2007 03 16 OPS G Forum DDOR

ESA Delta DOR: from implementation to operation : 

ESA Delta DOR: from implementation to operation R. Maddè, T. Morley ESOC, 16th March 2007

Summary: 

Summary DDOR definitions and system requirements System integration Receiver modifications and storage unit installation Communication links between DSAs and ESOC Correlator development Validation with MEX VEX operational campaign Results and comparison with JPL results Rosetta operational campaign Results and comparison with JPL results Current developments and the future Interoperability with other space agencies (JPL, JAXA) Use for GAIA, BepiColombo and other future ESA missions Support to Phoenix (next JPL mission to Mars)

DDOR definitions (1): 

DDOR definitions (1) D-DOR stands for Delta-Differential One Way Ranging DOR is the measure of the difference in signal arrival time between two stations. The observable is an uncalibrated delay between the two antennas “Delta” is respect to a simple DOR, and refers to quasar calibration of the S/C DOR Since the quasar signal is recorded on the same BW of the S/C channels, ideally any errors which are station or path dependent will cancel

DDOR definitions (2): 

DDOR definitions (2) The extent to which these error sources cancel depends on the angular separation of the two sources being observed. The maximum angular distance between S/C and Quasar should not exceed 10 deg. Thus, one is able to evaluate a potentially error-free relative station delay, which leads to an accurate determination of the S/C position in the plane of the sky The measurement accuracy is given by:

Slide6: 

Observation Sequence: S/C – Quasar – S/C (or Q-S-Q) DDOR definitions (3)

Driving requirements : 

Driving requirements Capability of simultaneous multiple channel (up to 8) open loop recording in the receiver (IFMS) Timetag synchronisation among channels better than 1 nanosecond Capability to format and store huge amount of data (tens of GB) Raw data transfer and processing within 24 hours from acquisition Fast data transfer Fast correlation process System to be developed, tested, deployed and validated in 10 months Delta-DOR overall accuracy: 1 nanosecond

Implementation approach: 

Implementation approach Modifications to implement Delta-DOR capabilities: Receiver (IFMS) modification to record in open loop up to eight channels simultaneously Installation of an External Storage Unit (ESU) for fast formatting and storage of the data Upgrade of the communication links between DSAs and ESOC to transfer up to 5.5 GB of data in 10 hours Development of a S/W correlator for Delta-DOR raw data processing

IFMS/ESU modifications: 

IFMS/ESU modifications IFMS GDSP Modifications 4 channels per GDSP with BW from 1kHz to 4MHz (depending on quantisation) Quantisation 1,2,4,8,16 bits (normally 8-bit for S/C acquisitions and 2-bit for quasar acquisitions) Synchronisation of GDSPs internal clock to a level better than 1 nanosecond 2 GDSPs per IFMS modified, to allow reception of up to 8 channels Installation of two ESUs Release the IFMS from storage burden DOR data fast formatting and storage

Communication links: 

Communication links Data volumes to be transferred from each station S/C observation: 0.25 GB Quasar observation 2.6 GB Worst case (Q-S-Q sequence): 5.5 GB Requirement: data to be transferred and processed within 24 hours from acquisition

DDOR Correlator (1)I/F definitions: 

DDOR Correlator (1) I/F definitions Correlator interfaces Station1, Station2: open loop data FD (input): orbital data to help the correlation FD (output): the final product of the correlation CP: configuration parameters S/C signal correlation Set of TM harmonics: signal characteristic permits phase extraction Quasar signal correlation Noise-like signal totally embedded in receiver noise: signal characteristic forces to go for a direct correlation method

DDOR correlator (2) S/C signal structure: 

DDOR correlator (2) S/C signal structure

DDOR correlator (3) S/C signal correlation: 

