Will there be a ".mars"?

Network World Fusion, 02/01/99

With the ongoing construction of the International Space Station refocusing the world's attention on space exploration, scientists at NASA's Jet Propulsion Lab (JPL) are working on an "Interplanetary Internet." Such a system would establish gateways in outer space to help send and receive data to and from the nether reaches of our solar system and possibly beyond.

Who better to help lead this effort then Vint Cerf, the man known as the "father of the Internet?" Last August, Cerf, the senior vice president of Internet Architecture and Engineering at MCI WorldCom, was appointed "Distinguished Visiting Scientist" at NASA's JPL in Pasadena, Calif. He and JPL's Adrian Hooke lead a small team of experts looking to merge the work of the Internet and space communities.

Cerf recently partcipated in an e-mail-based Q&A on the subject with Network World Fusion Staff Writer Jason Meserve:

Q. What is the project's objective?

A. The principal aim of this project is to design and build an extension of the Internet concept to work between the planets and their satellites. Included in this vision are mobile vehicles that operate between the planets, satellites, asteroid belt and so on. A secondary but also important aim of the project is to make it possible to use ordinary Internet technology aboard space craft and also on-planet (or what we sometimes refer to as "dirtside") so that our research stations, and eventually our habitats on other planets, can be as linked together as Earth is becoming today with standard Internet.

Q. Is there a time line set for any of the objectives? Or is this more of a think-tank/long range thing?

A. Well it has both components. There are tests underway which might allow us to evaluate variations on TCP/IP for Earth/Moon operation, and longer term opportunities to put the Interplanetary Gateway concept to the test with a Mars mission starting in 2003. In the long term (2020-2040) we hope to see the Interplanetary Internet (InterPlaNet) become a reality as parts of it can be put into place, piece by piece, as new missions into space are launched. My personal hope is that we will be able to put up a backbone for the solar system, to include interplanetary gateways operating out of the plane of the ecliptic so as to keep all planets "in view" all the time (not counting antenna shadow as planets and satellites rotate them out of direct view).

Q. What kind of technology are you envisioning? I read that a standard communication to Pluto takes 6 hours, making IP a bad choice. What are the alternatives?

A. Actually IP isn't a bad choice - it is TCP that gets bogged down with such long round-trip times. However, we've tentatively concluded that each dirtside and spacecraft Internet will have its own address space to avoid trying to do interplanetary zone transfers among DNS servers. Whether we use standard IPv6 (our preference over IPv4 for interplanetary work) or some variation in the "between planets" component of the system is still a little open. What we do want, however, is to unlink the distinct internets so they can evolve freely, and to bind them at the domain name or similar level (we have some non-domain name ideas which we think might be useful for identifying end-points in the system).

Q. In the long run, will this help improve data communications on earth as well?

A. Yes, we think so, because as speeds go up (data rate) in terrestrial systems, the relative delay due to speed of light limits also goes up. Eventually, at terabits per second, round trip times of 60 milliseconds across the US look a lot like much longer times as lower data rates. That means the methods we use to overcome the long round trip times over interplanetary distances may also apply to super high speed systems here on Earth.

Q. How and why did you get involved in the project? Are you spending time at JPL in Pasadena, or working from afar?

A. I began talking about this idea in October 1997 and early in 1998, one of my engineers who had worked with JPL alerted them to my interest. Adrian and his team had already started work on adapting Internet protocols to deep space use. We got together, it clicked, and after a lot of spadework on Adrian's part, we got support from the head of NASA to proceed. I am out at JPL every month or so but there is a local team at MITRE in the Washington DC area as well as a team out at JPL. We do a lot of email between face to face meetings.

Q.How many companies are working on this? What is the size of the project team in terms of people?

A. It is a small operation, by choice - you cannot do design with a large team. JPL and MITRE and some consulting from key Internet luminaries.

Q. Where does the funding come from: the government or are other companies like MCI WorldCom contributing as well?

A. MCI WorldCom is NOT contributing, except for the fact that they allow me the time to work on it. JPL, NASA, and we expect, DARPA, will be providing basic funding for the design work.

Q. What is your response to the detractors who say this will never work?

A. Some people said that packet switching was crazy, too! Many people opted for OSI rather than TCP/IP. Well, guess what?