Summary of the VSOG Mission Operations Meeting, held at ISAS, Friday 13th and Saturday 14th, December 1996 -------------------------------------------------------------------------- Objective of the meeting was to review operation plans, procedures, and practices of interactions between VSOP Ground Support mission elements and VSOG. This includes data files, formats, flow and timing. The rationale is to ensure smooth data flow and sound mission operations from launch, through in-orbit checkout, and science operations. Attendance: H.Hirabayashi, L.d'Addario, B.Carlson, W.Cannon, P.Edwards, G.Feil, L.Gurvits, M.Inoue, S.Kameno, N.Kawaguchi, T.Kii, H.Kobayashi, G.Langston, P.Newby, Y.Murata, D.Murphy, R.Okayasu, M.Popov, J.Reynolds, K.Shibata, D.Traub, J.Ulvestad, T.Umemoto, R.Wietfeldt Agenda 13th December, 9:00 - 12:30 (ISAS 5F meeting room) ~~~~~~~~~~~~~~~~~~~~~~~~~~~ Review of Agenda for the meeting. (9:00-9:15) Review of the Document Status (9:15 - 10:30) -- break -- Review of data flow between VSOG, correlator and tracking stations: (10:45 - 12:00) Review of the plan General Plan (1) Tracking station I/F's (Murata) naming conventions of the data (Murata) comments (Murphy) 13th December, 13:00 - 17:30 (ISAS 5F meeting Room) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Review of data flow between VSOG, correlator, and tracking stations: (continued 13:00 - 12:00) timelines (Murata) Socorro requirements for timeline(Ulvestad) Plan of feedback from Correlator to VSOG (Ulvestad/Carlson) Server structures ISAS server (Murata) JPL server (Murphy) Greenbank server (Langston) AOC server (Ulvestad) Calibration files Calibration file requirements from correlator (Ulvestad) Review of Data contents/format(1) (14:00 - 16:30) Reviewed by Murata + comments from ALL. Logs (Murata) logs from VSOG to Correlators logs from copier 14th December, 9:00 - 12:30 (ISAS 7F meeting room) ~~~~~~~~~~~~~~~~~~~~~~~~~~~ logs from Tracking stations to VSOG SRT schedule file Schedule to the correlators delta-T file Review of the meeting at Socorro in October. (Ulvestad/d'Addario) discussion Test plans: Status of the tests: Review Murata comments from ALL Future plan: Tracking Station Selection Rules Requirements (Ulvestad) How to select TS's in the scheduling Software. 14th December, 13:30 - 17:00 (ISAS 7F meeting room) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Initial Orbit Checkout Plan Detailed Timeline (Kobayashi) Discussions Communications (Status report / feedback) Support Plan "Operation Activity" documents Update plan Other documentation plan (Ulvestad) Action Items Review(all) Separate meeting. (16:00 - ) (ISAS 7F meeting room) ************************************************************* 1. Review of the Document Status Ulvestad presented the list of VSOG I/F Formats with the current status of development for every documents: (we present this list here with the corrections that were applied during the meeting) (1) SRT Schedule File (from VSOG to mission elements) last formal version from May 20 1996 new version (v.3.3) is prepared by Murata to be distributed in December 1996 (AI No.15) (2) Data Processing Log (from TS to VSOG) (includes correlator log and calibration data) last version May 17 1996 Log for S2 correlator shall be updated (AI No.12 for B.Carlson) (3) Performance Log (from TS to VSOG) last version November 21 1995 approved, not changed. (4) Telemetry Headers (from TS to VSOG) last version September 1, 1995 approved, not changed. (5) Time Corrections File (from TS to Correlators) last version April 24, 1995 new version November 13, 1996 (new keywords) was reported and accepted by VSOG during this meeting (6)Time Components File (from TS to Correlators) new file! November 15, 1996 (7) SRT Input Log File for VLBA Correlator (from VSOG to Correlator) last version July 1, 1996 very minor revision from March 1996 (thin tape speed) (8) Recording VSOP Data on VLBI Tapes (from TS/Copiers to Correlator) last version April 10, 1996 no changes. (9) VSOP Correlator Input Log (from TS/Copiers to Correlator) last version August 10, 1996 (internal to Japan). (10) S2 Correlator Input Log (from TS/Copiers to Correlator) last version August 14, 1996 to be updated by B.Carlson (AI No.12) (11) S2 Tape ID and User Field (from TS to Correlator) last version July 30, 1996 (12) PCAL, GAIN CAL, FLAG TABLE (from VSOG to Correlator) Socorro specification: October 11, 1996 required by AIPS (13) Flag Dictionary (from VSOG to Correlator) last version February 19, 1996 no changes. (14) Orbit Files (from Navigation to Correlator) SPICE/NAIF 1993 no changes. (15) Status Reports (from Mission Elements to VSOG) under development by Y.Murata (AI No.14) (16) Correlation Parameters (from VSOG to Correlators) under development by P.Edwards (AI No.13) (17) Schedule Summary (from