Marc's Blog

About Me

My name is Marc Brooker. I've been writing code, reading code, and living vicariously through computers for as long as I can remember. I like to build things that work. I also dabble in brewing, cooking and skiing.

I'm currently an engineer at Amazon Web Services (AWS) in Seattle, where I lead engineering on AWS Lambda and our other serverless products. Before that, I worked on EC2 and EBS.
All opinions are my own.


My Publications and Videos
@MarcJBrooker on Twitter

My Proposal for Arecibo: Drones

With apologies to real radio astronomers

Last night I finally got around to watching Grady Hillhouse's excellent video on the collapse of the Arecibo Telescope. At the end of Grady's video he says:

I hope eventually that they can replace the telescope with an instrument as futuristic and forward-looking as the Arecibo Telescope when first conceived.

I hope so too. While I've never worked in radio astronomy, my PhD supervisor and a number of my colleagues were involved in MeerKAT1 and HERA, so I developed a real interest in the field. I'd love to see radio, and radar, astronomy gain a great new instrument.

The Next Generation Arecibo Telescope white paper is one such proposal, written by a group a scientists who seem to know what they're talking about. Their concept is for an array of 1,112 closely-packed 9m dishes, either built in the same sinkhole as the original telescope, or elsewhere on the site. Just like with KAT and SKA, their proposal starts with a small array and builds up over time.

Some of the exciting new possibilities with this instrument include searching for pulsars orbiting Sgr A*, observing molecular lines from early Universe, climate change, ISR studies both parallel and perpendicular to the geomagnetic field, space debris characterization, ac-curate velocity measurements of a larger fraction of near earth objects, space weather forecasts, dark energy and dark matter studies, gravitational wave detection through pulsar timing, etc.

The whole thing seems smart, and sensible, but a little conservative.

So this is my semi-serious pitch for a sci-fi alternative: drones.

The expensive and hard part of building a telescope like Arecibo or FAST isn't the dish itself2, despite that being the large and iconic part. Instead, it's the platform and gear that houses the receivers (and, in the case of radar astronomy, the transmitters). That stuff is heavy, needs to be able to move around while being held firmly in position, creates a lot of data, and takes a lot of power. The dish itself is just a bunch of aluminium panels, mostly near the floor. What's inside the suspended structure is an antenna, or array of antennas, equipment to keep that antenna cool, equipment to receive signal from the antenna, digitize it, denoise and compress it, and get that data back to somewhere more sensible for further processing. It needs to be held quite still, way up in the air. To change where the telescope looks, you need to move the antenna. To focus it, you need to move the antenna. To change what frequency band you're working in, you need to change the antenna.

It's the same with radars: the traditional design is for one very complex and expensive receiver. Array telescopes change this up by having an array of dishes, and one slightly less complex receiver per dish, but the complexity is still quite high. Similarly, multistatic radars change it up by having a lot of receivers. What if we could go even further, and have a much higher number of much cheaper and less complex receivers? We could, in theory, build a radar or radio telescope that's orders of magnitude cheaper than what we're building today. Or orders of magnitude more capable for the same price.

If I have a really big hemispherical-ish dish, like what was at Arecibo, I can turn it into a telescope by having one big receiver in the air space above it, or by having many, many receivers in the same space. The advantage of one is that I get to make that one very good indeed. The advantage of many is that I get to sample the space at a lot of points, which (combined with fancy signal processing) gives me a huge amount of flexibility about where I look, and what energy I pay attention to. Plus, I get redundancy, which is nice.

People don't tend to build the "one dish, loads of receivers" kind of radio telescope, partially because the structure needed to keep tens of thousands of antennas floating in the air would be quite complex. All that structure would get into the way of the radio waves, and make the whole thing work much less well.

Which bring us to drone shows.

Drone shows, like the Intel Drone Light Show that featured at the opening ceremony of the Tokyo olympics. Or this record-breaking show:

Drone light shows put thousands of small things into the air, and get them to execute precise and complex coordinated movements. What if we put an antenna on each drone, and flew it above a huge reflective dish? By moving the drones around, we could sample the radio energy at thousands of locations above the dish, and send the data back for processing. With the right kind of processing, we could correct for errors in the drone's location and movement, and combine all their signals into a single view of what the telescope is seeing.

It's obviously hard, but I suspect it's possible. And likely way cheaper than the $454MM budget expected for NGAT. Cheap enough for an Olympic opening ceremony stunt.

What might make it impossible? Drones are quite unreliable, at least per-drone. They have limited battery life. Drones are quite electrically noisy. Drones don't hold station particularly well, certainly not well enough for long-term coherent sampling the the GHz range. It'd probably be hard to put a cryo-cooled antenna on a drone. But all those seem kinda surmountable. If there's one thing that would kill the idea, it's probably noise. There's not a lot of energy in those radio signals from the sky, so they are easy to drown out3. But maybe we could shove the electrical noise into the bands we're not interested in.

So, NSF, how about replacing Arecibo with a drone array?


  1. MeerKAT is not only a great scientific instrument, but also one of the greatest puns of all time.
  2. On the other hand, if you're building a big moving dish like DSS51 then the structure is very tricky, both in just getting it to stand up, and getting it stiff enough that it doesn't go all fuzzy when the wind blows or the sun shines. One interesting relationship is that HartRAO is where it is fro the same reasons Arecibo is where it is.
  3. One of the NGAT proposal's advantages over Arecibo is "Capable of mitigating radio frequency interference (RFI) through phased nulling". What they mean here is that they can use the fact that it's an array rather than a single antenna to better reject nearby sources of RFI, by steering a null in the direction of the noise source. The more antennas you have, the easier that is to do.