In the latest Quest

Posted by admin in Science in Society, Space, Technology

Radio frequency interference and radio astronomy: why the fuss?

The radio telescope arrays that are to be build in the Karoo as part of MeerKAT and the SKA will be among the most sensitive telescopes ever built. Richard Lord explains the significance of radio frequency interference.

The Karoo Array Telescope (KAT) is an initiative of the South African government to further world-class scientific research. In the next few years, an array of 60 – 80 antennas will be built in the Karoo. This radio telescope array is called MeerKAT, and it will be among the most sensitive telescopes ever built.

It is no secret that radio frequency interference (RFI) and radio astronomy are not the best of friends. However, the extent of their animosity is often suppressed or mis-understood.

The famous Jansky

Karl Guthe Jansky was an American physicist and radio engineer who first discovered radio waves coming from the Milky Way in August 1931. He is regarded as one of the founding figures in radio astronomy.

The non-SI unit used by radio astronomers for the strength of radio sources is the Jansky, defined as:

1 Jy is equivalent to 1.10^-26[W/m²/Hz]

Where W=Watts

Hz=Hertz

Some of the strongest radio sources at 1 420 MHz are:

* Our Sun: ~ 700 kJy

* Virgo A: 201.81 Jy

* Hercules A: 46.52 Jy

* Hydra A: 42.37 Jy

A drop in the ocean

It is difficult to understand just how small a Jansky is. Let’s do a simple calculation. Imagine we had the luxury of pointing the 26 m HartRAO dish (assume 70% efficiency) in South Africa to one of the brightest radio sources out there, namely Virgo A. Now assume that we could do so for the entire lifetime of the dish, approximated to be 30 years.

Every day, we would be able to receive radiation from Virgo A for a maximum of 12 hours (since the Earth rotates, the source would be hidden below the horizon for the other 12 hours). Furthermore, let’s assume we have an instantaneous collecting bandwidth of 256 MHz.

Leaving out the maths of the calculation (just trust me) the total energy collected over this timespan from Virgo A would be 25nWh – or 25 nano-Watt hours. To put this into perspective, a 100 W lightbulb, switched on for one second, uses a whoppping 28 milli-Watt-hours of energy – more than a million times what our 26 m dish recieves from Virgo A in 30 years.

And just as an aside, radio astronomers are not really interested in these strong, well-researched sources. The sources they are interested in are in the milli-Jansky range, and for the SKA, in the micro-Jansky range and less.

Landing on the Moon

One of NASA’s goals is to return humans to the moon by 2020. Will they carry cellphones? A typical cellphone has the following radiation characteristics:

• Transmit power: 0.5 W

• Bandwidth: 12 kHz (The GSM standard provides for a channel spacing of 200 kHz, with each channel having approximately a 12 kHz bandwidth.)

The average distance from the Earth to the Moon is about 380 thousand kilometres. We can now calculate the approximate spectral power flux density (another way of saying RFI) as seen on Earth, assuming that the cellphone transmits isotropically, i.e. equally strongly in every direction (let’s look at the maths this time):

^Scellphone = Pt/4πR²B

= 0.5/4π (380•10⁶)²•12•10³

= 2.3•10^-23 [W/m²/Hz]

= 2 300 [Jy]

In other words, a radiating cellphone on the Moon will appear to be ten times brighter than Virgo A, which is already one of the brightest galactic sources out there.

What about nylon chairs?

Nylon chairs can accumulate static charge, leading to static discharges. During the RFI measurement campaign in the Karoo, the RFI measurement system (which is much less sensitive than a radio telescope) picked up these static discharges, and the nylon chairs were subsequently replaced with canvas equivalents.

Nylon chairs weren’t the only culprits. A portable gazebo, used to provide protection for vehicles, consisted of steel pipes strung together by a combination of metal chain and elastic cord that was threaded through the pipes. It was found that the contact of the links of the metal chain with the inside of the pipes and with each other caused RFI, especially when the wind made the frame shake. The gazebo had to be removed from the site.

So where do we go?