VSOG to Mission Elements) under development by P.Edwards (AI No.13) (18) Tape Handling (from Mission Elements to Correlators) under development by J.Ulvestad and B.Carlson 2. Review of data flow between VSOG, correlator and tracking stations Murata presented for discussion the document "ISAS Server Operation Plan" which explains data flows between VSOG and operation elements, orresponding responsibilities, and operational time line. During most of the meeting this ocument was discussed in detail. The following issues were given the most attention: (1) Naming of SRT file consisting of plain figures was considered to be not very convenient for human treating. But the naming was already accepted by NEC as standard, and it was agreed do not introduce any changes. Murata informed the meeting about introduced format changes of SRT schedule which appears to be necessary during the development of operational software. All changes are in commands for VSOP satellite only. Updated version (ver.3.3) will be distributed to the mission elements soon (see AI No.5). (2) Extensive speculation concerning convention on treatment of outputs from tracking stations took place. Should the Data Processing Log from TS be based on one tracking pass (Murata), or one experiment (Gurvits), or one recorded tape (d'Addario)? How to create Log in a case when one tracking pass consists of two experiments (for example, the same source at two different frequencies)? If tracking station shall start a new tracking pass for the second experiment in the last case, it will take some time to initiate new pass, which will result in the unneeded lost of link with spacecraft. Fortunately, VLBA correlator is able to accept really any kind of log, unless it would require to change tapes every 5 minutes or would cover many days of correlator job (Ulvestad). Finally the agreement was achieved to apply one Log to one tracking pass, as defined by the BGNLK and ENDLK commands of SRT file. In a case of two experiments per one track (very seldom if ever to be scheduled) tape footage for the beginning of the second experiment shall be indicated in the Log, and "obscode1" being changed to "obscode2" at proper time. Murata has proposed naming convention for the Data Processing Log which seems able to resolve all ambiguities. It was unclear how to treat two VLBA tapes being recorded simultaneously. Murata proposed that Tracking stations should generate a log per recorder. In a case when tracking station will recording on two VLBA tapes, two VLBA logs will be generated; and file names are 9612122345-1.klg and 9612122345-2.klg. For example, -1 and -2 are assigned from COREL#, BGNRC#, ENDRC#, CONFG# in SRT schedule file. Murata also asked to have the Barrel roll information in FORM entry in the Log. Greenbank will supply FORM and VSOG can put "Barrel roll OFF" for DSN/JPL, because it is always "OFF". 3. Time line of delivery of operation files during the mission. Murata presented the time line of delivery of all operation files from VSOG to mission elements. The VSOP mission will be operated on a weekly basis (from Monday to Sunday). First version of SRT Schedule File for a given week will be available at VSOP Server on Tuesday (19:00 UT) four weeks in advance, orbit prediction will be ready two weeks before the observations, and final version of SRT Schedule File will be placed at VSOP Server on Tuesday (19:00 UT) on the previous week to observations. Correlator Input Log File was supposed to be constructed in two weeks after observation. That was a subject for long discussion. Ulvestad has explained the AOC Requirements on data delivery to VLBA correlator based on the demands of tape turnaround. Requested deadlines are: for VLBI tapes - two weeks after observations, for reconstructed orbit - two weeks after observations, for correlator input log - one week after observations. Finally VSOG agreed to provide Correlator Input Log File on the seventh day after observations on the basis of the best efforts, since it will depend on the availability of initial information coming from other mission elements. For example, Performance Log File from tracking station may be delayed up to 72 hours at weekends and holidays because of operator limited working hours, while Ku Telemetry Data extracted automatically from science header are available at Green Bank TS every 15 minutes. It was agreed also that tracking station shall supply Time Correction Files to correlators in seven days after observation. Data processing log will be generated after 4 DAYS after the tracking pass. 4. Feedback from Correlator to other mission elements. Plan of feedback from Correlator to VSOG and other mission elements was considered to be very important for the successful mission operation especially at the early mission phase. Ulvestad