Interview Scientists and engineers have been taken aback by the amount of radio interference generated by satellite constellations, and many are calling on standards bodies to improve operator performance.

The photobombing of ground- and space-based telescopes by proliferating satellites in Low Earth Orbit has long vexed astronomers. As well as optical annoyances, radio astronomers are also encountering interference due to satellites communicating with ground stations.

In an interview with The Register, Tudor Williams, CTO of high-frequency RF communication company Filtronic, explained the problem, which is mainly related to satellite-to-ground transmissions (many large constellations, such as SpaceX's Starlink, use optical links for satellite-to-satellite communication, which don't cause the same issues.)

According to Williams, the problem comes from the side lobes of poorly designed antennas, where signals are unintentionally spread. The effect can be bands used for communications overlapping with observation bands, causing headaches for radio astronomers.

"In a badly designed antenna," Williams explains, "you get fairly strong lobes of signals that get sent in different directions."

But surely licensing and regulations should ensure this isn't an issue?

Yes, but the issue, according to Williams, is that until the satellites began launching, the issue had not been fully understood or anticipated from ground testing. "Maybe the regulations weren't as stringent as they should have been at the start," he says, "And as we start to get more data on this, we can obviously have more stringent guard bands and more stringent regulations going forward."

A "guard band" is a slice of radio spectrum between active bands that acts as a buffer to prevent signals from leaking from one channel into another.

"The designers," he says, "do the best they can, but sometimes there are errors in the process which mean leaks into the spectrum … the authorities will define how much interference can be tolerated."

As the problem becomes better understood and companies like SpaceX work with scientists and regulators to devise improved hardware, the situation should start being mitigated. This, however, is small comfort to astronomers dealing with thousands of new sources of interference.

Williams acknowledges that retrofitting hardware to existing satellites is not a viable solution, but says, " They can do updates to things like the waveforms. The type of waveforms you've got through the amplifier has quite a big impact on spectral spreading. It'll stress the amplifiers in different ways.

"So you potentially could optimize the waveforms, using AI or whatever to actually control that and try and cause less distortion with even existing hardware.

"So you could potentially solve it from a baseband perspective."

It isn't only altruism that is encouraging good behavior from operators. "The more we see this happening," explains Williams, "the more that will push back into satellite operators because they'll be under pressure to behave. Because they have to control their spectrum.

"Everyone understands the problem. Everyone's aware of the problem. It's just… how bad is it? What problems is it causing? What do we need to do to be better?"

• Square Kilometre Array is so sensitive, its datacenter needs two Faraday cages to stop RF leaks
• Musk is messing with the Cosmic Dawn. Will alien hunters save the day for all mankind?
• DARPA skips the lab, will head to orbit to test space manufacturing tech
• Brit soldiers tune radio waves to fry drone swarms for pennies

Williams points out that satellite operators need their licenses, which will be tricky to renew if they behave badly and stomp over places where their signals don't belong. "I think," says Williams, "we need to have a co-existence of scientific missions and general broadband comms."

Starlink satellites, for example, are designed for a relatively short operational lifespan of around five years. As the interference problem becomes better understood and regulations tighten, there is an opportunity for hardware changes. "Would there be cost implications if you have to control it [the signal] better?" muses Williams. "Yes. Probably. Because it makes the design more complicated … volume manufacturing and lots of satellites can add quite a lot of cost in the process."

Ultimately, while the light pollution of the night sky by the rapidly growing constellations of satellites is visible even to the naked eye, the scale of RF interference has been unexpected by many. According to Williams, "feedback into standards bodies will be pushed out to the satellite operators, and they'll have to have better performance systems going forward." ®