Transcript Folie 1

Optical Clocks in Future Global Navigation Satellites
U. Hugentobler (1), M. Plattner (2), D. Voithenleitner (1), M. Heinze (1), V. Klein (2) and S. Bedrich (2)
(1)Technische
Universität München, Arcisstrasse 21, 80333 Munich, Germany, Email: [email protected]
(2)Kayser-Threde GmbH, Wolfratshauser Strasse 48, 81379 Munich, Germany, Email: [email protected]
1 MOTIVATION: CLOCKS AND NAVIGATION
4 STABLE CLOCKS FOR PRECISE ORBIT DETERMINATION
The primary payload of a navigation satellite is its
clock. Navigation and positioning performance
depends on high precision time measurements and on
the quality of the frequency generator.
ditional servicing installations. This would be very
helpful e.g. for a tsunami warning system with remote
sensors.
Highly stable clocks such as optical frequency generators can support precise orbit determination. In order to
investigate the potential, e.g., for the system operator, simulations were performed involving a Galileo constellation
of 27 satellites tracked by eight or by five globally distributed stations (s. Fig. 4,5).
Similarly precise orbit determination will profit from the
increased stability of the observation system (s. 4).
If no or only few satellite clock parameters need to be
estimated in the orbit determination process, an
improvement of the radial orbit error can be expected.
A different a priori radiation pressure model and more sophisticated troposphere modelling was applied to simulate
the observations than for the later analysis. Pseudorange and carrier phase observations were simulated for E1 and
E5a with measurement noise of 20cm and 2 mm (1 sigma) respectively.
In order to reach the highest accuracy for positions, all
involved clocks (satellite and receiver) need to be
synchronised for each measurement epoch. Today,
only relative positions are obtained at cm-level and to
retrieve absolute positions known coordinates for
reference sites are needed.
Satellite clocks that are stable enough in order to
synchronize them at the level of a few picoseconds for
hours or days would fundamentally change this
picture. Optical clock technology would allow such
accuracy and long term frequency stability. With further
progress of clock technology a future GNSS system
could be equipped with such optical clocks, thus
installing clocks of utmost stability in space.
A large variety of applications can be envisaged:
Such a scenario would enable the absolute positioning
at highest accuracy in real time without requiring ad-
The advantages of highly stable satellite clocks can
only be materialized if error sources, e.g., in
propagation delays or satellite orbits are controlled at a
precision similar to the clock stability. Furthermore,
relativistic clock corrections have to be modelled at a
higher level of accuracy than today.
In order to investigate the effects of such a scenario
with optical clocks, the project ‚OCTAGON‘ (Optical
Clock Technologies and their Applications for Globally
Optimized Navigation) was initiated by the partners TU
Munich – Research Institution of Satellite Geodesy and
Kayser-Threde GmbH. The used software is the
Bernese GNSS Software, V5.0.
Fig. 4,5: Depth-of-Coverage for the simulated eight stations network (left) and for a five stations network (right),
adopting an elevation mask of 10°: Number of stations seen from any satellite location.
Realistic clock performances were used for rubidium frequency standards, hydrogen masers, and optical clocks to
simulate clock offset time series for satellites and ground stations. In a first run the ionosphere-free linear
combination of the simulated observations were used to estimate orbit parameters, differential code biases for
receivers and spacecraft as well as epoch-wise receiver and satellite clocks corrections.
2 OPTICAL CLOCKS
Contrary to microwave clocks, where an atomic transition in the microwave range is excited, optical clocks are based
on an atomic transitions in the optical range. Using optical frequencies as clock oscillator can yield higher accuracy
since the ticks of the clock have much shorter timing intervals. A typical block diagram of an optical atomic clock is
shown in Fig 1.
Optical clock technology in principle can be divided into
two main implementations: single ion or optical lattice
Clock Laser
clocks.
(ultra-stable)
Rf output
Frequency
Comb
Cooling
Laser
Cooling
Laser
Feedback
Electronic
Trap
Single ion clocks use a magneto-electrical trap like a Paul
trap in order to confine a charged atom locally. Optical
lattice clocks require an additional laser for generating an
optical lattice wherein a bunch of neutral atoms is
trapped.
Although single ion clocks benefit from their simpler and
more mature setup, lattice clocks potentially will yield a
higher stability. This is due to the fact that in a lattice
clock a high amount of atoms is used and thereby a high
signal to noise ratio is achieved by averaging.
An ultra stable clock laser is used to probe the transition
of the ion or neutral atoms. Laser excitation of the
reference atom is measured by a photo detector. A
feedback electronic automatically tunes the laser
frequency in order to maximize the excitation probability.
The laser frequency is then “down-converted” to radio
frequency signals with an optical frequency comb.
Reference
Atoms
Detector
Fig. 1: Block diagram of an optical clock. The main
components are clock laser, atom or ion trap and
feedback electronic for closed loop control of the
laser frequency. The frequency comb is used for
transforming optical to microwave frequencies.
