4-4 time comparison equipment for ets－Ⅷ satellite part 1 ...€¦ · comparison equipment (tce)...

135

1 Introduction

The ETS－Ⅷ (Engineering Test SatelliteⅧ)[1], scheduled to be launched in 2004, isdesigned to contribute to the development ofcutting-edge, universal, basic technologiesthat will be useful in future space activities. Avariety of experiments will be carried outusing this satellite, including a mobile com-munications experiment[2] in which a largeexpandable antenna will be deployed.

Designed in line with the mission goals ofNational Space Development Agency of Japan(NASDA), the ETS－Ⅷ will be the first Japan-ese satellite equipped with an atomic frequen-cy standard (atomic clock). Applicationexperiments will include basic research onsatellite positioning technology[1].

To evaluate the performance of the atomicclock aboard the satellite, the CRL has pro-posed a high-precision method of comparingtime between the space-borne atomic clockand a ground-based atomic clock. Thismethod was subsequently approved, and the

corresponding system has been slated forinstallation aboard the ETS－Ⅷ. Developmentand manufacture of this onboard system isnow complete, and the equipment is undergo-ing evaluation as part of overall pre-deploy-ment satellite testing.

This paper describes the outline, operatingprinciple, and structure of the high-precisiontime comparison equipment (TCE) to beinstalled aboard the ETS－Ⅷ, as well the workleading to its development. We will alsodemonstrate how time comparison in the orderof 10 picoseconds becomes possible throughthe minimization of errors relating to propaga-tion delay, ionosphere delay, and delays in thetransmitter and receiver.

2 Japanese satellite-based posi-tioning technology

GPS[4]-based systems such as car naviga-tion systems, ship positioning systems, andtime-data dissemination systems are currentlyin wide use throughout Japan. GPS technolo-

TAKAHASHI Yasuhiro et al.

4-4 Time Comparison Equipment for ETS－ⅧSatellite―Part 1 Development of Flight Model―

TAKAHASHI Yasuhiro, IMAE Michito, GOTOH Tadahiro, NAKAGAWA Fumimaru, KIUCHI Hitoshi, HOSOKAWA Mizuhiko, AIDA Masanori, TAKAHASHI Yukio, NODA Hiroyuki, and HAMA Shin'ichi

The Engineering Test SatelliteⅧ (ETS－Ⅷ) is a Japanese geostationary satellite which willbe launched in 2004. Its missions will include mobile communication experiments andapplication experiments using Cesium atomic clocks in space. Using this satellite, the CRL(Communications Research Laboratory) and the NASDA (National Space DevelopmentAgency) is planning to conduct a precise time and frequency transfer between an atomicclock on-board the satellite and a ground-reference clock. This paper describes the systemfor precise time transfer between the ground reference clock and on-board clock. The sys-tem is designed to make the time transfer at a precision of around 10 ps.

Keywords ETS－Ⅷ, Satellite positioning system, On-board atomic clock, Time comparison

136

gy is also used in a number of earthquake-observation networks and in internationalstandard time comparison[3].

GPS is, however, a system developed bythe United States military. To date, the USgovernment has not charged users, for politi-cal reasons, but its use in the future or in theevent of a defense emergency is not guaran-teed. A similar satellite-based positioning sys-tem―Russia's GLONASS[5]―presents con-cerns in terms of future stability due to poten-tial political instability. In Europe, a satellite-based positioning system has long been underdiscussion. To reduce the risk of relying toomuch on GPS, the Europeans have begundevelopment of the GALILEO project[6], nowin the system-design phase. Commercialoperation is planned to begin roughly by 2008.

Japan had lagged behind other nations inresearch and development of satellite-basedpositioning systems, but in around 1996 dis-cussions began on a future Japanese satellite-based positioning system. In particular, aworking group of the National Space Devel-opment Agency of Japan presented a report inMarch 1997, in which they agreed as an initialstep to develop the following basic technolo-gies[7] in satellite-based positioning.- Space-borne atomic clock- Time-management technology for satellitenetworks- High-precision satellite orbit-determinationtechnology

It was agreed that the CRL would developthe hydrogen maser space-borne atomicclock[8], which offers frequency stability high-er than that of the Cs atomic clock and Rbatomic clock aboard the GPS satellite, and thatNASDA would push forward with researchand development of time-management tech-nology required for satellite networks, in addi-tion to technology to enable high-precisionsatellite orbit determination.

