[Image: D. Smyth, CSIRO]
Hopes are running high for the Square Kilometer Array (SKA), an audacious plan to build the world’s largest radio telescope by linking together thousands of widely dispersed individual radio astronomy dishes in Australia and South Africa. Making the plan operational, however, requires a stable reference frequency standard shared by the widely dispersed telescopes, to allow the signals from the facilities to be aligned in time and combined coherently.
A team of scientists in Australia has now demonstrated an approach for reliably transmitting such a frequency reference across hundreds of kilometers of installed, standard telecom fiber—even when the fiber is simultaneously transmitting live telecom traffic (Optica, doi: 10.1364/OPTICA.5.000138).
The need for a stable reference
The SKA idea rests on the technique of very-long-baseline interferometry (VLBI). In VLBI, numerous, geographically separated radio astronomy facilities train their scopes on the same patch of sky at the same time. The time-correlated signals are then interferometrically combined, to provide an angular resolution far superior to that of a single, much smaller individual radio dish. The SKA, for example, will provide a “virtual dish” with a total collecting area of around a million square meters, allowing it to peer into radio signals with a sensitivity 50 times greater than that of the Hubble Space Telescope in the optical domain.
To make the system work, the radio signal from each telescope is downconverted and sampled, using a local oscillator that’s frequency-referenced to a highly precise atomic clock, for time control. Heretofore, radio telescopes have been equipped with their own atomic clocks to provide that stable reference frequency.
But there are big disadvantages to that multi-clock approach in a VLBI setup. One is cost—maintaining an atomic clock at a given telescope facility has a cost on the order of US$200,000. Another disadvantage is that the system at large depends on all of the clocks having superb long-term stability; drift or losses of coherence between clocks at separate facilities can degrade the interferometric signal.
A “real-world” fiber solution
A better solution would be to find a way to send a frequency reference from a single atomic clock to multiple radio telescopes via optical fiber. Numerous research groups have taken on the problem, with some impressive results. In a test published a year ago (Sci. Rep., doi: 10.1038/srep40992), for example, a European research group succeeded in transferring a remote atomic-clock reference 550 km over optical fiber to a single antenna at a radio telescope facility in Italy. This and other long-distance demonstrations, however, have involved sending the reference frequency from a clock to a single antenna, rather than between two separate radio telescopes. And these demos have used dedicated fiber for the project rather than “real-world” fiber in the installed telecom network.
In the new work in Optica, the Australian team used a setup involving two telescopes in the CSIRO Australia Telescope Compact Array (ATCA), separated by 155 km of standard “real-world” telecom fiber carrying regular traffic, for a 310-km round-trip distance. One telescope facility included a hydrogen maser atomic clock that provided a master radio frequency signal, which was frequency mixed with a master laser. The other, remote facility used a high-quality quartz local oscillator to create a slave radio frequency, which was mixed with a slave laser.
In the setup, the two laser beams counter-propagate through the optical fiber, and each is received by a photodetector in the other facility. The received signal from the slave laser at the master facility includes both the known slave oscillator frequency and a term related to the noise in the fiber attributable to temperature, vibrations and other environmental issues.
With that information in hand, the signal of the master laser can be amplitude-modulated to algebraically subtract out the fiber noise term. So modulated, the reference frequency from the hydrogen maser can then be transmitted free of fiber noise from the master to the remote location, to lock the slave frequency to that of the master facility.
More stable than Earth’s atmosphere
Using the stable frequency reference provided by the hydrogen maser at one facility to calibrate both telescopes, the researchers were able to establish that the relative frequency stability exceeded that obtained with separate hydrogen masers at the two facilities. Indeed, the team found that the frequency stability transmitted over fiber was “significantly better” than the error introduced by atmospheric perturbations between the two telescopes.
The researchers believe that the system—which does not require changes to the rest of the fiber network and can be implemented relatively easily—could prove considerably more cost-effective for a big VLBI project such as the SKA than maintaining an atomic clock at each facility. The approach, according to OSA Fellow Kenneth Baldwin of Australia National University, a coauthor on the study, “allows an atomic clock, which costs around two hundred thousand dollars, to be replaced by a system that costs only a few tens of thousand dollars.”
The team also believes that, beyond VLBI, the system could find use in transmitting frequency standards “for other applications such as the precise calibration of environmental, industrial, and laboratory-based molecular-spectroscopic sensing.”
The research consortium involved in the work included Australia’s Academic and Research Network (AARNet), the Australian National University, the Commonwealth Scientific and Industrial Research Organisation (CSIRO), the National Measurement Institute, Macquarie University and the University of Adelaide.