DDOR correlator (3) S/C signal correlation Phase extraction by means of a S/W frequency estimator on the carrier and the lowest order TM even harmonic (highest in S/N) Drive the phase extraction of all other tones with the model (properly Doppler-scaled) built on the lowest order TM harmonic Multiplication of the extracted phasor with a phasor computed using the FD model At this point, we have the phase sequences of each TM harmonic at each Station, sufficiently corrected for phase uncertainties The correlation result as such is “modulo 2” ambiguous After an ambiguity resolution process a S/C DOR delay (S/C) is obtained

DDOR correlator (4) Quasar signal structure: 

Quasar signal Random noise-like signal Quasar signal is totally embedded in antenna noise Need for a wider channel BW to accumulate enough Signal-to-noise DDOR correlator (4) Quasar signal structure

DDOR correlator (5) Quasar signal correlation: 

DDOR correlator (5) Quasar signal correlation Each data stream (channel) from each station is delay- and Doppler- compensated (using the model provided by FD). The delay is mostly due to Earth rotation After delay- and Doppler- correction, each data stream of Station 1 will be correlated with the corresponding data stream of Station 2 for a range of delays (few s) around the expected value (provided by FD) The analysis of the observed data is split in observation periods (“accumulation periods”) of typically 1 s, in order to keep a tolerable level of error in Doppler compensation Correlation is performed for a suitable integration time (typically 10 min) in order to maximise the signal-to-noise ratio Delay resolution is improved by the use of available multi-band recordings (enlarging the total spanned bandwidth: “bandwidth synthesis”) The result of the correlation is then added to the value given by FD model As a result, one obtains a Quasar DOR (Q)

DDOR correlator (6) Correlator output: 

DDOR correlator (6) Correlator output The final output of the correlator is a file containing three DOR measurements (in the table only the most important parameters are reported).

Orbit determination: 

Orbit determination To obtain a DOR observation, the two S/C observations are linearly interpolated to the time of one Quasar observation. Direct differencing of the observations can then be made. The obtained time is used to correct FD estimation of the delay we are looking for. The corrected delay is then used to calculate the angle between the line orthogonal to the baseline and the S/C direction Results are then calibrated for media effects Overall accuracies are mainly driven by: Maximum spanned BW SNR of S/C and Quasar signals

Example of error budget : 

Example of error budget Main Parameters

Slide19: 

System Validation and tracking campaign results

DDOR system validation with MEX: 

DDOR system validation with MEX Quasar1 – Quasar2 – Quasar1 trackings A series of Quasar-Quasar tracking was performed during December 2005 in order to validate the Quasar correlator System validation with Mars Express Orbiting S/C with an a-priori trajectory known with an accuracy of 200 m Mars ephemeris known with an accuracy of the same order of magnitude

VEX operational campaign (1): 

VEX operational campaign (1)  = ESA NNO-CEB Baseline (both 35m antennas)  = DSN GOL-MAD Baseline (two 70m antennas)  = DSN GOL-CAN Baseline (two 70m antennas)  = DSN GOL-MAD Baseline (at least one 34m antenna)  = DSN GOL-CAN Baseline (at least one 34m antenna)

VEX operational campaign (2): 

45 JPL DOR data points Overall JPL data RMS post-fit residual = 0.8 ns JPL data RMS (34-metre antenna only) = 1 ns 15 ESA DOR data points (in 5 DOR sessions) ESA data RMS post-fit residual = 1.2 ns (data weighted with 1/ of the measurement – with 0.4 <  < 2.1 ns ) 35-metre antennas only Smaller total spanned BW with respect to JPL All JPL data weighted either using  = 0.25 or  = 0.35 ns Both JPL and ESA results are below expectations because of probable mismodelling of S/C accelerations VEX operational campaign (2)

Test with Smart-1: 

Test with Smart-1 Smart-1 is (was) the only ESA spacecraft equipped with a transponder capable of generating dedicated DOR tones (placed at ±2 and ±16 MHz) Higher signal-to-noise on the spacecraft signal Wider total spanned bandwidth (better accuracy) Only pre-fit residuals available Pre-fit plane of the sky residuals (RMS = 30m) < uncertainty in orbit determination (150m)

