E-EVN and VSOP-2 Zsolt Paragi (JIVE), for the EXPReS project.

e-EVN and VSOP-2

Zsolt Paragi (JIVE),for the EXPReS project

2008 May 14-16 VSOP-2 meeting, Bonn Slide #2

Relevance

The EVN will be a key partner to VSOP-2. Current developmentswill have a direct impact on VSOP-2 operations.

Talk

This talk will focus on data transfer and network reliability related issues. The EXPReS e-VLBI project will be introduced. Challenges, solutions, and the most recent results will be shown. Will closewith a few thoughts about the (e-)EVN and VSOP-2.


What is EXPReS?

• EXPReS = Express Production Real-time e-VLBI Service• Three year project, started March 2006, funded by the European

Commission (DG-INFSO), Sixth Framework Programme, Contract #026642

• Objective: to create a distributed, large-scale astronomical instrument of continental and inter-continental dimensions

• Means: high-speed communication networks operating in real-time and connecting some of the largest and most sensitive radio telescopes on the planet

• Initiated and formerly coordinated by Mike Garrett, now coordinated by Huib J. van Langevelde

http://www.expres-eu.org/

http://www.expres-eu.org/



e-VLBI for science• Regular e-VLBI test observations in every six week (on average)

• 24h time is pre-allocated for science observation during each e-VLBI test

• Observing proposals submitted up to normal deadlines - short projects, preparing for normal EVN runs - spectral line experiments - continuum, including ToO and triggered projects

• May trigger proposals 24h hours before observations

• Science operations so far at 1.6 and 5GHz at 32-512 Mbps (Cm, Jb2, Mc, On, Tr, Wb)

• More info at http://www.evlbi.org/evlbi/evlbi.html

http://www.evlbi.org/evlbi/evlbi.html


Science projects in 2006/2007• Cyg X-3, 20 Apr/18 May, 128 Mbps, Tudose et al.

• GRS1915+105, 20 Apr, 128 Mbps, Rushton et al.

• LSI +61.303, 256 Mbps, 26 Oct, Perez-Torres et al.

• Algol, 26 Oct/14 Dec, 256 Mbps, Paragi et al.

• Calibrators near M81, 14 Dec, 256 Mbps, Brunthaler et al.

• INTEGRAL microquasar candidates, 14 Dec, Pandey et al.

• “double header” run, 15 XRBs, 29 Jan, Rushton & Spencer

• Calibrators, 21 Feb, 256 Mbps, Tudose et al.

• J2020+3631 microquasar candidate, 28 Mar, 256 Mbps, Martí et al.

• Cyg X-3, 12-13(?) Jun, 256 Mbps, Tudose et al.

• Stellar maser search, 22-23 Aug, 32 Mbps, Langevelde et al.

• INTEGRAL source redo, 6-7 Sep, 256 Mbps, Pandey et al.

• Type Ib/c SN 2007gr, 6-7 Sep, 256 Mbps, Paragi et al.


EXPReS challenges and solutions• TCP protocol operating on normal network limited operations to 256 Mbps even within Europe

- lightpathes to telescopes (dynamically allocated in the future?)- UDP-based protocol developed

• Mark5A software was not stable enough, modified software• Station Unit crashes: rewrote software, SU reboot during

correlation job is now possible; Mark5B will retire SUs• Station Mark5As may be rebooted from JIVE without stopping the correlation job• Longest uninterrupted e-VLBI job is 12h 43m!• 1024 Mbps formatted data over 1 Gbps equipment does not

work: drop few percent of the packages (but keep headers)• Connection very limited to some telescopes: drop subbands

(being tested at the moment)


Test results

APAN live demo in China (August 2007): fringes between EU telescopes and Shanghai and Mopra. Demonstrated feasibility ofglobal e-VLBI observations with an ad-hoc array. Note Australiandata format was converted in real-time.

Happy JIVE staff in Dwingeloo


Science results from the ‘Oz-demo’

ATCA, Mopra and Parkes observations at 512 Mbps, correlated

at JIVE. Target: the nearby supernova in the LMC, only visible

from the South.

One of the breathtakingHubble Space Telescopeimages of SN1987A


And with VLBI…

Highest resolution image of SN1987Aso far (first VLBI!) – not bad for a three telescope array. Countours showsan earlier ATCA image.

Various fits to the data to measure the sizeof the expanding remnant.

Tingay et al., (in prep.)


Tests in 2008, happy new year for EXPReS

First 1 Gbps e-EVNimage with packed dropping.

Cm, Jb, Mc, Tr, Wb participated. Not allhad fringes in allsubbands, but from thedata transfer point of view it was a 1 Gbps result.


New e-telescopes every week!

• Effelsberg joins the e-EVN on 1st of April 2008, in all subbands up to 1 Gbps data rates.

• Most recently Hartebeesthoek live demo fringes at 32 Mbps

• …and Arecibo-TIGO (6m dish in Chile) fringes at 64 Mbps


Gentlemen pleased in South Africa


…and in JIVE

This was a few minutes with fringes between Arecibo and Hart


Science highlight from 2007

Supernova 2007grdetected on 6 Sep.Just ~400 uJy!256 Mbps, no Ef.

First ATel messagefrom the e-EVN.

Followed-up with aglobal VLBI array.

Paragi et al. (2007)


Science in 2008• Arp299 SNe by Perez-Torres et al. in April;

very robust operations (longest correlation job at JIVE; see

MERLIN image of target and e-VLBI throughput graph to the right )

• Followed by the first ever triggered e-EVN triggered project: Cyg X-3 accretion disk state change (Tudose et al.)

ATel publication. • In May Cyg X-3 flared, three epochs e-

EVN ToO – the first ToO observations on unscheduled date. All of these at 512 Mbps. First pipeline image from 4h of data produced during the experiment.


JIVE software correlator• JIVE software correlator replaced NICT (CRL) software for ftpfringe tests.

• Doing R&D on GRID-based software correlation.


e-EVN and VSOP-2: data transfer

• The e-EVN will aim for 10 Gbps operations, which is feasible already now from networking point of view. VSOP-2 will be

limited to 1 Gbps.

• The EVN will not abandon disk recording completely, which ensures operations with VSOP-2 (needs accurate orbit for correlation, thus cannot be real-time); but data may be put to the network immediately instead of local storage at the telescopes.

• Central data storage and distributed software correlation on the GRID is a possibility. Data acquisition system difference is not a problem, but software must be there to translate data on the fly.


e-EVN and VSOP-2: operations

• Although real-time space VLBI is not possible, must establish a quick fringe checking mechanism like the ftp tests in the EVN. This will allow monitoring of the array performance.

• Automatic data pipeline and easy access through data archive like in the EVN would help space VLBI users.

• e-EVN quick fringe-test results prior to VSOP-2 survey (or ToO) observations may be desirable; note that because the candidate sources are very variable at these frequencies a pre-launch survey may not be useful at all. Even total flux density measurements prior to VSOP-2 observations will not guarantee source compactness (although likely will be sufficient most of the time).

E-EVN and VSOP-2 Zsolt Paragi (JIVE), for the EXPReS project.

Documents

mbps slide

mbps formatted data

expres project slide

dwingeloo slide

hart slide

moment slide

expres evlbi project

uninterrupted evlbi