Slashdot Mirror

Effort has been underway for quite some time - by folks such as Vint Cerf, no less - to facilitate Internet over long delays. Surprisingly, there has been terrestrial (or aquatic) applications in the research as well, for example solar-powered sensor networks that can only transmit during daylight hours.

There's a nice overview architecture draft from 2003, especially interesting bits are in the routing section (12.3-12.4), see https://tools.ietf.org/html/dr... - the eventually published RFC https://tools.ietf.org/html/rf... has nowhere such interesting figures about routing between Earth and Mars :)

Anyway, the underlying arch is relying on putting a "bundle layer" between applications and transport, a layer 5 if you will - and the bundling will attempt to hide the long latencies. Naturally, for interactive applications this won't work, but for everything else why not...There are some implementations at http://www.dtnrg.org/wiki/Code.

:( :( :(

A couple places to start would be an explanation by NASA and the organisation for developing the protocols at the Delay Tolerant Networking Research Group.

http://www.dtnrg.org/wiki TFTY

The article is correct, the DSN has just a couple of letters in common with DTN, and nothing to do with the Bundle Protocol.

Delay/Disruption-Tolerant-Networks have been researched and developed by the DTN Research Group and the Bundle Protocol has been an RFC since 2007. It's possible to download an open-source reference implementation from SourceForge.

Actually NASA also use their own protocol, called ION (Interplanetary Overlay Network).

Related links for this article:
DTN Research Group: http://www.dtnrg.org/wiki
lots of docs: http://www.dtnrg.org/wiki/Docs
overview presentation: http://jeroen.massar.ch/presentations/files/CCC2007-DTN-Upgrading-Martian-Carrier-Pigeons.ppt

The book: http://www.amazon.com/dp/1596930632
Wikipedia: http://en.wikipedia.org/wiki/Delay-tolerant_networking

Source code: http://dtn.dsg.cs.tcd.ie/sft/ltplib/

Oh and yes, theoretically this extends the Internet in the same way that various other protocols do, eg 6lowpan etc.
And yes, as it is store-and-forward it looks an awful lot like SMTP.

Enjoy ;)

Related links for this article:
DTN Research Group: http://www.dtnrg.org/wiki
lots of docs: http://www.dtnrg.org/wiki/Docs
overview presentation: http://jeroen.massar.ch/presentations/files/CCC2007-DTN-Upgrading-Martian-Carrier-Pigeons.ppt

The book: http://www.amazon.com/dp/1596930632
Wikipedia: http://en.wikipedia.org/wiki/Delay-tolerant_networking

Source code: http://dtn.dsg.cs.tcd.ie/sft/ltplib/

Oh and yes, theoretically this extends the Internet in the same way that various other protocols do, eg 6lowpan etc.
And yes, as it is store-and-forward it looks an awful lot like SMTP.

Enjoy ;)

3 is basically already done. See Delay Tolerant Networking, which is already used for orbital packet switched networks.

than current rovers without the 10-minute time lag to Earth.

At opposition, the average round trip time (RTT) to Mars is 9 minutes.

At superior conjunction, the average RTT to Mars is 42 minutes.

At other times, the RTT will be in between these two values.

Both of these numbers will vary at the 10% level due to orbital eccentricities and inclinations, but, clearly, most of the time the RTT will be greater than 10 minutes.

However, this is almost irrelevant. All currently and planned rovers or landers use "bent-pipe tracking," where data is sent to an orbiter, and then the orbiter, sometime later, sends it to the Earth. This greatly increases the effective RTT (there are not orbiters passing over any given surface location at any time).

I believe that the Phoenix, the current rovers, and the Mars Science Laboratory all basically plan on an effectively daily RTT (i.e., at best one up and down link per day). These long effective RTTs and the use of orbiters to store-and-forward data are part of the motivation behind the efforts around Delay / Disruption Tolerant Networking (DTN) - AKA the Interplanetary Internet.

The trick is that you don't have to use TCP as your transport layer. DTN bundles can be transmitted over UDP, NORM, sneakernet, carrier pigeon, or anything else you can write a convergence layer for. Since DTN abstracts away the lower levels, each hop can use the transport layer that is most appropriate, like TCP on an internet hop, SCPS on a satellite hop, etc...

More information is available on the DTN Research Group's homepage: http://dtnrg.org.

Delay Tolerant Internet (or DTN) is the current version of Vint Cerf's
"Interplanetary Internet" - basically, making a TCP-like protocol in situations where there may be long delays and no end-to-end connectivity. I thought that there was a test of this on a shuttle flight but cannot find a link, Vint Cerf last year talked about a test in 2010.

To me, that is a lot more interesting than just having a switch in LEO.

