behavior of the setup.
Satellite ranging measurement were then carried out at Fucino Space Center which provided
uplink and downlink in Ku-band, as well as a dedicated band on Eutelsat's Telstar-12 satellite
(15.0°W) “bent-pipe” transponder.
The entire measurement chain is shown in Fig. 3. The transmitted QPSK signal was
composed by two PN sequences, as described in the previous section. The signal had an IF
carrier frequency of 70MHz, a power level of -30 dBm at the modulator output, a rate of
2.048Mchip/s, and a bandwidth of about 2.7 MHz.

Fig. 3 Measurement setup used at Fucino Space Center.

A Comtech modem was used to simulate the main signal to which low-power ranging signal
should have been added. It generates a DVB-S signal with the following characteristics:
carrier 70MHz, power level -19dBm, bit rate 2.048Mbit/s, FEC ¾, and roll-off 0.35.
The sum of main (payload) signal and the low-power ranging signal was sent to the E/S
upconverter and High Power Amplifier (HPA). The ranging signal power at the HPA output
was +19 dBm, while the power of the main signal was +30 dBm. The antenna was a 5m
diameter Cassegrain dish, having a gain of 54 dBi. Looking at the receiving chain, it is
possible to observe that part of received signal came back to the Comtech, allowing to
calculate important parameters, such as Bit Error Rate (BER) and Eb/N0. Measures automation
was made possible by use of a PC, connected to the MS2830A via an Ethernet link.
The DVB-S received-signal spectrum (magenta) and the PN ranging signal (blue) can be
observed in Fig. 4, together with adjacent signals sharing the same transponder.
Results
A preliminary calibration was performed to determine and cancel the contribution of cables
and devices between our system and the antenna. A loop-back was connected from the
upconverter to the downconverter and the round-trip delay associated with this path from the
signal analyzer to the antenna measured. An average delay of 4.114 µs (standard deviation 19
ns) was found, corresponding to an equivalent length in air of 1234 m (standard deviation
5.7m). All satellite ranging measurements are affected by a constant time offset due to
satellite group delay. This delay is a characteristic of each satellite and is given by the satellite
manufacturer.
Fig. 5 shows, in blue, the ranging measurements results for a period of three days, and, in
black, the range predicted with the simplified model perturbation propagator (SGP4/SDP4)
from Keplerian Elements (in TLE NASA format) for an “epoch” close to the period of
interest. Our ranging measurements were not continuous, since it was necessary release the
satellite link from time to time. Ranging data were acquired every 15 seconds. Looking at Fig.
5, it is possible to observe the typical sinusoidal pattern of geostationary satellites orbit with a
period of the order of 24 hours. Peak-to-peak swing was of 18 km. After interpolating the
measured data with a fitting curve, a r.m.s. error of 4.8m was observed, which indicates a very
small dispersion of measurements. This campaign also confirmed a known fact: SGP4/SDP4
prediction accuracy for geostationary satellites is in the order of a few tens of kilometers [9]-
[11]. Unfortunately it was not possible to compare results with actual measurements since the
Fucino Space Center doesn’t perform ranging measurements on Telstar 12, which was the
only satellite with available bandwidth at the test time.
Moreover, it was observed that performance of the DVB-S signal is not substantially
degraded by the ranging signal. The latter is transmitted to a power level comparable to that
of thermal noise, as shown in Fig. 4. In particular, the Eb/N0 parameter at the DVB-S receiver
demodulator was impaired by less than 1dB.

Fig. 5 Geostationary satellite Telstar-12 range measurement performed, with the proposed
approach, in July 2013 from Fucino Space Centre (FOC), Italy. Thick blue curve shows
measured data, while thin black one is the prediction based on SGP4/SDP4, which is known
to have an accuracy of few tens of kilometers, when used for geostationary orbit. The inset
shows how each measurement point, acquired every 15 seconds, was spread with respect to
the red fitting curve. A r.ms. spread of about 5 meters can be observed.

