The InterPlanetary
Internet (IPN) project tries to make the Internet work on interplanetary
connections. This is the current obsession of Vint Cerf, one of the
people who really invented the Internet. It's a magnificent
obsession, and it might even work.
You may think that current Internet technology (i.e the IP protocol
and the layers above) could be used on radio communications since they
have been used for several decades to communicate between spacecrafts and
the Earth. But this solution suffers from major
drawbacks:
- delay. Internet protocols are based on repeated requests and
answers at several layers. On Earth, the delay is usually one fraction of
a second. Between two planets, it will take minutes or hours. Interactive
protocols cannot be used, and bandwidth will be rare.
- power supply. Photovoltaic energy is very low beyond
Mars. And hydroelectric power stations are not an option in deep
space.
- maintenance. You may complain about your PC manufacturer
support, but imagine your Web server sits somewhere between Jupiter and
Saturn? You don't want to wait for the next mission in five years.
- moving targets. Planets move and satellites move
around the planets, so your server may be invisible from time to time.
The IPN team makes the following propositions:
- delay: use gateways. Instead of communicating directly
from your computer to a server on Saturn, you will interact with a
gateway on Earth, which will interact with another gateway on Mars, and
so on until the request gets to the server. Then the answer will follow
the reverse path. Because TCP uses too many negociations between the
client and the server, it will be replaced by a new "bundle
protocol". The new protocol will not be faster than light, but it will ensure some reliability. Furthermore, you will
have to pay for bandwidth; interplanetary Internet may never be free as
in free beer.
- power supply: power will be saved by using less verbose
protocols and planning the communications.
- maintenance: gateways will be duplicated so that redundancy
ensures long-term operations. If necessary, aliens will be hired to
go and fix the servers.
- moving targets: relay satellites will be positioned
at LaGrange points, which are special orbital locations where they will
always remain visible.
Ok, but do we really need an interplanetary Internet?
Is there any decent Web site on Uranus? Some people may need IPN for
other reasons:
- to facilitate the communication between the Nasa satellites and the
Earth. The communication hardware could be much lighter, so that more
space could be devoted to scientific hardware.
- to call Web servers on other planets. It may be only for fun, but
imagine a webcam on the Moon, directed at the Earth...
So is it working now? Can I ping Mars? Not yet. But it may
start working after 2005, when the Nasa will or should deploy the first satellite
network around Mars.
That's very exciting, but I have not won the Nobel prize yet, so I
suppose I cannot be involved? On the contrary, the IPN
team may offer projects to you if you are a post-graduate.
But first read the IETF draft at
http://www.ietf.org/internet-drafts/draft-irtf-ipnrg-arch-00.txt. If it's a
little too technical for you, try http://www.spaceref.com/focuson/ipn/.