If scientists at the U.S. Defense Advanced Research Projects Agency (Darpa) get their way, in a few years there may be networked clusters of dozens or even hundreds of small, cheap, easily replaceable satellites working together in place of the large, expensive and difficult-to-replace birds now in orbit.

Darpa researchers have worked for several years and spent tens of millions of dollars on a program known as “System F6” (Future, Fast, Flexible, Fractionated, Free-Flying Spacecraft), whose goal is to create clusters of satellites linked by ad hoc wireless networks allowing them to autonomously share tasks such as processing, data storage, sensing, communications relay and navigation, while trading off missions if any satellite fails or falls out of orbit. Darpa wants to conduct an orbital demonstration in 2014-15. Raytheon announced in July that its BBN Technologies segment has been awarded $2.4 million to design the network the satellites will use.

“We're building an embedded computing cloud in the sky with real-time requirements and multiple users,” says Craig Partridge, principal investigator at Raytheon BBN Technologies.

While Darpa hasn't specified the project's size, Raytheon is designing the network for clusters of up to 100 satellites in which each would have 10 applications running at one time, both consuming and producing data. And while each node on the network operates independently, Partridge says that if parts of the network go down, content will still flow across those that remain. With the satellites in orbit, users on the ground “can always upload new missions to them. . . .We view it as a dynamic programming environment,” he adds.

Darpa didn't return repeated calls for comment. Nevertheless, the idea here is to use clusters of cubesats, tiny satellites that cost anywhere from a few thousand to a few hundred thousand dollars each, and weigh a few pounds at most. University research teams have been launching the mini satellites into space for several years, and the Army, Air Force and Navy have experimented with the technology. (For an article on the U.S. Army's work with cubesats, see DTI June 2009, p. 14.)

Even U.S. special forces have cubesats in low Earth orbit. Last December, U.S. Special Operations Command (SOC)sent up four cubesats on the SpaceX rocket to demonstrate the transmission of tagging, tracking and locating data, according to Doug Richardson, a civilian official at SOC, who revealed the information at the Special Operations Forces Industry Conference in Tampa, Fla., in May. Richardson did not provide additional details.

“The military is moving toward using cubesats more often,” says William Ostrove, space systems analyst at consultant Forecast International. It's not only the cost or the ability to upgrade quickly, he says, but the fact that they're tough targets for enemies to shoot out of the sky.

“It's hard to hit a target that is roughly the size of a coffee cup screaming across the sky,” says Brian Zufelt of the University of New Mexico, who works with the Air Force Research Laboratory and NASA on cubesat programs. Zufelt's lab will orbit three 4 X 4 X 4-in. cubesats this fall. The satellites will be built and launched for $35,000.

Zufelt says one of the ideas kicking around his team is to use a cubesat to develop a way to shoot a 1-watt laser from low Earth orbit to a soldier on the ground, who would receive the transmission with “a simple telescope and a photon detector attached to the lens.” While the laser would disperse at that range, “the photons would make it to the telescope and be able to transmit communications. So you would have an unjammable communication link that is pretty much undetected,” Zufelt says.

Commercial off-the-shelf equipment works well in the cubesats, so much so that “you can put what you find in your typical Android phone and shoot it into low Earth orbit,” he adds.

Currently, the biggest limitation to the technology is power-generation capabilities that keep the applications on the vehicle functioning, since the satellite is too small to equip for long missions with the power technologies currently in use. “That's going to be the limitation for a long time,” Zufelt advises. “Anything the military wants, the technology is there that can fit into that structure, but the issue is power.”