Interplanetary network: It's the data, stupid!

DW: What's the state of space communications today?

Matthew Cosby: At the moment, communications use point-to-point links. So there's a manager on the ground who schedules link sessions with a spacecraft, whether it's around the moon, Mars or the Earth, and it's all done "mandrolically" [a long, tedious task performed manually - the Eds.] With networks around Mars, it's done the same way, where you've got the rovers communicating to orbiters, and the orbiters communicate to Earth. They are all ad-hoc networks. They're either managed in Germany at ESOC in Darmstadt, or by NASA. 

And the vision is to try to deploy a space internet - an interplanetary internet - that replicates the terrestrial internet, where you don't need to do this mandrolically. Instead, it would all be automatic.

But there are fundamental differences between the terrestrial internet and space internet. With the terrestrial internet, most of the infrastructure is there, and it's error free. And in space that's not the case. Sometimes you lose connectivity.

What are the issues in space?

The issues are long delay times - so-called light-times - and the environment has "burst errors," it is not error-free. We do modulations and coding to ensure you get an error-free transfer, but there are delays and you still get errors. Take, for instance, rovers and orbiters. An orbiter will come over the horizon and the link improves, and then degrades again as the orbiter goes over the horizon, and your network is gone.

Now what we're working on at the moment is to try to cope with "delay tolerance." That's where the links appear, disappear and reappear, but you're still able to transfer your image. It's called "disruptive tolerant networking."

Usually if a link disappears, everything stops. You wait until the link comes back, and then you have to start all over again.

With delay tolerant networking, you try to cope with any point - or node - in the connection disappearing at any time. The data is stored and when the link comes back, or there's a new link, your data transfer continues. It's better suited for the space environment.

So do we need new technology to make this work, or is there existing technology?

No, we have existing technology. We just have to move away from the point-to-point, mandrolic way of doing it. The software and hardware does exist, and it's being tested - it's been tested on the [International] Space Station. If that is adopted internationally - if a private company, or America, Europe, China, India, or whoever sends something to the moon, and they agree to take the equivalent of a WiFi router or something, then every spacecraft builds up that network - and it becomes a standard, and once it's a standard, people will use it more.

What's the object of having an interplanetary internet? Is it for FaceTiming with people in the Mars colony … ?

Ultimately, everything is about data. When you do a scientific experiment, you generate data, and that's only going to increase. So you've got to have more efficient ways of transferring the data. I think FaceTiming from the surface of Mars is a long way off, and probably the need for it is also a long way off. The light-time from Mars varies from 8 minutes to 40 minutes so your FaceTime is going to be particularly troublesome at that point!

But transferring data files, more intricate data files from Mars or the moon, is what we're going to need going forward. And in ten years, we could, in theory, have private people wandering around on the moon. Seven years ago people thought that having commercial re-supply missions to the ISS was a fantasy, and now it's routine.

How big a priority is an interplanetary internet? Where does it fit in with all the other priorities - Earth observation, reusable rockets, human spaceflight, small satellites …

Well, communications tend to be the last thing to be thought about when people design missions. It's always assumed it's there. And hopefully we'll get to a situation where the network is there and you can just assume it's there. But it is going to be a high priority for the people developing the communications side of things, because scientists and engineers want to get more and more data and we have to build more infrastructure to get that data back.

Matthew Cosby is the chief scientist at Goonhilly Earth Station in the UK. He also represents the UK Space Agency at the Operations Advisory Group (IOAG) and Consultative Committee for Space Data Systems (CCSDS) on space communication standards. The interview was conducted at the 2017 UK Space Conference in Manchester.