Unlike TCP/IP, DTN does not assume an continuous end-to-end connection. Delays can occur in a host of situations (like when a spacecraft moves behind a planet or when a solar storm occurs) so the protocol has to be robust...and patient. With DTN, data packets are not discarded when a destination path cannot be found. Instead, each network node stores data until it can safely communicate with another node. It may take a little longer, but the data will eventually get there. The second round of testing will occur next summer when DTN software will be installed aboard the International Space Station. Over the course of the next few years, NASA hopes to have the technology aboard a wide variety of space missions. [NASA]

Communicating in space presents a bunch of problems—the Earth's rotation causes senders and receivers to be constantly changing positions, and the long distance causes equally long delays. Our current radio-based network is tailored to almost every new mission, meaning that older equipment can't be repurposed for newer shuttles.

Cerf, who more or less co-created the internet, is now figuring out new protocols that'll work in the final frontier. The project, called Interplanetary Internet, will be tested aboard the International Space Station in 2009. If it works out, space missions in the future will be able to use the same systems, ultimately making communicating from above much, much cheaper and easier. I wonder what their ping rates will be. [Technology Review]