Techscape: Vint Cerf on the Interplanet

The Register , 31.08.05

By Bill Robinson

Interview In the second of TechScape's three exclusive interviews with Vint Cerf, Bill Robinson investigates the story of the InterPlanet (or Interplanetary Internet as it is sometimes called).

The first time I met Vint Cerf was about five years ago when he was speaking at Cambridge University . Being a Yank over here, I decided I should hear him speak, particularly given the topic: The InterPlaNet.

What, I wondered, was this guy smoking? It all sounded so bizarre, so surreal. An internet out in space?

I told him years later of my initial cynicism, and he replied simply: “Yes, I get a lot of that.”

When he first began speaking at Cambridge that day, there was an auditorium full of skeptics like me who wanted to see a scientist meltdown publicly.

But no such meltdown was forthcoming, only a man articulately describing his vision for projecting a communications network into outer space.

“In late 1997,” Cerf began during his recent interview with TechScape, “I began thinking about how the internet started 25 years earlier and what I could get working on that might be important 25 years later.”

Hmmm, this is already beginning to get scary, I thought, recalling an interview I did with interview of Sir Arthur C. Clarke. Here, Cerf is trying not only to predict the future as Sir Arthur would undoubtedly try to do, but also how he could contribute to it.

“My mind wandered into the possibility of expanding the internet out into the solar system,” Cerf boldly went where no other man has gone. He knew an engineer working at JPL (Jet Propulsion Labs in Pasadena , California ) which is a close partner of NASA's in all of its space missions and this man was a very willing and enthusiastic partner for laying out the blueprint for this extraterrestrial creation.

And the InterPlaNet's visionaries knew that NASA was sending missions to Mars. This seemed like a natural opportunity to begin laying the railroad tracks, infrastructure or IP backbone for the solar system's first human communications system.

But where to start technologically?

Initially, “flow control”, or the time it takes for a signal to reach and return from an exploratory mission in space, was Cerf's challenge to overcome. If he could reduce or minimize this crucial “downtime” period, he could vastly improve the efficiency of interplanetary communications just as he did with the internet and intra-planetary communications here on Earth.

Also essential, surprisingly, were the availability of domain names for the space communications network. “At the outset, we thought the domain-name system may not work,” Cerf said, “but we realized that not using it meant one might spend 40 minutes looking up an IP address for a certain location in space. We didn't want that. We also worried that due to the popularity of the internet, Earth may use up all possible domain names. This is why ICANN set aside the .mars and other planetary domains for this initiative.”

The next key issue was satellites and how to position them.

The time astronomical and physics complexities of sending data (and instructions) to and from Mars are infinite and most often involve reception-quality issues such as we Earthlings experience terrestrially on our mobile phones. “If you're transmitting data to Mars, Mars is spinning of course. If your receptor is on the far side, you have to wait until it comes back to the close side again; or have a satellite to bounce it back from the dark side,” Cerf declared.

The challenge, according to Cerf, was an extension of communications reception problems here amongst us humans. “Think about mobile operation and connectivity here which is not great - when we go under a bridge for instance.” Going under a bridge is one thing; trying to call Mum from Mars another thing entirely.

“It can take 40 minutes round trip for data to/from Mars,” Cerf said, “and there's a minimum of six minutes even when Mars is at its closest to Earth.” So any reduction in wait time can be important particularly if you're remote-controlling a Mars rover.

Imagine sending an instruction through the joystick only to find out 12 minutes later that you've just sent your $60m Mars rover into a two-mile-deep Martian chasm.

“To overcome some of these issues, we created a new protocol we called Delay & Disruption-Tolerant Networking (DDTN) which we believed would offer ways around some of these communications problems,” Cerf explained.

Shortly thereafter, Cerf contacted his old friends at DARPA (Defense Advanced Research Projects Agency). He told them of his idea to create an interplanetary communications system and they sat up and took notice.

Of course they were motivated by a rather intense interest in the military, battlefield and tactical applications of such a communications technology; just as the internet held their attention for many of the same reasons.

DARPA funded the initial research into the InterPlaNet to the tune of approximately $500,000.

Cerf credits many people and initiatives for contributing to this development including The Center for Embedded Network Sensing at UCLA which is headed by Deborah Estrin a renowned figure in Silicon Valley networking circles; and some joint work done at UC Berkeley in cooperation with Intel.

Meanwhile, Cerf and crew have been looking at more Earth-bound ways to generate solutions to cosmic complications; low satellite angles to connect, for instance. Even precipitation can cause a problem when connecting. “I've spent some time above the Arctic Circle recently,” Cerf reveals.

It turns out that the areas around the North Pole have good potential for addressing the low satellite angle issue (not to mention the “precipitation problem”) and provided a good test bed for Cerf.

“I was working with the Sami people, a group of indigenous reindeer herders in Lapland .”

Lapland extends from northern Norway to the Kola Peninsula in Russia . The Sami are nomadic people who are always on the move; consequently there are few Sami towns but rather communities which are highly-mobile. Estimated Sami population figures are: 25,000 in Norway , 17,000 in Sweden , 4,000 in Finland and 2,000 in Russia .

Working with the University of Lulea in northern Sweden , Cerf installs laptops with 802.11 capabilities on their snowmobiles. I had to chuckle at the image of the Samis cruising around the tundra with better, more consistent access to a connection than I enjoy in London .

“Testing on Earth,” Cerf avows, “is so important, before we send something out into space which cannot be modified later.” Good point.

So Cerf's InterPlaNet and DDTN protocol has now evolved into an integral part of the communications system for the International Space Station and the Space Shuttle while the orbiter system is in full-swing and rapidly expanding. “There will be three orbiters around Mars,” Cerf proudly exclaimed, “the last one just launched and will take seven to eight months to get there.”

In conclusion, what I thought was a pipe-dream several years ago has now turned into reality with the subsequent Mars missions by both NASA and now the European Space Agency dropping off the necessary IP hardware, software and antennae to begin to transmit data from Mars more expediently, more efficiently in terms of power consumption and more clearly in terms of interference.