Standards still emerging for military wireless nets
Civilian technologies are being deployed piecemeal, but military wireless networks are still a work in progress
No one would ever mistake the battlefield for a Starbucks. But the Defense Department is aiming for the same sort of ubiquitous wireless networking that the average person finds at the corner coffee shop. To achieve that, the military wireless network of the near future will incorporate civilian technologies such as Wi-Fi while accounting for all the ways that networking a battlefield is not like networking a coffee shop or even a city.
Although many wireless mesh networks and mobile ad hoc networks, or Manets, have been deployed, military systems architects continue to struggle with fundamental issues of security and reliability that must be resolved before the technology can be deployed more widely and for more critical missions.
Ultimately, the goal is to support the network-centric ideal of information superiority. The voice, video and data that warfighters need for situational awareness range from the ability to locate friendly forces on a moving map to pulling video feeds from unmanned aerial vehicles and ground-based robots.
For projects that attempt to adapt commercial technologies for the military using networking equipment based on 802.11 (Wi-Fi) and 802.16 (WiMax), security is a major consideration but not the only one. Other challenges include figuring out how to offer wireless networking that supports a high degree of mobility when nodes might be helicopters, tanks, trucks or soldiers on foot. Such networks must also be able to operate with little or no reliance on a centralized infrastructure and must fit the Joint Tactical Radio System (JTRS), the next-generation architecture the Defense Department has been trying to define for a decade. JTRS is designed around software-based radios that can be reprogrammed to work with multiple protocols and modes of communications, including Manets.
This technology is not in the field to a large, large extent, and we’re just beginning to field mesh networks,” said Gayle Grant, program director for tactical wireless networking at the Army’s Communications-Electronics Research, Development and Engineering Center (CERDEC), which is one of the main centers of activity for defining future military wireless networks.
JTRS technologies will start making their way into the field to a greater extent in the next five years, she said, adding that demonstrations have been under way for years. Sharon Mackey, chief of CERDEC’s Commercial Wireless Branch, said Army units in the field bought most of the commercial wireless technology that’s been fielded to date and had the vendors install it rather than operating under the Army’s central contracting oversight.
“It’s not gone through the rigors, the testing, that we would require,” she said.
“One of the reasons there are problems is there is no standard yet,” said Tom Badders, director of wireless networking solutions at systems integrator Telos. “There is some work on a specification called 802.11s for mesh networks, but the standard has not yet been ratified.”
As a result, vendors are building mesh networking solutions using technology that is proprietary or only loosely aligned with the proposed standard, “and that means you have to pick a vendor and stick with it,” Badders said.
Telos recently won a contract to provide at least 13,000 transportable wireless base stations for Army logistics operations. They will incorporate wireless bridge hardware from Fortress Technologies. By taking advantage of a mesh networking upgrade to the Fortress equipment, Telos was able to demonstrate that it could support a wireless network that covers as much as a seven-mile radius and provides service to hundreds of users. The equipment will be deployed under the Army’s Combat Service Support Automated Information System Interface program.
Other vendors, such as Aruba Networks and Rajant, also have technology that meets military requirements, Badders said, but having a standard would make it easier to deploy multivendor solutions.
The mesh networks that some municipalities use to provide wireless access to residents or city workers rely on routers that wirelessly relay signals from each node to its neighbors. But although they minimize dependence on a wired infrastructure, conventional mesh networks typically still rely on a wired telecommunications infrastructure for back-end access to enterprise networks and the Internet.
To be suitable for tactical deployment, a military equivalent must be able to substitute a satellite uplink for that backend network connection and continue functioning if it loses the connection. To make things tougher, many military scenarios would require mobile routers that wouldn’t stay strapped to telephone poles, as in a municipal Wi-Fi system.
“The original definition of mesh networks did not address mobility,” said Charles Graff, a senior engineer at CERDEC.
