P. Tortora, PhD

Scuola di Ingegneria Aerospaziale, UniversitÓ di Roma “la Sapienza”, Italy


Deep space navigation systems of interplanetary spacecraft currently rely on X-band transponders, used for both range and range-rate estimation. These devices are characterized by a frequency stability (measured by the Allan deviation) whose typical value never falls below the limit of 10-13, which broadly reflects in a range rate accuracy of about 15.10-4 cm/s. Moreover, for missions in the ecliptic plane, a critical phase is represented by the solar conjunctions which, depending on the spacecraft trajectory, can last up to two weeks. In such conditions the radio data are inherently corrupted by solar plasma noise, which causes a dramatic decay of the obtainable navigation accuracy. As a matter of fact, in the orbit determination process, radio data collected when the line of sight falls within 40 solar radii are usually discarded, leading to long time spans during which navigation cannot rely on actual data.

For the Cassini spacecraft, currently in cruise flight to Saturn, 30 days of data across the 2000 and 2001 solar conjunctions were removed for navigation purposes. While this strategy is widely acceptable and proven during the cruise flight, it is not recommended during critical mission phases, when frequent ground-commanded maneuvers are executed.

The Cassini tour of the Saturn system will begin with the Saturn Orbit Insertion (SOI) maneuver, scheduled on July 1st 2004, a few days before a solar conjunction. A significant improvement of the navigation accuracy would be achieved using, for the Orbit Determination process, all radio data collected up to the SOI.

The Cassini spacecraft is currently testing a novel RF multilink technology to perform radio science experiments (RSE). The on-board configuration is based on a X/X transponder, which generates a reference signal to the Ka-Band Exciter (KEX) for the X/Ka link; furthermore a coherent frequency translator (KaT) is used for the Ka/Ka link. The primary goals of RSE are the measurement of the solar gravitational deflection during the Solar Conjunction Experiments (SCE), and the search for gravitational waves, (due for example to the coalescence of massive black binaries) during solar oppositions.

During the SCE, the estimation accuracy of the post-Newtonian parameter g will be strongly enhanced by the plasma calibration, made possible by the multifrequency link. With this method, sky frequencies in the three bands (X/X, X/Ka, Ka/Ka) are coherently combined to remove the effects of the solar plasma, the major noise source in the Doppler observable.

The preliminary analysis of the 2001 Cassini solar conjunction data, using the multifrequency plasma calibration scheme, has shown an improvement of a factor of 8 over the noise of the bare Ka/Ka observable and a factor of 100 over X/X data. At an impact parameter of about 25 solar radii, the Allan deviation is as low as 2.10-14 at integration times of 1000 s, offering a range rates accuracy of about 0.003 mm/s. As the calibrated frequency residuals exhibit a nearly white power spectrum over a broad frequency range, at 10s integration time the accuracy is degraded by a factor of 10. At an impact parameter of about 6 solar radii, the Allan deviation is on the order of 4.10-14 (1000 s integration time), still well below the corresponding uncalibrated X-band value. Further processing, including the use of the available advanced media calibrations (using water vapor radiometers and microwave temperature profilers developed for the Cassini radio science experiments) may lead to even better values.

An immediate consequence of such results is the dramatic navigation accuracy improvement which can be obtained calibrating for solar plasma the Doppler observable used by the Orbit Determination Program (ODP).

In the seminar, the steps required to compute a “plasma free” observable are described and discussed in detail. First, the algorithm to reconstruct the sky frequencies from wideband open loop data are illustrated and, using a simple model of the orbital dynamics, the obtained residuals are characterized in terms of Allan deviations and power spectra. Then, it is shown how the plasma calibration scheme, used to generate plasma free observable for each band, leads to a much improved signal stability and to a reduced spectral density.

Using the raw and calibrated sky frequencies in the ODP, a thorough comparison is performed among the Doppler frequency residuals obtained from 1) uncalibrated X-band data, 2) uncalibrated Ka-band data, and 3) calibrated X and Ka-band data. The results show that the plasma free observable generation allows the use of actual data in the orbit determination process, even when the S/C has an impact parameter of about 5 solar radii, well into a superior conjunction.