described the approach to the matter at VLBA Correlator. The problem is to make some informative fraction of correlator output be available to the mission elements without significant burden on correlator operations. Access to the data may be organized by providing some dedicated disk storage with contents available by anonymous FTP. Other mission elements could also be authorized to use Space VLBI user-support workstations by remote login. There are two kinds of data that can be provided as feedback from correlator output: a) Copies of complete correlator FITS data product, and b) Copies of routine diagnostic results. In the former case the criteria for the selection of FITS data product are not well understood. It can be done automatically or on basis of: expected strong fringes, or variety of orbital phases, tracking stations, and spacecraft elevations. The great advantage of this first approach is its completeness: any aspect can be explored over full range of possible interest. On the other hand, the big volume of data (typically 200-300 MB) represents strong disadvantage: it may take long time to transfer and/or load into AIPS. In the latter case, all plots and other results produced in routine validation of correlator operation may be presented to mission elements. The products may be "sniffer" output, intermediate data files, and plot files. They can be used as AIPS datasets via direct 'tar' transfer to compatible computer. The advantage of this approach is moderate data volume; but much information of potential interest will have been lost in these heavily processed data. Carlson explained the treatment of feedback issue at S2 Correlator. He proposed to provide standard correlator output data in the "new" UV FITS format. A filter program will be written that can extract only portions of data based on (a) source name and/or (b) UT timerange. A maximum of 2 man months work would be necessary to develop the filter program. The Canadian correlator can generate subsets of a job's FITS data based on a UT timerange right now. Output should be kept less than 10 Mbytes so that it can be transferred via internet in a reasonable time. Correlator will produce these files based on email request from the VSOG. Files will be placed in correlator's anonymous ftp areas. Reply email to each request will indicate file name and ftp location. D'Addario criticized the approach based on email request for the feedback data via VSOG. The volume of 10 Mbytes was considered to be too small to represent valuable information. The issue was not solved, and AI was set to VSOG to determine specifications on content and format of feedback information at correlators (AI N4). 5. VSOP servers status and structure. Murata, Murphy, Langston, and Ulvestad outlined the present status of VSOP file servers at ISAS, JPL, Green Bank, and AOC respectively. FTP addresses and directory structure were presented for every site. Some issues related to the management of the servers (like time terms of data updating and archiving) were also discussed. 6. Calibration files Calibration file requirements from correlator were outlined by Ulvestad. Calibration data at VLBA Correlator are normally presented in "TSM" file, which contains several sections. Currently, TSM file is edited by P.I. to create 3 different input files, for 3 different AIPS tasks. Eventually, TSM file will be processed automatically to generate AIPS binary tables, but this will cause no change in how the data should be supplied to correlator. Data from VSOG may be supplied as TSM file or 3 separate files: edit section includes spacecraft or tracking station codes as defined in flag dictionary; Tsys section must be edited to add gain record; pulsecal section includes frequency group and pulse-cal group (tone definition and actual measurements). All specifications are presented in the "NRAO AOC Requirements on VSOP Calibration Data" prepared by Ulvestad, Romney and Diamond on October 11, 1996. Gurvits pointed out that non-VLBA GRTs generate calibration file in different format, and he proposed to maintain a means for format transformation, but Ulvestad assured that VLBA Correlator will not accept different formats. Gurvits sugessted that the decisions of the meeting are valid only for SRT operation. GRT issues should be discussed later somewhere. D'Addario tried to raise the discussion on the general aspects of VSOP calibration, but there were no time for such discussion. Calibration issues were discussed by Larry, Dave Murphy, and Kobayashi at separate meeting. Some result of the discussions were reflected in a short memo written by d'Addario: "Further Comments on VSOP Calibration" which is available from Larry on request. Finally VSOG agreed with one-file mode (TSM file) of presentation of calibration data. 7. Delta