Rosetta operational campaign (1): 

Rosetta operational campaign (1)

Rosetta operational campaign (2): 

46 JPL DOR data points Overall JPL data RMS post-fit residual = 0.16 ns 31 ESA DOR data points (in 16 DOR sessions) ESA data RMS post-fit residual = 0.53 ns (data weighted with 1/ of the measurement – with 0.4 <  < 3 ns ), all JPL data weighted either using  = 0.25 35-metre antennas only (some of JPL DOR performed using 70-metre antennas) Smaller total spanned BW with respect to JPL (ESA 9 MHz, JPL 15 MHz) Smaller channel bandwidth used for quasar tracking (ESA 2 MHz, JPL 4MHz) Two DOR performed with quasars having very low flux (< 0.1 Jy) Two DOR acquisitions were with the Rosetta transponder in self-lock status (dramatic degradation of signal quality). This never happened during JPL acquisitions Discarding these 4 DOR ESA data RMS post-fit residuals = 0.34 ns ESA results are well within the required accuracy (1 ns) but could possibly give better accuracies. This needs further investigations (ongoing) Rosetta operational campaign (2)

Slide26: 

On going developments and the future

Current activities (1): 

Current activities (1) Development of a data translator Being able to process data coming from other Agencies, ESA would be capable to provide Tracking support + data translation to the desired data format Tracking support + data correlation

Current activities (2): 

Current activities (2) Enhancement of S/W correlator capabilities The first version of the S/W correlator was able to support only a limited number of quantisation schemes and sampling rates (i.e. 50kHz with 8-bit for the S/C signal and 2MHz with2-bit for the quasar signal) The maximum number of channels to be processed was fixed to 4 The new S/W correlator will be capable to process data with all sampling rates and quantisation schemes possible with the IFMS The maximum number of channels to be processed will be up to 8 New operational modes will also be added

DOR future (1) - Interoperability: 

DOR future (1) - Interoperability Interoperability with other Agencies With JPL: ESA baseline CEB-NNO is almost perfectly orthogonal to JPL baseline Goldstone-Canberra, thus providing complementary information for OD With JAXA: NNO is almost in a North-South direction with Japanese stations (and declination is the coordinate which is worse determined by Doppler and ranging only solutions). CEB would provide a very long baseline when in conjunction with one Japanese antenna Why interoperability with VLBI network Being able to process data coming from other stations than ESA would enlarge the number of usable baselines It would also reduce the operational load on ESA stations when necessary On which missions First opportunity would JPL mission Phoenix (launch in August 2007, arrival at Mars in April 2008) All interplanetary missions

DOR future (2) – JPL Cross support: 

DOR future (2) – JPL Cross support ESA DOR support to Phoenix Phoenix has extremely stringent navigation requirement (requirement for JPL DOR is of 60 ps accuracy) It is equipped with a transponder capable to generate DOR tones at ±19MHz at X-band only JPL plan to get down to the desired level of accuracy is to use the second harmonic of the DOR tone (so to have a total spanned BW of 76 MHz). The theoretical uncertainty in this case would be of 80 ps with quasars having flux of 0.7 Jy ESA cannot use the same BW JPL is using. Limitation to 28MHz BW The use of this BW would anyway guarantee accuracies down to 0.22 ns with quasars having flux of 0.7 Jy ESA baseline is almost perfectly orthogonal to the Goldstone – Canberra, thus providing the complementary information for orbit determination We are currently developing a plan to demonstrate the feasibility of its support to Phoenix

Roadmap for JPL cross support (1): 