Here they are, at the home of the Delay Tolerant Networking Research Group:
http://www.dtnrg.org/

Whilst the Delay Tolerant Network Architecture (a store-and-forward overlay network, see RFC 4838) and Bundling Protocol are mainly there to solve different problems (mostly the potential lack of a known end-to-end path) the Licklider Transmission Protocol (LTP), is designed primarily to run efficiently over single link-hops with considerable round-trip times (typically introduced by light speed constraints) but also deals with bandwidth asymmetry and mixed reliable or best effort delivery.

Deploying BP over LTP on these kind of links seems to be the plan. Then BP can extend from the orbital node down to the ground over TCP or SCPS-TP (http://www.scps.org/) or similar. BP is designed to have a shim-like Convergence Layer (CL) to interface with the whatever stack underpins it on any specific link. The open source reference implementation (Apache 2.0 licence) currently supports TCP and Bluetooth (Linux only). I've implemented a mostly working CL for AX.25 which I hope to try out using ham radio gear soon.

Check out the Delay Tolerant Networking reference implementation at DTNRG -- maybe not consistently formatted through and through, but overall design is solid. Full Disclosure: I am not one of the primary authors, but I have contributed a few items.

From the site you referenced http://www.scps.org/scps/html/related_links.html:

While the SCPS protocol suite enables terrestrial networks to extend into cislunar space, extending the internet to reach other planets requires new techniques. In 1998, DARPA funded the JPL/MITRE/SPARTA SCPS team to explore how an Interplanetary Internet might be designed. Dr. Vint Cerf joined the team, and a number of new concepts were explored.

The primary challenge was in dealing with light-time delays to other planets - minutes or even hours could elapse between a signal and its acknowledgement. It soon became evident that if one could deal with minutes or hours of delay caused by interplanetary distances, the same techniques could be equally useful for dealing with delays caused by disconnection or other comm link disruption.

In recent years, this work has been generalized and brought into the Internet Research Task Force as a Working Group on Delay (or Disruption) Tolerant Networking. For further information see: http://www.dtnrg.org/

IPN is closely related to Delay Tolerant Networking (DTN), which has more potential applications than just interplanetary networking. There's a whole bunch of uses for this stuff that can include real-life sensor networks, healthcare scenarios, emergency situations, and periodic shipping of data to isolated communities. Many situations where you either don't have a network connection, or you can't guarantee a network connection, can make use of some of this stuff.

As with any news article, it is trying to explain the concepts to a general audience. This always leads to misconceptions about what the technical solutions and problems are. Primarily, DTNs are not designed to "fix" the internet, it is designed to deal with disruptions at the edges, and to deal with challenged networking environments (primarily mobile ones). If you are interested in some technical information (some shameless self-promotion as the DieselNet project mentioned in the article belongs to myself and my colleague).

http://prisms.cs.umass.edu/dome
http://www.dtnrg.org/

If you want to know more, there are a ton of good papers being published in networking conferences on DTNs

Actually, a planned convergence layer for the DTN project is sneakernet.

DTNs work by storing packets (well, "bundles," really) at the router, until an opportunistic connection is available. Bundles move from hop to hop, until they arrive at their destination.

This is accomplished over a variety of "convergence layers," such as TCP and UDP, with UDP being the most commonly used for transmitting bundles currently in research. However, other convergence layers for other uses are being planned. One of these is sneakernet, where data is copied to physical media, the media is physically taken over to another node, and the other node reads the data and forwards the bundles to the next hop as available.

The main research page is at http://dtnrg.org/.

http://slashdot.org/comments.pl?sid=161582&thresho ld=1&commentsort=0&tid=95&mode=thread&pid=13510971 #13512140

Which support the development of international standards for protocols that don't break over long distances with lossy data link layers. The point being that with a proper delay-tolerant protocol, reliability goes up and long-distance links become more efficient. No one intends to surf the internet from Mars, but it would be nice to reliably send commands to and receive data from a rover via a secure link on a computer with just a standard internet connection. On top of this, a good deep-space protocol would get the information from source to destination whether or not the rover has a line-of-sight link or must go through an orbiting probe, and it would not require the scientist to worry about the messy details of setting up the link.

All of this is missing from current space protocols. Interestingly enough, if you read through the delay-tolerant-networking research group's website ( http://www.dtnrg.org/ ), you'll see that these protocol standards have terrestrial applications with civilian, miltary and scientific projects.

http://www.dtnrg.org/
http://www.ccsds.org/

...which is already being researched by the NASA and the IETF: more info.

Anyone heard of delay (or is it disruption?) tolerant networking? I could've sworn I saw an article on it on /. a while back ... http://www.dtnrg.org/ ... is Vint in on that too?

See the Berkeley TIER Project and the Delay Tolerant Networking Research Group"