As part of these efforts NASDA willinstall a Cs atomic clock in the ETS－Ⅷ.While deployment of a Cs clock is not part ofthe mission, the introduction of this device―with its known track record within the GPS―

will provide basic information on the underly-ing techniques of satellite-based positioning.

The CRL has proposed a bi-directionalmethod of time comparison used in the TCEfor high-precision time comparison betweenthe on-board atomic clock (pre-launch) and aground-based clock. This method will eventu-ally be used to evaluate the performance of theatomic clock aboard the orbiting ETS－Ⅷ. Thegoal of the CRL in this context is to providethe equipment required for this experimentalsystem aboard the ETS－Ⅷ.

The decision was subsequently made tolaunch a quasi-zenith satellite in 2008, and thedevelopment of a satellite-based positioningsystem using this satellite is now underway.The CRL, which is responsible for the man-agement of the time standard for this system,will develop a TCE-based method for timecomparison between the satellite and a groundstation.

3 ETS－Ⅷ (Engineering Test SatelliteⅧ)[1]

An external view of the ETS－Ⅷ is shownin Fig.1. The satellite is scheduled to belaunched in 2004. A variety of experiments,including a mobile communications experi-ment using a large expandable antenna, willbe carried out using this satellite to developsophisticated, universal, basic technologiesthat will prove useful in a range of futurespace activities.

Journal of the National Institute of Information and Communications Technology Vol.50 Nos.1/2 2003

External view of ETS－ⅧFig.1

137

NASDA intends to equip the ETS－Ⅷ withan atomic clock-the first such deployment on aJapanese satellite-in order to examine its per-formance in orbit and to acquire the basic tech-niques of satellite-based positioning. In addi-tion, basic applied experimental research willbe conducted on satellite-based positioning.

3.1 On-board atomic clockNASDA has selected the US FTS Cs

atomic clock, often used in GPS applications,as the on-board atomic clock. The use of thisdevice will enable us to acquire the basic tech-niques of satellite-based positioning.

The specifications of the Cs atomic clockare as follows:-Frequency:10.23MHz -5.5×10-3Hz

(including relativistic correction)-Weight: 13.6kg-Certainty: ±1×10-11

-Stability: 1.0×10-11 (1 to 3.6s)1.89×10-11/√τ (3.6 to 105s)6×10-14 (105 to 106s)

The equipment NASDA will use includesan S-band transceiver, an L-band transmitter,an S-band/L-band common 1.0mφantenna,and SLR (satellite laser ranging) equipment.These devices will be used to acquire basicsatellite-based positioning techniques relatingto the following:- Performance evaluation of the on-boardatomic clock in orbit and acquisition of in-orbit management techniques- Precision management techniques of satellitenetwork time and ground time- Evaluation of a techniques of high-precisionorbit determination

3.2 Outline of high-precision timecomparison equipment (TCE)

Fig.2 illustrates the method for precisetime (frequency) comparison between theETS－Ⅷ-borne atomic clock and the referenceground clock. The satellite sends a signal fortime comparison to the ground station andvice versa. Under this method, referred to asbi-directional transmission time comparison,the difference between the space-borne atomic

clock and the reference ground clock is calcu-lated using received signals and the differencein the reception times. This method allows forhigh-precision time comparison becausedelays caused by the ionosphere and atmos-phere along the propagation path can be virtu-ally canceled out in calculation, even whenthese delays fluctuate; satellite motion cansimilarly be accounted for.

