The Square Kilometre Array (SKA) telescope project is sure to stretch local definitions of broadband should the project go ahead.
The project will link together thousands of radio telescopes to build a virtual telescope with a collecting area of one million square metres. Short-haul links across the inner array will need a capacity of 80Gbit/s per antenna, and long-haul links will need a capacity of 2Tbit/s per station.
High performance central computing engines — capable of 10 to the power of 15 operations per second — will be required, as well as world-wide transfers of 1TB of images every minute.
As a part of the Square Kilometre Array SKA project, two radio telescope stations could be sited in New Zealand. This would bring in opportunities for both New Zealand IT and engineering companies, says Sergei Gulyaev, director for the Centre for Radiophysics and Space Research at Auckland University of Technology.
The CRSR is part of the international consortium – with representatives from 18 countries – behind the project, which aims to build the world’s biggest radio telescope. The SKA will have a collecting area of one million square metres.
The project is based on interferometry, which means that signals received from a large number of widely spaced instruments can be combined digitally at a central processing site, using a technique called aperture synthesis. This will simulate a telescope with a diameter equal to the largest antenna separation, which will be more than 3,000 kilometres for the project. Hundreds of small stations will collect data to create a big picture, covering frequencies of between 0.1GHz and 25 GHz.
One of the proposed sites for the SKA is Australia and New Zealand. The core, which would feature an array, measuring five-kilometres of 1,000 dishes would be sited just outside Perth in Western Australia, according to Gulyaev. A hundred stations, with 12-20 smaller dishes each, would then spiral out from this core. At least two of these smaller stations would be sited in New Zealand. One of these would most likely be sited the North Island, and the other one the South Island, he says. Four sites are currently being investigated, including Warkworth, Ardmore, Rangiora and Awarua.
“The construction of SKA will start in 2011,” says Gulyaev. “[The consortium] is talking about building the stations in New Zealand [if selected] in phase one.”
The other proposed sites are in Argentina, South Africa and China. The decision on which site will be chosen will be made by the participating countries’ governments in 2007-2008. However, the optimistic Australians have already started building the core in the desert outside Perth, says Gulyaev.
Recent innovations in computing and radio frequency devices will enable the project to be completed before 2020, says Gulyaev.
“The SKA will be 100 times more powerful than today’s telescopes,” he says.
“It will be 50 times more sensitive and will be able to survey the sky 10,000 times faster.”
A spatial resolution of one mas (milliarcsecond) has already been achieved, says Gulyaev.
The SKA will be capable of addressing previously unanswered questions, says Gulyaev. For example, it will enable tests of general relativity with pulsars and black holes to be conducted. It could also be used to explore the question of how the galaxy evolved, as well as questions about dark energy, cosmic magnetism, the possibility of extra solar planets and, of course, whether there is extra terrestrial intelligence, says Gulyaev.
The SKA is sensitive enough to search for signals no stronger than those generated for television.
Extremely powerful broadband will be required to transport data from the antennas to the central processor. Short-haul links in the inner array will need a capacity of 80 Gbit/s per antenna, and long-haul links will need a capacity of 2 Tbit/s per station. High performance central computing engines – capable of 10 to the power of 15 operations per second – will be required, as well as world-wide transfers of 1 Tbyte images every minute. Gulyaev says that the processing and post-processing of the project will rely heavily on Moore’s law of the doubling of computer processing power every 18 months.
“Archiving and sharing the data will be a major challenge,” he says.
The budget for the project is one billion euros. Europe and the US will put in 30% each, Australia and New Zealand will contribute 10%, while the remaining countries will pay the remaining 30%.
“Ten million euros for a period of ten years will be expected from New Zealand if it is selected,” says Gulyaev. On the other hand, 999 million euro a year will come from overseas, he points out.
The planned stations, if they are sited in NZ, would have 12-20 antennas, each 10 or 15 metres in diameter. One station would be equivalent to a 100-metre diameter radio telescope.
“It would be huge for New Zealand to be involved in this project,” says Gulyaev.
There is an enormous potential for New Zealand companies to be involved in, for example, software development, data processing and data infrastructure around the project, says Tim Natusch, senior lecturer at the CRSR.
There would also be opportunities in areas like civil engineering, networking and timing systems, says Natusch. He adds that, if telescopes are wired, synchronisation of clocks is easy. But, clearly, it is impossible to wire all the telescopes in the SKA project.
“An absurdly high level of stability of clocks is required,” says Natusch. “And a hundred different stations will need to be working in synchronisation with that accuracy. There is scope for all kinds of clever inventions in that area.”
He thinks even small engineering firms have the opportunity to garner a piece of the SKA pie should Australia and New Zealand be selected for the project. “[And] there is a good case for money to be spent in Australia and New Zealand”, says Natusch.