Exelis' SpearNet Team Member Radio was designed to be part of a mobile ad hoc network. ()
The tourists probably missed it, but back in 2011, a Trellisware Technologies team boarded the USS Midway, a decommissioned aircraft carrier operating as a naval museum in San Diego, and conducted a test. Armed with Android devices tethered to tiny software-defined radios, the group wanted to see what it would take to push a signal from the bowels of the ship, past the pipes, bulkheads and metal doors, to the outside world.
And so one man stayed behind in the engine room while the rest of the team took positions throughout the aircraft carrier. As the person in the engine room started recording video, the imagery traveled over the impromptu radio network, hopping from node to node until it made its way to the flight deck, where it was zapped to someone else five miles away. In all, it took just four handhelds to communicate across a notoriously disruptive environment.
“That is the power of a mobile ad hoc network,” says Jeff Harris, vice president of Wireless & System Solutions at Trellisware Technologies.
Designed to support units operating at a rapid clip, a mobile ad hoc network (MANET) provides a radio link where there might not otherwise be one. For example, the operational area might lack fixed infrastructure or obstacles might block line-of-sight communications. A MANET overcomes these obstacles by having radios in the network serve as both relay stations and end users. Voice and data communication is then just a matter of hopping across available nodes.
Still, there are questions as to how far the technology can scale. The assorted waveforms and algorithms that support MANETs differ greatly in capability, and the networks of 100-200 nodes that contractors tout are often accomplished in labs or open fields. With more networked sensors and weapons entering the battlespace, the Defense Advanced Research Projects Agency (DARPA) has asked for new “clean-state ideas” on how to build a large MANET.
An alternative to solutions like mobile towers, the first MANETs were designed to provide frontline troops with “a reliable, real-time voice and data communications capability that wasn’t dependent on fixed infrastructure or line-of-sight communications,” says Clifton Basnight, lower tactical Internet engineering and integration branch chief at Army Program Executive Office for Command, Control, Communications-Tactical (PEO C3T).
The idea was to have a self-healing network that kept dismounted warfighters in touch with one another and the command element of their force. It was also meant as way to share location information so cases of fratricide could be avoided, says Chris Marzilli, president of General Dyamics C4 Systems.
MANET technology dates back to the mid-1990s — and even earlier if one counts the initial academic research. Following a Defense Science Board study of the command-and-control challenges faced by U.S. troops fighting in urban areas in Somalia, DARPA began working with ITT, now called Exelis, on the Small Unit Operations Situational Awareness System (SUO-SAS), a prototype radio with a brand-new IP-based waveform that established a MANET of up to 50 nodes.
In 2002, DARPA handed off SUO-SAS to the Army’s Communications-Electronics Research, Development and Engineering Center, which renamed the program the Soldier-Level Integrated Communications Environment, or SLICE. “And out of that SLICE program was the birth of the Soldier Radio Waveform,” says Eric Whitehill, chief engineer of International Communication Systems at Exelis’ Night Vision and Communications Solutions Division. (Full disclosure: The author of this article once worked for a public relations agency that counted Exelis as a client.)
As MANET developed, so did software-defined radios. Charting the history of these systems, Marzilli notes that the first DARPA programs — SPEAKeasy I and II — began in the early 1990s. These efforts eventually led to the Army’s Joint Tactical Radio System program, which, though flaming out spectacularly, did produce the current-generation PRC-154 Rifleman Radio and PRC-155 two-channel Manpack. Both run the Soldier Radio Waveform, with the Manpack using others as well.
Indeed, the parallel rise of MANETs and software-defined radios is no coincidence.
“MANET helped lead to the push for software-defined radios,” Basnight insists. “The Army began to look at MANET technology in order to provide a network that was not dependent on fixed infrastructure, and determined that software-defined frameworks would enable that vision with lower sustainment costs.”
A question of scale
A number of contractors have already sold radios that can form MANETs to the U.S. Special Forces and coalition partners in Afghanistan. And PEO C3T is busy fielding the Rifleman and Manpack radios as part of Capability Set 13 to Army brigade combat teams.
But how well do MANET radios really scale? Contractors say that they have no problem linking more than 100 nodes; some say more than 200. But DARPA doesn’t seem convinced. In April, the agency put out a request for information for new, non-Internet ways to erect a MANET, stating that current approaches “only scale to around 50 nodes before network services become ineffective.”
In early August, selected respondents participated in a two-day DARPA symposium on the issue. Industry attendees included Cubic, BAE Systems, Bell Labs Alcatel-Lucent, AT&T Labs Research, and Raytheon, which developed the Next Generation MANET Waveform.
While careful not to criticize existing solutions, DARPA program manager Mark Rich argues that the growth in the number of sensors, unmanned vehicles and networked weapons requires new approaches to MANETs.
“All those things lead to much larger networks than we’ve used in the past, and they are also driving much more dynamic networks,” he says.
However, barring a revolutionary discovery by DARPA, contractors will have to continue with Internet protocol-based approaches. The trick will be sorting out which waveforms and routing algorithms yield the best performance.
Harris, from TrellisWare, notes the term “MANET” spans many waveforms, some of which — like the Soldier Radio Waveform and Wideband Networking Waveform — are long in the tooth.
“You have a flip phone, an old Nokia flip phone,” he explains by way of analogy. “It’s a cellphone, right? I have an iPhone 5C. Mine is considered a cellphone, too. Your cellphone and my cellphone aren’t even remotely the same.”
Likewise, not all routing algorithms are equal. Link State, which is still in use with some MANETs, constantly queries nodes in the network to find their location and plots tracks to them, just in case someone wants to reach them. That eats up bandwidth and makes scaling upward extremely challenging, says Tim Strobel, technical director of Tactical Communications Systems at Raytheon.
Whatever the specific solution taken in the coming years, it will have to put an emphasis on network robustness because, says Strobel, the next enemy the U.S. encounters will likely be more technologically savvy than the Taliban.
“We expect to lose satellites and GPS, and the MANETs will have to survive through that,” he says.