In this method, both the satellite andground station are equipped with a high-stabil-ity atomic clock, allowing for the generationof coherent carrier signals and coherent modu-lation signals, as with GPS. Thus not only themodulation signal but also phase informationin the carrier signal can be used in compari-son. It then becomes possible to compare timeon the order of picoseconds, or to comparelength on the order or millimeters.

3.3 TCE principleAs shown in Fig.2, signals for time com-

parison are exchanged between the satellitestation and the ground station. The satelliteand the ground station measure the receivedsignals and the time difference.Subjects of measurement are as follows:τ1: Difference in time between the satelliteand the ground station, with reference to thesatellite time, Ts;τ2: Difference in time between the satelliteand the ground station, with reference to theground time, Te; and


Principle of bi-directional time com-parison

Fig.2

138

τg: Satellite-ground propagation time.

Then,

In this way, the subtraction of one meas-urement value from another provides the timedifference, while adding measurement valuesprovides the propagation time. The propaga-tion time in turn provides a value for satellite-ground distance.

In actual measurement, the code clock sig-nal of the 1.023MHz PN code (equivalent tothe GPS C/A code) (sent as a ranging signal)and its carrier wave-phase data. The formersignal is used to measure the approximate dis-tance, while the latter completes high-preci-sion measurement.

3.4 TCE configurationTable 1 summarizes the major specifica-

tions of TCE and the on-board atomic clocksystem.

Fig.3 is a block diagram of the TCE

(inside the blue frame) and of the on-boardatomic clock (outside the blue frame). Thesehave the following functions:- Transmission of a time comparison signalfrom the satellite to the ground (transmittingfunction)- Receipt of the time comparison signal fromthe ground (receiving function)- Measurement of delay in the satellite trans-mitter and receiver systems in real time (trans-mitter correction signal and receiver correc-tion signal)

These transmitter and receiver signalsenable calculation of time difference under the


Block diagram of high-precision time comparison equipment aboard ETS－ⅧFig.3

Major specifications of TCE and theatomic clock

Table 1

139

bi-directional transmission method. Thetransmitter correction signal and receiver cor-rection signal are used in parallel to correctpropagation delay fluctuations, due for exam-ple to temperature changes and other changesover time that are not common to the transmit-ter and receiver. For this purpose, the signalprocessing unit of the TCE (inside the greenframe) is designed to handle three channels ofsignals concurrently―the received signal, thereceiver correction signal, and the transmittercorrection signal.

The signals are handled as follows in eachof the signal routes.- Transmitter signal: The base-band signalprocessor generates transmission carrierwaves and a PN code, which are PN-modulat-ed and amplified by the S-band transmissionunit and then sent out from a 1.0mφantenna.- Receiver signal: The signals received by a1.0 mφantenna are branched by the S-bandreceiver unit. They are frequency-converted,PN-demodulated, and used for carrier-wavephase analysis and code clock measurement.- Receiver correction signal: The carrier wavesgenerated by the signal synthesizer and PNcode generated by the PN code generator arePN-modulated, inserted in the directional cou-pler in front of the S-band receiver unit, andthen processed along with the received signals.- Transmitter correction signal: The transmit-ted signals are branched by the directionalcoupler in front of the antenna, frequency-con-verted, and processed along with the receivedsignal and receiver correction signal.

3.5 Ground stationWe are in the process of constructing two

ground stations, each having an antenna with adiameter of 2 m, for experimental use: onewill be installed at the CRL Headquartersfacility and the other will be used as a mobilestation. The time comparison unit of theground station is a modified EM of TCE. Wewill conduct the same measurement, transmit-ter correction, and receiver correction as willbe performed in the satellite.

3.6 Correction of errors(1) Propagation delay

Because the signals from the satellite andground station take the same propagation path,the delay caused by the ionosphere and atmos-phere, including fluctuations in this delay willbe canceled out. Additionally, errors due tosatellite movements can be corrected by calcu-lation.