T File D'Addario and Ulvestad reported about specification changes to Delta T File introduced during the meeting at Socorro in October 1996. D'Addario gave a clear insight into an understanding of different components involved in total time delay at tracking station. This explanation was useful to understand the meaning of introduced new keywords. The changes were accepted by VSOG. New specifications are described in "Time Correction File Interface" (A34300N004F), prepared by d'Addario and Langston , 1996 November 13. The document is located at: gb.nrao.edu:ovlbi/doc/deltat-interface.spe 8. Plan for prelaunch tests of mission operations Murata presented plan for various tests of mission operation. He reviewed the status of the tests. He subordinated the status into four categories: test has been completed (o); test has not been completed (x); initial data has been sent (^); no data flow (-). The following status is now for the test of operations between Tracking Stations and VSOG: I/F mode Usuda DSN GB JPL-NAV SRT Schedule o ^ ^ ^ TLM Header o ^ o - Data Processing Log x ^ ^ - Performance Log x x x - the same for operations between tracking stations and correlator: This diagram is not correct. Correlator Usuda DSN GB VLBA x o o S2 x x x VSOP o x x (these would require copying) and between VSOG and Correlators: I/F mode VSOP S2 VLBA Correl.Input Log o x o Calib. table x x x For navigation related operations: XOE file (from VSOG to JPL) x Orbit prediction: ISAS to Usuda o JPL to DSN o JPL to GB o JPL to VSOG o Orbit Determination: ISAS to VSOP Correlator o JPL to VLBA Correlator o JPL to S2 Correlator o VLBI Tapes between tracking stations and correlators: Usuda DSN GB VLBA o (thin) o o S2 ? x x VSOP o * * (these would require copying) (* means thin/thick tape) A one day operational test is planned to be conducted at tracking stations at a week starting from January 6 1997. It was appeared that DSN tracking station could not track simulated VSOP pass if no downlink signal will be detected. Therefore, both Surfsat and VSOP SRT schedule will be constructed by VSOG for the whole week. JPL Navigation Group shall supply Surfsat predicted orbit to VSOG before December 20. VSOG will also supply VSOP orbit for the test only, to JPL for generating Orbit prediction file for the trackings stations. SRT Schedule file for this test will be available from VSOG on December 31 1996. All tracking station are requested to provide VSOG with their horizon masks before the end of December. At present 5 degree elevation constraint will be applied to every tracking station. 9. Tracking Station Selection Rules SVLBISCHED software does not select tracking station in a case when there is overlap of spacecraft pass at several tracking station. Selection of tracking station is being done by VSOG CHECK software. Possible criteria for such selection were described by Ulvestad. He proposed to distinguish three classes: 1) technical or operational constraints; 2)operational convenience; and 3) scientific and technical reasons. First category includes such things like: spacecraft safety, tests, or contingencies (Usuda); necessity of near-real-time correlator (Usuda, Green Bank); rapid correlation (Usuda for VSOP, Green Bank or Goldstone for VLBA or Canadian correlator); diversity of navigation data. Requirements to minimize tape-copying or distances of tape shipment fall into the second category. Demands to avoid low-elevation tracking, and preference for longest continuous tracking passes are in the third category of priority. Murata explained current treatment of tracking station selection at VSOG. Present software starts with TS available first (early than others); when the pass is close to the end software switches to the tracking station where spacecraft has a highest elevation. Check for longest continuous pass is planned to be introduced soon. 10. Initial Orbit Checkout Plan Kobayashi presented the block-diagram and elaborate table where tests of all operational mission elements could be identified. These tests were carefully discussed. Carlson pointed out that no tests for S2 Correlator were planned in the initial phase of IOC. D'Addario has raised the question whether several sequential tests could be done in parallel, which would essentially save time for IOC. The answer was that given plan anticipates the problems which might be solved only in sequential manner, but in the best case of "no problem" many tests joined in one or few could in principle solve the IOC tasks. Other important issue raised by d'Addario was the approach to solve contingencies during the IOC. What must be the reaction time? Would the Mission be satisfied with 24 hours? Or, can there be contingencies which require very fast response (one hour)? One suggestion was to have particular email address at any operational element with regular monitoring of email-box content: suppose once per 24 hours, or once per 6 hours. It was suggested also that post address and FAX number of any operation elements must be known at any operation element. Several ideas about the contingencies were prpoposed during the discussion: 1. To appoint Primary and Secondary contact person. 