Data speeds power up MWA radio telescope research
- 10 December, 2012 15:19
The Murchison Widefield Array (MWA) project in Western Australia.
Data processing rates of up to 4 gigabits per second will be realised next year when the Murchison Widefield Array (MWA) radio telescope correlator ramps up from February 2013.
The $51 million MWA project in Western Australia is an international collaboration between 13 universities across Australia, India, New Zealand and the US to construct a low frequency radio telescope as a precursor to the Square Kilometre Array (SKA), the world’s largest radio telescope.
MWA director, Professor Steven Tingay, told CIO Australia that data generated by the MWA correlator is being sent down a 10 gigabit per second line to Perth where the data is aggregated into an archive. Researchers will use the MWA project data for different areas of astronomy research.
“The first is to look back into the early stages of the evolution of the universe when the first stars and galaxies were being formed,” he said.
The second area of MWA is doing detailed studies of galaxies while the third area of research will identify the trajectory of solar storms, quadrupling the warning period currently provided by near-earth satellites.
Tingay added that the MWA radio telescope will have an operational life of five to 10 years as once the SKA radio telescope is built, it will be wound down.
As part of the MWA project, Cisco was contracted to provide 10 gigabit Ethernet networking switches, a unified computing system (UCS) and Nexus data centre series switches.
These form part of the MWA correlator which is aggregating and processing volumes of raw radio signals for analysis by Australian radio astronomy experts.
According to Tingay, the vendor was selected because off-the-shelf technology now meets the requirements of radio astronomy.
“It’s often the case that you can go out and buy something rather than go through a lengthy design and prototype cycle to do it yourself,” he said.
“We now do lots of the central signal processing using enterprise servers and central processing unit [CPU] platforms.”
Follow Hamish Barwick on Twitter: @HamishBarwick