The above will be true ifτg in uplink isexactly the same asτg in downlink. In fact,however, the satellite moves in orbit duringthe time from signal transmission to reception,and the ground station on Earth also movesduring this period, due to the Earth's rotation.These movements are not the same, and thusτg in uplink is different from that in down-link. These movements do, however, repre-sent errors that can be calculated from thepositions of the satellite and ground stationand can be corrected by calculation. Addition-ally, the satellite orbit is not an ideal orbit, andif this difference is largeτg in uplink will dif-fer further fromτg in downlink. This errorcan also be corrected through calculation ofthe satellite's orbit.(2) Influence of the ionosphere[9]

The S-band frequency used in our systemis 2,656.390 MHz for uplink and 2,491.005MHz for downlink; there is a difference ofabout 170 MHz between the two. With theemployed frequency, f, and the total numberof electrons per unit area in the ionosphere,Ne, then the delay distance,⊿L, whichchanges with the group delay in the iono-sphere, can be calculated as:

When the above frequencies are used inuplink and downlink,⊿L is calculated asshown in Table 2. The difference becomes0.024 m in the nighttime during a solar mini-mum, and 0.78m in the daytime during a solarmaximum. It is thus difficult to account fordelays caused by the ionosphere. On the otherhand, since the difference in delay betweenthe ranging signals of the two frequencies (S-


140

band and L-band) transmitted from the ETS－Ⅷ can be monitored in real time, the differ-ence in measured delay between the two fre-quencies, ⊿⊿L, can be calculated by the fol-lowing equation.

The total number of electrons in the iono-sphere per unit area, Ne, can then be deter-mined (fS is the S-band frequency, fL is the L-band frequency). In this way, delays can becorrected even when different frequencies areused in uplink and downlink.(3) Delay caused by the communications hard-ware[10]

Even in bi-directional time comparison,delay (and the change in delay over time) not


Example of channel design and accuracy in time comparisonTable 3

Difference in the delay distancedue to use of different frequencies

Table 2

141

common to both transmitter and receiver can-not be accounted for; these delays translateinto error in the measurement of time differ-ence. Our system, however, can monitor andcorrect these delays in real time.

3.7 Accuracy of time comparisonTable 3 shows an example of channel cal-

culation and accuracy of time comparison. Interms of channel quality, 65dB•Hz or higher isexpected, while the accuracy of time compari-son is 1 ns or so in an ideal, completely error-free case using the PN code clock for diffu-sion in an integral of one second. Use of thecarrier-wave phase data will further improvethe accuracy to the order of picoseconds. Inreality, some errors cannot be accounted for oreliminated, so the accuracy of time compari-son will be around 10ps. Nevertheless, thislevel of accuracy is much higher than previ-ously available under conventional systems.

3.8 Proto-flight model (PFM)Fig.4 shows an external view of the TCE-

PFM. The PFM will be examined to obtainapproval for deployment aboard the satellite.After approval is obtained, the PFM will beinstalled on the satellite. We have performedthe following PFM check procedures:

- Oscillation frequency, level, and stability ofeach local signal- PN code and carrier-wave phase measure-ment capability, which will prove importantwhen the transmitter correction signal andreceiver correction signal are overlaid on thereceived signal- Thermal-vacuum test, vibration test, shocktest, and EMC test

Fig.5 demonstrates the frequency stabilityof the 2,656.390MHz signal. The test resultsshow that PFM is sufficiently reliable for usein time comparison.

4 Conclusions

We have described the TCE principle,structure, and specifications, and have shownthat a time comparison in the order of 10 pscan be conducted by minimizing errors due,for example, to propagation delay, ionospheredelay, and delays in the transmitter and receiv-er. A manufacturing and performance checkof the PFM has been completed to prepare forthe 2004 launch. In the future, we will per-form an overall satellite test and a connectiontest on the ground, when the ground stationsare built.

We would like to thank those who assistedin enabling our system to be installed aboardthe ETS－Ⅷ.


External view of TCE-PFMFig.4

2,656.390 GHz signal frequency stabilityFig.5


References1 M. Homma, S. Yoshimoto, N. Natori, and Y.Tsutsumi, "Engineering Test Satellite-8 for Mobile Communica-

tion and Navigation Experiment", IAF, No.IAF-00-M.3.01, pp.256-263.