2. To set operation mail address which distribute the mail through this address to all (or some) people related in the mission elements. 3. Other informations (TEL, FAX num, Address, ... ) should be in operation handbook. Because of importance of this issue an action item was set to VSOG: to develop regulations of treatment of contingency and emergency situations (AI No.9). Kobayashi proposed that DSN tests will be started form Madrid becasue of the tracking station allocation problem. But Rick said that Madrid station will be the last ready for operation. Tracking station people suggested to use both Greenbank and Goldstone every two days in the first stage of the Ku link operations. 11. Communications There were two points of discussion. One is human communication issues discussed in IOC part, another is data I/F plan. Easy fast and efficient communication between operational elements is very crucial for such complicated system as VSOP. Meanwhile, the INTERNET node at ISAS is rather slow for such communication. There were suggestions to investigate possibility of using JPL node. There is also plan to install "dial-up PPP" connection, and secondary data server at ISAS. ISDN backup line between ISAS and JPL has already been set. Many participant supported the request from Ulvestad to establish as soon as possible communication personnel at operational elements. 12. "Operation Activity" documents update plan At the end of the meeting Ulvestad and Murata reviewed once again the status of operation documents and interfaces development. In this summary the final list and status of all documents are presented in item 1. Main discussion was held about development of VSOP Operational Handbook. First incomplete draft was composed at JPL based on the so-called "Green Book" (VSOP Operational Activities). Finally it was proposed to merge this draft with the document: "ISAS Server Operation Plan" presented by Murata and discussed during the meeting. Action item was set to Edwards to be responsible for the final addition (AI No.11). Edwards is also working on the preparation of Schedule Summary and Correlator Parameters (AI No.13). 13. Separate meetings There were several small meetings: one was devoted to calibration issues (see above); second was about the S2 technical issues discussed by ISTS people (Kawaguchi, Kameno) and Brent Carlson; and the third was about the "Time component file" (Ulvestad, Kawaguchi, Kameno, and Carlson). 14. Action Items 1. Kawaguchi: To write specifications on format for bit-counting statistics results, and distribute it to all tracking stations and correlators. December 20, 1996 2.Kameno: Provide delta-T file, obtained at Usuda in simulated test of VSOP tracking, to mission elements January 1997 3. Wietfeldt: Provide for Mitaka correlator: 1) test S2 tape with "dirty time", 2) S2 tape with Surfsat tracking from Tidbinbilla January 21 1997 4. VSOG: Determine specification on content and format of feedback information at correlators. Determine rules of access to the information December 1996 5. VSOG: Distribute revised version of "SRT Schedule File" (ver.3.3) to operation elements December 20 1996 6. JPL Navigation Group: Supply to VSOG orbital elements of Surfsat December 20 1996 7. VSOG: Distribute schedule for TS tests starting on January 6 1997 (to all tracking stations) December 31 1996 8. Tracking Stations: Provide VSOG with azimuth-elevation constraints December 1996 9. VSOG: Develop regulations of treatment of contingency and emergency situations December 1996 10.Ulvestad: Present remarks on the content of "VSOP Operational Handbook" December 16 1996 (It has been done) 11.VSOG: Prepare final version of "VSOP Operational Handbook" January 21 1997 12.Carlson: Present updated specification for "S2 Correlator Log File" December 31 1996 13.Edwards: After consultation with representatives of all correlators agree upon a format for a summary of SRT files for distribution to correlators and other interested mission elements. Also establish means for passing correlation parameters to correlators. January 10 1997 14.VSOG: Develop specifications on content, format, and time line for Status Report from operational elements January 1996 15.VSOG: Update all documents and files in an accordance with the agreements achieved at this meeting December 31 1996 16.Popov: Distribute summary of the discussions that took place during the meeting to all participants December 24 1996