By Dr. Philip Baczewski, Director of  Academic Computing and User Services

Bits in Space

It's "Bits... in... Space..." with a booming voice and an echo effect.

More than four years ago, I reported on efforts to set up connections between various Mars orbiters in order for NASA to be able to relay data coming from planetary rovers. I dubbed this the Intergalacticnet, but noted that NASA had been operating a deep space network for over 40 years. More recently, I saw a note about efforts to design and deploy an "interplanetary Internet" which is proposed to support ongoing efforts in manned and unmanned space exploration.

A British company called Surrey Satellite Technology Ltd (SSTL) is experimenting with fault tolerant methods to send data from satellite observations to ground stations via a standard Internet Protocol. Networking in space is a bit more challenging than networking on the ground, but a bit less challenging than setting up your own wireless base station.

A satellite is not always in range of its communication base station as it orbits the earth. Across larger distances, asteroids, meteors, or even other planets may block a direct communication path to an earth-based receiving station. The solution being worked on for these cases is called Delay-Tolerant Networking (DTN, in case you were worried that we did not have enough Internet acronyms -- or initialisms if you want to be totally accurate.)

The idea behind DTN is that information should be able to be transferred when a communication window is available or opportune. This is done by defining a protocol supporting "bundles" of information that can operate within a store and forward service over "normal" Internet Protocols. There is a DTN Research Group who are working on developing such a protocol. Lloyd Wood and a bunch of other folks from NASA and SSTL have described such a protocol in a recent presentation. (I guess this really is rocket science.) According to Lloyd et al., "Key capabilities of the Bundle Protocol include:

• Custody transfer – the ability for a bundle node to take full responsibility for a bundle reaching its final destination.
   
• Ability for implementations to cope with intermittent connectivity if required.
   
• Ability for implementations to cope with long propagation delays if required.
   
• Ability to take advantage of scheduled, predicted, and opportunistic connectivity (in addition to continuous connectivity).
   
• Late binding of overlay network endpoint identifiers to constituent internet [SIC]  addresses."

All this talk about custody transfer, intermittent connectivity, delays, and predicted and opportunistic connectivity sounded oddly familiar, so I consulted the "Way-back Machine" and found FidoNet way back in the recesses of my (and the Internet's) memory. FidoNet was/is a store and forward network to relay e-mail between standalone computers that were able to make phone modem connections to other standalone computers (it was one of several such networks in operation back in the BBS era). One of the things proposed in that bygone era (1988) was the concept of "Bundles."

The FidoNet bundle was described as follows:

"Messages are transmitted in "bundles." A bundle is a sequence of 'packets.' Every bundle has at least two packets: a header and a footer."

The DTN Bundle is a little more sophisticated:

"A bundle is a protocol data unit of the DTN bundle protocol. Each bundle comprises a sequence of two or more 'blocks' of protocol data, which serve various purposes. . . . A bundle payload (or simply 'payload') is the application data whose conveyance to the bundle's destination is the purpose for the transmission of a given bundle."

A translation of the DTN bundle definition could read, "a bundle is a sequence of packets used to transmit messages." Somehow, that sounds familiar. It just proves that there are no old ideas -- just old implementations. So, the next time you're whizzing around the solar system reading your e-mail, remember that it may not have been possible without FidoNet. That's right -- thank dog!