Musk is messing with Cosmic Dawn. Will alien hunters save the day for mankind?

Opinion Elon Musk's Starlink flying circus is many things. It's a multi-thousand satellite global internet provider, growing by hundreds of new orbiting relays a month. It's part of intricate geopolitical power games between the Pentagon, the US government, Ukraine, and Musk himself.

It loves photobombing astronomy photographs, as the satellites drift across fields of view that were intended to snap distant starlight. But for its latest trick, it's humming loudly over the faintest whispers from the dawn of the universe. 

The latest report from the wizards of Oz at Perth, Western Australia's International Centre for Radio Astronomy Research (ICRAR), is not pretty. A new multi-billion dollar, multi-year international radio observatory called SKA-Low – a Square Kilometre Array planned to cover the frequency range from 50 to 350 MHz – is being built. Its purpose is to catch the radio signals sent 13-plus billion years ago as the Universe cooled from the Big Bang and recognizable states of matter started to appear. 

By the time those signals get here, stretched out by cosmic expansion over all that time, they've gone from the ultra-high frequencies of atomic physics to good old fashioned VHF, home of two-way radios and FM broadcasts. Western Australia is mostly emptiness and sheep, neither of whom have much use for either. So that's a good place to build a VHF radio observatory. You'd think. 

It's always better to check – one of science's better ideas, that. The SKA-Low team built a prototype instrument called EDA2 and ran some tests. As reported in Space Australia, they found unquiet skies. Over 29 days and 78 million full-sky images later, they found more than 112,000 signals from 1,506 different Starlink satellites. These could be five orders of magnitude stronger than target signals. 

This isn't supposed to happen. Radio astronomy has its own protected bands, and satellites have to conform to the frequencies allocated to them as well. On a regulator's spectrum map, it looks lovely, with sharp black lines cleanly separating each Hertzian fiefdom. Reality is nothing like that. 

Astronomers classify signals from satellites into the intended and the unintended. Regulators and satellite radio engineers concentrate on the intended, those used to communicate with the ground. Those will generate some signals outside their nominal bandwidth, because reality hasn't read the theory, but this is part of radio engineering and is taken into account by everyone. Unintended emissions are very different. 

If you take an AM radio and hold it close to your computer, you'll hear a lot of noises Guglielmo Marconi never intended. That's physics again, as inside your computer innumerable signals are switching on and off very fast. That's what makes radio waves. Inside a Starlink satellite are very many very fast, very switchy signals. In your computer, there are various kinds of shielding to keep the signals in; and the same is true of Starlink. There are normal limits that normally keep normal things working normally. 

Only Starlink is not normal – there are thousands and thousands of them. And radio observatories are not normal, they're designed to pick up the tiniest of signals there are from the furthest away that there is. Starlink is that tinted-windowed over-specced car, blowing out windows with its bass. 

Away from the perils of SKA-Low, things aren't so great in the intended signals segment either. The direct-to-cell service is a real doozie. With ordinary Starlink broadband, the customer terminal has a high-gain antenna that can form concentrated beams and track its orbiting target to both receive and transmit. This means the satellite doesn't have to work too hard to receive, nor use very high powers to transmit. With direct-to-cell, the ground terminal is the mobile's tiny internal antenna system and battery powered transmitter, designed to work with cell towers a kilometer or so away at worst. 

The Starlink satellite is 500km away at best. Here's where the gods of radio physics get nasty: the inverse square law says that signals get weaker as a square of the distance they travel. A signal 500 km away isn't 500 times weaker than one at one kilometer, it's 250,000 times weaker. The satellite's transmitter has to compensate by upping its power accordingly. This gets beamed straight back to Earth, to the cellphone user and, oh, any exquisitely sensitive radio telescope straining to hear the worst of all weak signals. 

It's like having a giant searchlight in the night sky pointed directly to you, only there are lots of them and more of them each week. There's a reason radio telescopes live in radio quiet zones, where visitors have to turn all their devices off and Wi-Fi is as welcome as a fart in a spacesuit. But there's not much point if Musk is running strings of ultra-bright radio emitters across the sky like Christmas lights in suburbia. 

Even when the signals being scanned by the telescopes are far away from those of the satellite transmitters, there are problems. Normally, the energy from signals elsewhere in the spectrum is removed by antenna design and electronic filtering. Radio telescope detectors, however, are very sensitive, and can be overwhelmed very easily. As electronic filtering can only be added after the sensor has turned radio signal into electrical signal, there's not much to be done. 

Here's the good news. SpaceX is extraordinarily responsive to these problems. Some of this is the RF Priesthood. Radio frequency engineering is seen as a dark art, and those who practice it have a natural bond. Some of it is the SpaceX culture of moving fast and breaking things, and the bigger the thing broken — like the dawn of creation – the faster people move. Finally, given the launch cadence, SpaceX engineers can get a modification into a satellite and test in orbit faster than it takes the Pentagon to sign off on a screwdriver.

• Starlink says SpaceX targeting 2026 for launch of Starship-ready terabit satellites
• Bezos beams up batch two as Project Kuiper plays catch-up with Starlink
• Eviden unveils satellite monitoring tool, as Starlink asks UK for E band access
• SpaceX resets 'Days Since Last Starship Explosion' counter to zero, again

Here's the better news. SpaceX has teamed up with the SETI institute to form an industry group that will focus on these issues and start to formalize the matter as an engineering disciple. If you don't know the SETI  Institute, you should: Started in the early 1980s to encourage the scientific hunt for aliens when expressing an interest for the search for extraterrestrial intelligence could be a career-limiting move for a scientist, it has become beyond respectable now that everyone is feverishly searching for life signals in exoplanetary atmospheres. 

This alliance is desperately needed. Swarms of satellites are pouring ever more powerful, ever more spectrum-gobbling signals at the Earth, while radio observatories are chasing more and more signals across more and more spectrum at ever higher sensitivities. It remains to be seen whether or not all players will participate. Many have a history of not respecting all concerns that might get in the way of their intentions, so building strong, responsible institutions now is essential. After all, if you can get Musk on the side of the angels, you have some pretty cosmic powers yourself. ®