Within the OCTAGON project we are estimating the
achievable performance benefits for navigation satellites
with different optical clocks. These results will then be
compared to the technical effort which is necessary to
space qualify the different optical clock technologies.
Fig. 6,7: Modified Allan deviations for different types of simulated space or ground clocks (left) and the
corresponding time series of clock corrections (right). The optical clock characteristics adopted in the simulation
corresponds to the Optical EGE.
The estimated clock corrections in Fig. 8 (top) show a periodic variation with an amplitude of about 0.5 ns and a
period corresponding to the orbital period. Evidently, radial orbit errors are absorbed by the space clock parameters.
In fact, the red curve in Fig. 8 (bottom) shows the radial displacement of the orbit with respect to the true orbit. The
green curve in the same figure shows the radial orbit errors obtained when in a second run all clock values were
fixed on their nominal values and optical clocks are used. Obviously the radial orbit error that is caused by
uncertainty in estimation of orbit parameters is no longer present.
Fig. 9 (top) shows the epoch-wise clock corrections for different types of space clocks. A once-per-revolution signal
indicating absorption of radial orbit errors can be found for optical clocks and for passive space hydrogen masers.
A degradation of the orbit has to be expected for rubidium frequency standards (blue curve).
The radial orbit errors for the 27 simulated satellites for the 5-stations-solution are shown in Fig. 9 (bottom). Radial
orbit errors reach several meters when clock values are estimated epoch-wise (red curves) and are well below 1 m if
nominal clock values are kept fixed and optical clocks are used.
3 SIMULATION OF TIME SERIES
Electromagnetic signals are represented by power-law noise processes. The ‚Discrete Simulation of Power Law
Noises‘ [Kasdin N.J, Walter T. 1992] presents an algorithm for simulating different atomic clock signals. This is done
based on the spectral density of a noise process in the frequency domain which is transformed to the time domain.
The result is a time series (s. Fig. 2) of defined length, sampling interval, noise type and scaling.
Noise types are characterized by the linear slopes of the frequency stability of the clock oscillator (s. Fig. 3) and the
scaling at the beginning of the averaging time of the Allan deviation. Usually, the stability of a specific type of atomic
clock is governed by two or three noise processes. The signals simulated for this paper are shown in Fig. 6,7. The
optical clock frequency responses are adopted from the project ‚Einstein Gravity Explorer (EGE)‘ [Schiller et al.,
2009] which defines expected and desired future optical frequency stability figures for scientific space applications.
Fig. 8: Estimated epoch-wise clock corrections for
satellite E01 based on optical clocks (top) and
corresponding radial orbit errors (bottom). The green
curve (bottom) shows the radial orbit error for nominal
readings of the optical clocks introduced as fixed.
Fig. 9: Epoch wise clock corrections for different
frequency standards (top) and radial orbit errors
(bottom) for 27 satellites for a sparse tracking network
and estimating epoch-wise clock corrections (red) or
fixing clock corrections on nominal values when optical
clocks are used (green curves).
5 CONCLUSIONS
Clocks are the main instruments required for
navigation with GNSS satellites. Highly stable and
accurate satellite clocks bear a large potential for
improvements of user positions in real-time.
If clocks are stable enough such that epoch-wise clock
synchronisation is no longer required for precise
positioning involving carrier phase, precise point
positions at cm level becomes possible in real-time
using only broadcast information from the GNSS
alone, i.e., in single receiver mode.
Fig. 2: Generation of pure power-law noise processes.
The results are time series of different noise types like
the well-known white phase noise (red). With the
combination of different noises, the simulation of
typical atomic clock values (as of a Rubidium standard
or a future optical standard) is possible.
Fig. 3: Modified Allan deviation for the main noise
types. The types are distinguished by the linear
slopes (m) of the frequency stability, represented by
the Allan deviation. Further, the slopes enable the
simulation of time series for various atomic clock
standards.
GON’ is supported by the Space Agency of the German Aerospace Center (DLR)
A large variety of applications could profit from such
improved space infrastructure, ranging from precise
navigation, e.g., for automatic navigation of ships in
ports without reference station installations, tsunami
warning systems with sensors at remote locations,
space applications such as docking manoeuvres, and,
last but not least, distribution of atomic time and a
high-precision frequency standard from space.
The paper demonstrates as an example improvements
expected for precise orbit determination based on
optical clocks. GNSS operators could profit from a
possible reduction of ground infrastructure.
Upon the result of our ongoing investigations, it can be
decided whether single-ion or optical-lattice clocks
shall be used for future GNSS. Components of the
different types of optical atomic clocks are thus
evaluated for their current technical maturity and
potential space qualification.
Depending on the technical readiness level of the
subsystems of optical clocks concerning their space
qualification status and performance requirements for
optimized navigation, the best suited optical clock
technology for future GNSS can be identified.
ACKNOWLEDGEMENT
with funds from the Federal Ministry of Economics and Technology (BMWi) based on a resolution of the German Bundestag unde
2nd International Colloquium – Scientific and Fundamental Aspects of the Galileo Programme,
14 – 16 October 2009 in Padova, Italy