MANET and the art of communication

Emerging ad hoc networks may help first responders

Imagine police, fire and emergency response units rushing to the scene of a disaster and being able to spontaneously establish an interoperable communications network, even if the local infrastructure has been destroyed.

That's the vision of mobile ad hoc networks (MANETs). The concept has been the subject of university research and thesis papers in recent years. But MANET technology (pronounced the same as the last name of French painter Édouard Manet) is beginning to make its way into demonstrations and pilot programs.

The Homeland Security Department and the National Institute of Standards and Technology are among the federal entities that have taken notice.

Industry and government executives say potential beneficiaries of MANETs range from military personnel deployed in remote areas to users in developing nations. In the homeland security context, the technology is viewed as a natural fit for emergency workers.

"If you can easily establish communications on a moment's notice, you gain a significant time element that is important in emergency situations," said Donna Rhode, director of product strategy and marketing for Cisco Systems Inc.'s Government Systems Unit. Over time, she believes first responders will migrate from fixed-infrastructure networks to the use of MANETs.

Before that happens, MANETs must mature significantly. Among the challenges are scalability and security. In addition, a standard MANET protocol has yet to emerge, although efforts to promote a common approach and ensure interoperability are under way.

As MANET development continues, other wireless methods are putting communications tools in the hands of frontline forces. Mobile IP routing, tested by NASA and the Coast Guard, is one example. Other advances include communications systems deployable in the field that blend satellite links and IP.

A Failure to Communicate

When emergency personnel from different jurisdictions converge on a disaster scene, they can't count on their communications systems to talk to one another. When public safety organizations bought radio systems in the 1960s, interoperability wasn't on the agenda, noted Alan Caldwell, director of government relations at the International Association of Fire Chiefs, Fairfax, Va.

"Most [radio systems] are still in business today and most can't talk to each other," he said.

Tapping the local communications infrastructure to coordinate relief efforts may not be a viable alternative, however. In the case of the World Trade Center attack on Sept. 11, 2001, telephone capabilities were destroyed or hopelessly overloaded in lower Manhattan.

MANETs may address the need for a technology that transcends jurisdictional differences but is independent of fixed infrastructures.

"In some instances, even in the same jurisdiction with radios purchased from the same vendor, police can't talk to firefighters," said Nader Moayeri, a manager in NIST's Wireless Communications Technologies Group. "What fire, police and [emergency medical services] have today is quite limited. There is a lot of room for improvement." He added that first-responder communications systems are typically voice-oriented, with limited data capability.

Against that backdrop, NIST has tested and deployed a MANET using personal digital assistants equipped with wireless local-area network cards. The pilot project automatically established message transmission routes among Hewlett-Packard Co. iPaq PDAs, without the need for a fixed network infrastructure, according to NIST. The MANET was able to transmit voice, text, video and sensor data.

More MANET pilot projects are set to follow. CoCo Communications Corp., a Seattle-based start-up company focusing on MANETs, plans to conduct tests in at least two cities this year, according to Peter Erickson, the company's vice president of business development. One of the network projects will focus on public safety. The company has been discussing its MANET technology with DHS.

The Pentagon is also interested in MANETs. A number of proof-of-

concept projects testing military applications are under way (see "Military applications," Page S8).

How MANETs Work

A MANET consists of autonomous mobile users and their communications devices (PDAs, for example), which all act as nodes on the wireless network. When users arrive on the scene and activate their devices, the network self-organizes and the nodes find one another automatically, Moayeri said. Once the network topology is discovered, the nodes collaborate to establish a stream of communication.

In that stream, each node can act as a source, relay point or destination. The communication flow starts with the source node and, in the case of out-of-range nodes, may hop across a couple of intermediary nodes before reaching the destination node.

NIST officials believe this "multihop" method can carry more traffic than a direct link, according to Moayeri. That's because the shorter hops use less power, cause less interference and provide better frequency usage, he said.

Another edge for MANETs is survivability. The distributed nature of MANETs means there is no single point of failure. Wireless local-area networks, in contrast, have access points that link to a wired network. Cellular communication, meanwhile, relies on base stations. "Most of the wireless networks are networks that have some sort of infrastructure," Moayeri said.

If a MANET node is destroyed or otherwise leaves the network, the MANET can reconfigure itself. "It adapts itself nicely to node link failures," Moayeri said.

The self-organizing nature of MANETs stems from having routing functionality embedded in each node. MANET routing protocols provide that functionality. The Dynamic Source Routing protocol, for example, is designed for use in "multihop wireless ad hoc networks of mobile nodes" and allows those networks to self-configure, according to a draft specification submitted to the Internet Engineering Task Force's MANET Working Group.

Other protocols submitted to the working group include Optimized Link State Routing, Topology Dissemination Based on Reverse-Path Forwarding, and Ad Hoc On-Demand Distance Vector. All provide a mechanism for finding the best path through a MANET.

Elsewhere, Cisco officials are eyeing a MANET extension for Open Shortest Path First (OSPF), a routing protocol for IP networks. "We are planning to extend OSPF to have MANET capabilities similar to those we see in research protocols," said Ed Carney, vice president and general manager of Cisco's Government Systems Unit. He said the technology eventually will be submitted to IETF for standardization.

CoCo Communications, meanwhile, has developed its own MANET routing protocol. The company has applied for a patent.

Limitations

Considerable work remains before MANETs can leap from laboratories to the front lines. To prepare MANETs for prime time, developments in several areas are necessary. Scalability is one consideration. Thus far, test implementations have generally involved only a handful of nodes. Moayeri said NIST has experimented with 10 to 15 nodes. He expects to take up the scalability issue later this year. "Any sort of architecture, at some point, will break," he said. Whether that breaking point is 100 nodes or 1,000 nodes is still to be determined, he added.

Jeremy Bruestle, CoCo Communications' chief technology officer, said scalability is among the company's key pursuits. "If [a MANET] starts to break down as more people come on to the scene, it will be more of a problem than helpful," he said. Bruestle said the company has tested its protocol on 300 nodes and believes the technology can extend beyond that number.

The scalability question is part of a broader issue: reliability. Ensuring a network can take on more users without breaking down is one concern, but other considerations include the length of time it takes to establish a route and the probability of route error, noted Taylor Salman, director of business development at OpNet Technologies Inc. OpNet provides software for simulating MANETs.

Then there's the problem of securing a network of independent nodes (see "Security concerns," Page S6). "It's not easy to secure an ad hoc network," said Will Ivancic, a senior research engineer at NASA's Glenn Research Center. Ivancic is focusing on mobile IP research, noting that MANETs are far from mature deployment in his work areas.

Other observers cite the limited nature of MANET pilot programs. They would like to see tests involving more types of devices (radios, for example) and different wireless technologies such as Code Division Multiple Access and Global System for Mobile Communications.

Tests based on 802.11 represent "a very preliminary substantiation of MANETs," Cisco's Rhode said. But "it's a good basis for identifying

requirements."

Moore is a freelance writer based in Syracuse, N.Y.

***

MANETs: The virtue of ad hoc

Mobile ad hoc networks provide "multihop" communication among autonomous nodes, allowing a low-power network to be established at a disaster site. Each node can act as a source, relay point or destination. This method also can extend the range of radios within large buildings.

NEXT STORY: Defense to unveil GIG plans