Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.atnf.csiro.au/management/atuc/2005dec/docs/3.5_VLBI_report_nov05.doc Дата изменения: Mon Nov 28 02:02:19 2005 Дата индексирования: Sun Jun 27 16:03:47 2010 Кодировка: Поисковые слова: п п п п п п п п п п п п п п п п п п п

ATNF ATUC MEMORANDUM
To: ATUC
From: Tasso Tzioumis
Date: 25 November 2005
Subject: VLBI report - LBA Operations & eVLBI

1) Operations
In the past 6 months there was only 1 VLBI session, 1-6 November 2005.
No major technical problems were reported from any observatory,
although some observations were affected by high winds at many
telescopes. The high winds also resulted in some loss of power at
Hobart.

The next VLBI observing session is scheduled for 9-17 March 2006.

Starting in the OCT2005 observing semester, the disk-based VLBI
recording system has been officially offered as a National Facility
observing mode. Correlation will be done on the Swinburne software
correlator. Available time is restricted primarily by the expense and
availability of disk units.

There are problems with an H-maser at Narrabri, the H-maser at Ceduna
and the H-maser at Mopra. All these will need refurbishing and
arrangements are in progress to bring a team from Russia to refurbish
the H-masers. This may not be possible by the March 2006 VLBI run but
alternative arrangements will be made to support these observations.

2) VLBI tests
a) Real-time fringe testing
Real-time fringe testing using the disk systems has now been
implemented at all antennas. Fringe tests are routinely performed for
every LBA observational setup. This has greatly increased the
reliability and robustness of the LBA network.

b) mm-VLBI fringes
At the end of August the first mm-VLBI tests in Australia were
performed between Narrabri and Mopra. Fringes were detected very
quickly using the real-time fringe testing machinery on the Swinburne
software correlator. These tests used the old SIS receiver at Mopra and
it is planned to repeat the tests with the new mm receiver.

c) NZ VLBI tests
Some test observations between Australia and NZ were done in May, July
and November. The antenna in NZ is only 6m diameter, operating at 1650
MHz. The NZ VLBI hardware was extensively tested at Hobart and Parkes
and performed well in all these tests. Fringes were eventually found
using the Swinburne software correlator after integrating for a few
minutes. This demonstrated the feasibility of VLBI to NZ but also
highlighted the fact that a more sensitive antenna is needed.

d) Spacecraft Navigation
After the Huygens success last January, interest in spacecraft
navigation using VLBI has been maintained. Tests are planned with the
Japanese and discussions have been held with NASA at JPL. NASA is
interested in developing better tracking facilities in the southern
hemisphere. We have agreed to start low-level feasibility tests, which
may lead to substantial future collaborations.

3) eVLBI Project
a) Fast network connections
The broadband links to the telescopes were funded by CSIRO in mid-July.
The funding formula provides for equal sharing of the ~$2M cost between
CSIRO IT and ATNF over the next 15 years, and it will cost ATNF about
$150k per year. This project would lead to 1 Gbps links in 2006 and
facilitate 10 Gbps links in a few years.

Work on the "last-mile" fibre connections to the antennas has begun.
The
Parkes work has been delayed by wet weather but is expected to be
completed early in January. The Mopra and Narrabri connections are in
progress and should be in place before the end of the year, with
testing and commissioning in January.

Contracts for lighting-up the fibre "backbone" had some delays, but now
a contract is in place with Cisco. The target date for completion is 30
January 2006 and work has already commenced.

Overall, it appears that the network should become operational around
February 2006. Tests are planned near the VLBI session in March.

b) Disk-based recording
New computer systems for VLBI recording have been acquired for all
antennas. These have been thoroughly tested and commissioned at most
antennas in time for the November 2005 VLBI session. The same systems
will be used for connections to the fast network when it becomes
available.

Tests are continuing to achieve 512 Mbps rates for a single DAS. This
will enable 512 Mbps operation at all LBA antennas and 1 Gbps operation
at the ATNF antennas. There are some problems with timing and
connectors, and a new system is under design.

c) Software correlators
Software correlators are under development by the Swinburne University
of Technology using their supercomputer cluster and a Cray computer at
UWA. Such correlators will be the only VLBI correlators for the disk-
based system and for high data rates, until the new CABB correlator at
Narrabri comes online in about 3 years.

The original XF software correlator has been operational at Swinburne
for about two years. A new FX software correlator has been implemented
recently and this can perform very close to real time! An FX software
correlator has also been tested on the Cray supercomputer at WA, using
FPGAs and the Opteron processors.

In the November 2006 VLBI session test observations were simultaneously
recorded on S2 tapes and disks. These will be correlated on the LBA S2
correlator, the FX software correlator on the Swinburne cluster and the
correlator on the Cray supercomputer, to provide an inter-comparison
test for all systems.


d) 4th e-VLBI workshop
The 4th international e-VLBI workshop was held at Marsfield 12-15 July
2005. About 50 people attended, 20 from overseas. It was a successful
workshop and in particular it facilitated cross-links with the
networking community. The workshop was followed by a 3-day tour of the
ATNF observatories and the participants were very enthusiastic. The
proceedings of this workshop are available on
(http://www.atnf.csiro.au/vlbi/evlbi2005/).