Roadmap for JPL cross support (1) Currently, the interoperability has been tested in only one direction (ESA spacecraft tracked by JPL antennas and correlation results provided to ESA OD team by JPL) Progressive steps of interoperability validation are: “calibration step”: ESA planetary orbiting S/C simultaneously tracked by 2 ESA and 2 JPL antennas. Independent correlation processes performed by JPL and ESA. Final DDOR results validation by ESA OD team  step completed ESA S/C tracked by one JPL antenna and one ESA antenna. Exchange of raw data and correlation performed at both ESA and JPL. DDOR data analysis by ESA OD team on going JPL S/C tracked by two ESA antennas, correlation results and OD provided by ESA to JPL  ready for Phoenix support JPL S/C tracked by one JPL antenna and one ESA antenna. Exchange of raw data and correlation performed at both JPL and ESA. DDOR data analysis by JPL OD team JPL S/C tracked by two ESA antennas, correlation results and OD provided by ESA to JPL

Roadmap for JPL cross support (2): 

Roadmap for JPL cross support (2) Step 2 would need an ESA orbiting S/C to be tracked using the four stations: - CEB (ESA) - NNO (ESA) - Robledo (JPL) - Goldstone (JPL) The tracking should be structured as follows: Simultaneous tracking from CEB-NNO and Robledo Simultaneous tracking from Goldstone-Robledo and CEB This step would validate the ESA data translator and correlator with JPL correlator

Roadmap for Jaxa cross support: 

Roadmap for Jaxa cross support Jaxa would like to use K5 raw data format This is a VLBI compatible data format (very similar to Mk5 format) The translator under development will be able to treat the Mk5 format An agreement for the formalisation of cooperation on delta-DOR has been reached in January 2007 Definition of interfaces on going Future steps: Definition of level of cooperation Identification of suitable missions, for cooperation demonstration Preparation of a suitable test campaign

Example of JPL and JAXA cross support: 

Example of JPL and JAXA cross support

DOR future (3) – Inside ESA: 

DOR future (3) – Inside ESA DOR with Kourou Kourou has been recently upgraded with Hydrogen Masers and IFMS/ESU. The station is potentially ready to support DOR BepiColombo DOR will be needed as well for BepiColombo navigation during critical phases (Venus fly-bys) and Mercury orbit insertion BepiColombo will be equipped with a transponder capable of generating dedicated DOR tones in both X- (total spanned bandwidth of 40 MHz) and Ka- (total spanned bandwidth of 160 MHz) bands. The use of a much larger spanned BW will permit better results in terms of S/C position accuracy (estimation below 100 ps in Ka-band) GAIA GAIA has very stringent navigation requirements when orbiting in L2, making this mission a perfect potential user of DOR EXO MARS, LISA-Pathfinder, LISA, Don Quijote, SOLO

DOR future (4) – Related techniques: 

DOR future (4) – Related techniques Station location DOR could be used to check the position of ESA stations This would be done using for each observation one ESA and one VLBI station which position is already well known Phase referencing Phase referencing is a powerful interferometric technique to get absolute differential phase measurements It was used to get a-posteriori results for the descent orbit of Huygens Same Beam Interferometry (SBI) SBI is another interferometric technique that makes use of a planetary orbiting S/C to precisely locate another S/C which is in the same antenna beam The technique could be used at Mars, where there are several S/C already orbiting

Conclusions: 

Conclusions A lot of work has been done in the last 2 ½ years Much more work is going to come Currently ESA is the only place investing on cross-support Such (moderate) investment will allow ESA to be in a favourite position for the years to come However, the current activities are only marginally improving the quality of our service, in terms of end-to-end performance In order to improve in performance, we need to invest more on the following areas: Troposphere calibration techniques (a very simplified model is used today) Ionosphere calibration techniques (we currently depend on JPL for that) We need to develop a wider band receiver (IFMS limitations)

Acknowledgments: 

Acknowledgments Delta-DOR Team OPS-GS Mattia Mercolino Marco Lanucara Ricard Abelló Gunther Sessler Javier De Vicente OPS-GF F. Budnik N. Schlecht OPS-ECT M. Bertelsmeier G. Buscemi