The military is more interested in developing Manets for which the mobility of network nodes is a key feature. They aren’t new, Graff said. Early designs were based on packet radio technology that dates to the 1970s. “But the questions revolve around scalability – how big can it be, how many nodes – and how fast can it recover [from the failure or destruction of a node], and, of course, security.”
Mesh networks are supposed to reconfigure themselves as necessary to adapt to the loss of a particular node. If one node discovers that its neighbor has dropped off-line, it will attempt to reach out to the nearest mesh node within range of its signal. But although that design confers a certain resiliency, it still assumes that the mesh will be relatively stable, with only an occasional need to reroute to avoid trouble spots.
Such designs can be useful in a base setting, where mesh networking makes it possible to provide network access over a large area with a minimal need for wiring. But they don’t lend themselves to a scenario in which many, if not most, nodes will be continually moving in and out of range of other nodes.
Therefore, military Manets must be designed so that the network will always be reconfiguring itself. Each networked device must be able to move between connected environments and quickly reorient itself to communicate via a different set of network nodes while also authenticating to the network.
The security challenges are considerable, said Stephen Lucas, chief engineer at CERDEC’s Information Assurance Division. “Everybody can be a router in these types of networks, so issues of trust and determining who can be a user are a big concern for us,” he added.
In a wired setting, military networks often verify access rights against a centralized database or directory. Authorized users are given a Common Access Card containing a cryptographic key that they use to authenticate themselves to the network. Security keys can also be built into devices such as radio handsets rather than issued to specific users.
But in a wireless setting in which the only link to centralized information resources is via satellite, looking up security credentials would be too slow and expensive, Lucas said. One of his major efforts in the past year has been to develop a decentralized security and authentication scheme, which is now being incorporated into the Army’s Future Combat Systems program.
Instead of relying on cryptographic keys issued by a central authority, the ideal system would issue keys and authenticate them locally. However, that approach carries its own risks because it moves a crucial piece of the security infrastructure into the field.
“These are trusted systems so we don’t want them to fall into the wrong hands,” Lucas said. “We have to do a lot of trade-offs coming up with the acceptable levels of risk we’re going to assume when we deploy these networks.”
In the first place, establishing the relationship among keys, devices and individuals is much tougher when users roam around on a battlefield, Lucas said. “They may lose connectivity at different moments in the mission, and then you have to determine did their link just fade out. Or were they overrun? Was that node physically captured? Do we need to get that node off the network? And that’s very hard to do.”
CERDEC researchers also contributed to defining the Soldier Radio Waveform, the infantry communications protocol for JTRS, and they support the command, control, communications, computers, intelligence, surveillance and reconnaissance On-The- Move test bed program for the new technologies at Fort Dix, N.J.
Meanwhile, as developers prepare those technologies for deployment, independent research and development projects elsewhere are seeking other alternatives. For example, under a $1.4 million DOD grant, the Navy is working with Fortress to create a prototype Manet whose nodes could communicate via a variety of radio frequency protocols. It would also use what the company promises will be superior routing algorithms.
Roger Kuhn, a Coast Guard reservist and the senior project engineer overseeing the Navy initiative, said the company was able to secure a grant from DOD’s Quick Reaction Fund, which seeks to speed the development of commercial technologies that can benefit the military by streamlining the procurement process. If the prototype lives up to expectations, Fortress will have an opportunity to sell or license the resulting technology.
Kuhn said the goal is to create a model for Manet nodes that can communicate with one another regardless of whether they connect via an 802.11 Wi-Fi link, a longer-range WiMax connection or a military radio frequency.
Magued Barsoum, Fortress’ chief technical officer, said the company’s wireless nodes would not only see other Fortress nodes but also other data radio technologies in the area.
“We create adjacency to everybody we can see and then figure out what is the best path to reach the ultimate destination,” he said. “We constantly evaluate what is the best path and keep track of all those adjacent nodes so that if one goes away, we always have a backup.”
He said he hopes to demonstrate prototypes based on the Fortress ES520 wireless bridge by the end of the year, with a software upgrade to support the new capabilities available commercially in 2009.