Conclusions
The proposed PN technique for ranging measurement of a geostationary satellites by means of
a laboratory signal generator/signal analyzer and a Software-Defined approach produced
satisfactory results even if a general purpose instrument, such as the Anritsu MS2830A-021
was used instead of an expensive and hard to find, dedicated hardware to the specific
application. Our system allows to obtain a r.m.s. errors of about 5 meters by exploiting
channels that already occupied by other satellite links, in the specific case a DVB-S signal,
without the need of a dedicated bandwidth.
Acknowledgment
The authors would like to thank Gianni Riccobono for his positive response to their request to
use Telespazio facilities, and Guido Mancini and Fabrizio Mancini for their excellent
technical support.
References
[1] H. B. O. Bekhit, E. El Diwany, S. H. El Ramly, “Design of Ranging Codes for Low-Earth
Orbit Satellites”, Recent Advances in Space Technologies (RAST), 2011 5th International
Conference on, pp. 324-329, Istanbul, 9-11 June 2011.
[2] A. Weinberg, “Precise satellite ranging and timing system using pseudonoise bandwidth
synthesis” United States Patent: 5,041,833, Aug. 20, 1991.
[3] J. Berner, “Regenerative pseudonoise ranging”, Jet Propulsion Laboratory, National
Aeronautics and Space Administration, 2001.
[4] R. L. Pickholtz, “Theory of spread spectrum communications – a tutorial”, IEEE Trans.
Comm., Vol. 30, pp. 855 – 884, May 1982.
[5] C.G. Lehr, “Geodetic and Geophysical Applications of Laser Satellite Ranging”,
Geoscience Electronics, IEEE Transactions on
vol. 7, pp. 261–267, 1969.
[6] T. Johnson, H. Plotkin, P. Spadin, W. Wells, S. Moss, “A laser satellite ranging system”,
Quantum Electronics, IEEE Journal of, vol. 3, pp. 233-234, 1967.
[7] P. Maldari, “ESA's new high-performance tone-ranging system”, ESA Bulletin (ISSN
0376-4265), no. 34, pp. 54-59, May 1983.
[8]http://www.timetech.de/
[9] R.B. Ward, F.L. Strubel, “Precision Earth-to-Satellite Range Measurements Using Delay-
Lock Techniques”, Instrumentation and Measurement, IEEE Transactions on, vol. 19, pp 118-
124, 1970.
[10] D. Wei, C. Zhao, “An Accuracy Analysis of the SGP4/SDP4 Model”, Chinese
Astronomy and Astrophysics, vol. 34, pp. 69-76, 2010.
[11] D. Vallado, “SGP4 Orbit Determination”, AIAA/AAS Astrodynamics Specialist
Conference and Exhibit, AIAA 2008-6770, Honolulu, Hawaii, 18-21 August 2008.
Ilaria Lucresi: Ph.D student Department of Industrial and Information Engineering,
University of L'Aquila: data processing systems, telecommunications systems and networks.
Thecnical adviser for electronic devices development at Antares s.c.a.r.l. (Nov.2012-
Apr.2013) : ISHTER Project - time and frequency dissemination system.
Project Planner of Space Satellite Equipment at Thales Alenia Space (Jun. 2011-Jun.2012):
planning, schedule tracking and project reporting of design, engineering, procurement,
manufacturing, integration and test acitivities. Planner of trasmitter production line.
Master's degree, Master of Science in Electronic Engineering cum laude and academic
mention, in 2011.
Alessandro Di Carlofelice: received the Laurea degree in Electronic Engineering in 2006
from the University of L’Aquila, Italy. He achieved the Ph.D degree in microwave radiometry
for remote sensing and bio-medical applications in 2011. In July, 2006 he joined the ESMO
(European Student Moon Orbiter) program supported by ESA to design a microwave
radiometer payload. Since November 2006 he joined the EXOMARS program with
Telespazio and his specific task is the detailed analysis and trade off of the different
communication scenarios options for the ROCC (Rover Operation Control Center) in the
support of the rover mission operations. He joined the testing of hybrid subsystems
(Transmitter in Ku-band) with Thales Alenia Spaces (Sept. 2007 – Dec. 2007). Since
February 2011, he has been with the Department of Electrical Engineering and Information,
University of L’Aquila, as a Postdoctorate Researcher. His research activity is in space
systems for telecommunications and remote sensing. He is member SIEm (Italian Society of
Electromagnetics).
Alessandro Alilla: In 2009 obtained his Master of Science in Electronic Engineering at
University of L’Aquila. His thesis was centered on FPGA digital design in collaboration with
R&D Flash Memory at Micron Italy. In April 2013 he finished his Ph.D. focused on FPGA
digital systems. In 2008 he got a specialized certificate in Nanotechnology. Since July 2013
he has worked as Product Engineer at Infineon Technologies Italy in power supply division
for automotive applications.
Marco Faccio: is currently Associate Professor of Digital Electronic Systems at University of
L'Aquila (Italy). His present research interests include the hardware architectures for fast
signal processing, the digital sensor interface systems, the AtoD conversion for smart sensors,
CMOS image sensor technology and interface, and low voltage digital systems. He is co-
author and author of more ninety international papers in this research fields.
 Piero Tognolatti (M’83) received the Laurea degree in Electronic Engineering from the
University “La Sapienza” of Rome, in 1981. From 1982 to 1984, he worked at the Research
Division of Telespazio S.p.A., the Italian Satellite Communication company. From 1984 to
1992 he was at the University of RoTor Vergata. In 1992 he moved to the University of
L’Aquila, as Associate Professor of Microwaves, where in 2000 he became Full Professor.
From 2007 to 2012 he serves as Head of the Electrical Engineering and Computer Science
Department (DIEI) at University of L’Aquila. His main scientific interests are numerical
techniques for electromagnetic field computation, medical application of microwaves,
interactions between electromagnetic fields and human body, signal processing,
electromagnetic compatibility, and design of passive microwave circuits including film bulk
acoustic-wave resonators (FBAR). His interests also include software defined radio issues. He
has consulted for Italian industries involved in Aerospace and Avionics. Piero Tognolatti is a
member of IEEE (since 1983) and SIEm (Italian Society of Electromagnetics).

View publication stats