2 Y. Kawakami, S. Yoshimoto, Y. Matsumoto, T. Ohira, and T. Ide, "S-Band Mobile Satellite Communications

and Multimedia Broadcasting Onboard Equipment for ETS－Ⅷ", IEICE Trans.Commun. , Vol.E82-B, No.10,

pp.74-81, Oct. 1999.

3 M. Aida, " 5.Generation, Keeping and Dissemination of Time and Frequency Standards 5.1Automatic Pro-

cessing System for Standard Time-Signal Generation and Data Publication", Review of the Communications

Research Laboratory, Vol.45, No.1/2, March/June 1999. ( in Japanese)

4 B. Hofmann-wellenfof, H. Lichtenegger, and J. Collins: GPS Theory and Practice, Spring-Verlag Wien, New

York, 1994.

5 A. Yasuda, "Outline of GLONASS and the report on the mission toward Russia", Japan Institute of Naviga-

tion GPS Symposium '99, No.1.3, Nov. 1999.

6 K. Ito, "GNSS '99 and a trend of GALLILEO", Japan Institute of Navigation GPS Symposium '99, No.1.2,

Nov. 1999.

7 Space Development Committee, Report on satellite positioning technology Subcommittee, " Policy on the

development of satellite positioning technology in Japan", 1997.

8 T. Morikawa, K. Takahei, M. Uehara, and M. Tsuda, "Design Analysis of a Sapphire Loaded Cavity for a

Spaceborne Hydrogen Maser", Conference on Precision Electromagnetic Measurements, No.MOP4-3,

pp.74-75, Sydney, Australia, May 2000.

9 T. Tanaka, "Erroe budget", Chap. 7, Precise positioning system by using satellites, Ed. The Geodetic Soci-

ety of Japan, Tokyo, 1986.

10 M. Imae, Y. Takahashi, M. Aida, T. Gotoh, Y. Shibuya, K. Mano, and N. Kurihara, "Delay Variation Earth Sta-

tion for Two Way Satellite Time Tranfer and Its Measurement Method", Proceedings of the 2001IEICE Gen-

eral Conference, No.B-2-3, Mar. 2001.

TAKAHASHI Yasuhiro

Senior Researcher, Time and Frequen-cy Measurements Group, AppliedResearch and Standards Division

Satellite Communication, SatellitePositioning System

GOTOH Tadahiro

Reseacher, Time and Frequency Mea-surement Group, Applied Research andStandards Division

GPS Time Transfar

IMAE Michito

Leader, Time and Frequency Measure-ments Group, Applied Research andStandards Division

Frequency Standards

NAKAGAWA Fumimaru, Ph. D.

Research Fellow, Time and FrequencyMeasurements Group, Applied Researchand Standards Division

Satellite Navigation, Satellite TimeTransfer

143TAKAHASHI Yasuhiro et al.

TAKAHASHI Yukio

Senior Researcher, Keihanna HumanInfo-Communication Research Center,Information and Network Systems Divi-sion

Posting Technology, Astrometry, VLBI

KIUCHI Hitoshi, Dr. Eng.

Senior Researcher, Optical SpaceCommunications Group, WirelessCommunications Division

Radio Interferometry, Optical SpaceCommunication

AIDA Masanori

Senior Researcher, Research PlanningOffice, Strategic Planning Division

Frequency and Time Standards

NODA Hiroyuki, Ph. D.

Japan Aerospace Exploration Agency

Development of the Satellite Position-ing Experiment System

HAMA Shin'ichi

Leader, Quasi-Zenith Satellite SystemGroup, Applied Research and Stan-dards Division

Satellite Communication, VLBI

HOSOKAWA Mizuhiko, Ph. D.

Leader, Atomic Frequency StandardsGroup, Applied Research and Stan-dards Division

Atomic Frequency Standards, SpaceTime Measurements

4-4 time comparison equipment for ets－Ⅷ satellite part 1 ...€¦ · comparison equipment (tce)...

Documents