It seems RFI is everywhere. Even the Antarctic isn’t immune from RFI from satellites. Then there are airplanes, criss-crossing the skys while radiating merrily. There are very, very few places left on Earth that are radio quiet. One such place is the Karoo. It is the (relative) lack of RFI, more than anything else, that has led to the decision to build

a radio telescope array in such a deserted place. Not because there is no light-pollution from nearby cities (only optical telescopes care about light-pollution). Not because it is cheaper to build (due to the lack of roads, power lines and other infrastructure, it will be much more expensive). Not because we love the Karoo (we do, but it is not the reason). No, the reason is the (relative) lack of RFI.

Keeping it clean

There are many reasons why it is important to devote so much effort to keeping the Karoo site as RFI-clean as possible, some of which are:

1 The receiver has to remain linear, i.e. there may not be any amplifier saturation. A strong RFI source, even if it only occupies a tiny fraction of the spectrum, will corrupt the data across the whole spectrum if the amplifiers are saturated.

2 The receiver may not introduce any inter-modulation distortions. Therefore strong sources outside the bandwidth also need to be regulated.

3 The above two issues require RFI transmissions to be reduced, even if they cannot be removed entirely, i.e. even if they are still detected by the radio telescope and (hopefully) flagged by the processing software. However, it is also feasible that some RFI transmissions can be avoided altogether, or shielded to levels below the radio telescope

sensitivity. This should be the aim wherever possible.

4 The radio astronomy protected bands must remain completely RFI-free. They have been hard-fought for, and any device emitting inside these bands, or producing harmonics or other non-intentional emissions inside these bands, must be removed or shielded.

5 The more RFI needs to be mitigated by the processing software, the more expensive the processing hardware becomes, and the more development time is needed for writing this software.

Keeping the site RFI-clean is not a one-man show. It requires a consistent and concerted effort that spans many disciplines,

for example:

• Political (e.g. instituting a ‘Radio Quiet Zone’)

• Infrastructure (e.g. power lines)

• Ancillary equipment (e.g. shielding of buildings and other equipment)

• Logistical (e.g. switching off cellphones and other radiating devices)

• EMC (e.g. shielding noisy equipment, implementing good design principles, testing in anechoic chambers)

• Monitoring and policing (e.g. illegal RFI sources need to be located and removed)

• Lightning protection

Threshold levels

OK, so we shield everything that radiates. Although a noble idea, the required levels of shielding are often prohibitively expensive and very difficult to achieve. The ITU Recommendations give firm guidelines regarding the threshold levels of harmful RFI radiation. These levels depend on the sensitivity of the radio telescope, which is a function of the telescope’s noise temperature (Tsys), instantaneous bandwidth and total integration time. Any device emitting more than the recommended threshold should be sufficiently shielded. Wherever possible, the design should make use of natural

attenuators, e.g. using mountains to shield antennas from equipment.

Epilogue

One of the science drivers for MeerKAT is to observe new phenomenon related to the transient radio sky, i.e. to discover sources that vary in brightness on time scales of  seconds to years. Many of these transient events originate in processes involving ultramassive black holes in the centres of galaxies, distant supernova explosions near the edge of the observable universe, and gamma-ray bursts, the most energetic events in the cosmos. The stars linked to these events could be billions of light years away. This means the light from them took that many years to reach us. The Earth itself is about four billion

years old, so some of these events occurred before our planet even existed, before the first microbes formed, before the oceans had formed.

Is it not tragic then, that these photons, after travelling for so many years through the universe, should fall prey to RFI in their last few milliseconds of existence. Forever lost. Undetectable in a sea of man-made interference.

Article based on a technical report prepared for the National Research Foundation.

Richard Lord completed his PhD in the field of radar remote sensing at the University of Cape Town in 2000. After completing post-doctoral research at the German Aerospace Centre in Munich, Germany, he returned to UCT where he worked as a Research Officer in the radar remote sensing group. He joined the SKA SA organisation in 2007, where he worked as a software specialist in the Computing Team during that year.  In 2008, he joined the Systems Engineering Team, and is currently the project manager and systems engineer for KAT-7.