Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×
The Internet NASA Networking Space

ISS Launches First Permanent Node of "Interplanetary Internet" 121

schliz writes "Researchers developing the 'Interplanetary Internet' have launched its first permanent node in space via a payload aboard the International Space Station. The network is based on a new communications protocol called Delay/Disruption Tolerant Networking (DTN). It will be tested heavily this month, and could give astronauts direct Internet access within a year. The Interplanetary Internet is the brainchild of Vint Cerf ('father of the Internet'), among others. Last year, NASA tested the technology on the Deep Impact spacecraft." Update: 07/13 20:01 GMT by KD : If by "permanent" we mean seven years.
This discussion has been archived. No new comments can be posted.

ISS Launches First Permanent Node of "Interplanetary Internet"

Comments Filter:
  • Define "permanent" (Score:5, Informative)

    by qengho ( 54305 ) on Monday July 13, 2009 @08:11AM (#28674613)
    It'll be permanent until the ISS is de-orbited in 2016, [washingtonpost.com] eh?
  • by Anonymous Coward on Monday July 13, 2009 @09:46AM (#28675441)

    it uses, of course, DTN, as the protocol. DTN sits in between IP and the other layers (UDP,TCP).. Think of it is as intelligent "store and forward" routing for long latency or sporadic links. It has a lot of flavors, but ultimately, you transfer a bundle of data to another node (perhaps on a predetermined schedule), when the other node agrees it's got it, "custody" of the bundle transfers.

    The physical layer is standard microwave for space (S and K band for station, X band and Ka band for deep space links of the future). It can also be carried over wired media (or heck, avian carriers as well)

  • by CarpetShark ( 865376 ) on Monday July 13, 2009 @09:47AM (#28675461)

    "It will be tested heavily this month" , so, they are going to post the URL on slashdot ?

    No, they're going to let slashdotters whine about it, and see if it cracks.

  • Re:I thought... (Score:3, Informative)

    by jandrese ( 485 ) <kensama@vt.edu> on Monday July 13, 2009 @09:59AM (#28675607) Homepage Journal
    A single DTN node is not terribly useful, where you see advantages is when you have multiple lossy or intermittent links along the path. In this case, TCP will perform very poorly as it can only even try to push data on the cases where all of the links are up and will suffer a lot from TCP's slow start.

    Think of it like a game of Frogger: TCP only knows how to play by waiting until all of the cars and logs are line up just right so it can jump all of the way across at once. DTN plays like a human player, one step at a time while it's safe and then waiting for it to become safe for the next step.

    One of the interesting aspects of DTN is that it is transport layer agnostic. This means the very same bundle can be sent over TCP, then SCPS, then USB key or whatever and the applications never have to deal with it. The applications only have to speak DTN. There are local Daemons in the network that choose the most optimal transport protocol for whatever the next hop is, and then send the bundle over that.
  • by scubamage ( 727538 ) on Monday July 13, 2009 @10:04AM (#28675673)
    Also, despite the bad form of replying to myself, I found the RFC's specifying the protocols here: RFC 4838 (DTN) [ietf.org] and RFC 5050 [ietf.org].

A morsel of genuine history is a thing so rare as to be always valuable. -- Thomas Jefferson

Working...