Remember the cell phone owners who lined up at pay phones throughout the Northeast during last summer's blackout? TV images of anxious subscribers who couldn't call home made one thing clear: Traditional wireless networks can fail in an emergency.
When a power outage, earthquake or malicious attack knocks out service at cellular towers, it renders handsets useless. The same goes for wireless devices that communicate with wired transceivers, whether they're voice radios, mobile e-mail platforms or notebook computers with Wi-Fi cards. While a breakdown in commercial service triggers frustration, a similar failure in public-safety communications can have dangerous consequences.
An alternative wireless model might help first responders and other government employees exchange information without a fixed infrastructure. It could also extend communications to places that can't be reached by signals from radio towers.
Called a mobile ad hoc network (MANET), this technology establishes high-speed, peer-to-peer communications among mobile devices, each of which doubles as a router. When one device needs to send information to another, low-powered radio signals relay the data in short "hops" from one node to the next until it reaches its destination.
The system calculates the most efficient route for each data packet, changing that route on the fly as the devices move about. MANETs are designed to carry any sort of digital information, including text, voice, graphics and video.
The MANET concept grew out of research funded in the 1970s and 1980s by the U.S. Defense Advanced Research Projects Agency (DARPA). The goal then was to allow military troops to instantly establish voice and data communications where there was no fixed radio infrastructure.
In the mid-1990s, researchers also started to explore nonmilitary uses in emergency response and disaster relief, said Nader Moayeri, manager of the Wireless Communication Technologies Group at the National Institute of Standards and Technology (NIST).
Better Coverage in Buildings
NIST is one of numerous organizations, including government agencies, universities and corporations, that have conducted research on MANETs. NIST worked for a while on developing an ad hoc network for outdoor use, which could be used, for example, in the event of a hurricane, said Moayeri. "After September 11, we concentrated on doing that type of thing inside a building."
As the world saw during rescue efforts at the World Trade Center, conventional radio systems can't reach the higher floors of tall buildings, he said. "But if you have this natural relaying mechanism, you can penetrate. You can have coverage in larger buildings, and you extend the range of your radio."
Any mobile computing device that accepts the required communications card can serve as a node on a MANET. NIST has developed a pilot network using HP iPAQ Pocket PCs with 802.11b WLAN cards, with added software to handle MANET-specific functions. One protocol, for instance, determines the best route for a given transmission. A voice over IP establishes voice communications.
NIST also included location software, using triangulation techniques to determine the position of each device on the network. A commander at an emergency scene could use this to track responders within a building, Moayeri said. "This would be useful when it comes to saving lives, and also in terms of coordinating the activities of these people," he said.
Along with the mobile nodes, NIST's pilot network includes a fixed communications station to provide a connection to the outside world. In real-world use, this might be a computer in a command vehicle, linked to the Internet via satellite. Responders inside the building could place phone calls by routing communications through this link, Moayeri said.
NIST successfully tested a MANET composed of 30 devices, with 10 to 15 of them communicating at once. It's not yet clear whether MANETs can be built to support larger numbers, Moayeri said.
At least one commercial developer has created techniques to support any number of devices on a MANET. MeshNetworks of Maitland, Fla., markets a product that can operate in two modes: a pure MANET or a MANET supplemented by fixed routers. MeshNetworks promotes its Mesh Enabled Architecture (MEA) for use in public safety, emergency response, intelligent transportation and other government applications.
Rather than the open-standard 802.11 radio technology, MeshNetworks employs a proprietary radio protocol called Quadrature Division Multiple Access (QDMA). The company chose this course because the radio was designed to work in a mobile ad hoc scenario, which no existing, standards-based radio was designed to do, said Rick Rotundo, vice president of technical marketing at MeshNetworks.
The company is heavily involved in standards-making efforts in the hope that future wireless technologies will support mobile, ad hoc, broadband networking, he said.
"We would dump our radio if there was something better," Rotundo said. "Our routing technology, and our routing software, is radio-agnostic."
Like 802.11 radios, MeshNetworks' radio currently operates in the unlicensed 2.4 GHz band. Because devices in this band must operate at low power, networks using this frequency generally support broadband only over a short distance, such as the 300 feet between a user's notebook computer and a Wi-Fi access point, Rotundo said.
With transmission power held constant, the only way to increase the data rate is to decrease distance, Rotundo explained. Longer-distance transmissions mean slower data. "A Mesh network makes physics work for you," he said. "Instead of trying to transmit a mile directly, at 10 kilobits, we hop at 6 megabits, but we just hop short distances and tie those together like a bucket brigade."
Orange County Pilot
The Fire Rescue Department in Orange County, Fla., is testing MeshNetworks' technology for locating firefighters inside a structure. During a pilot last summer, it deployed a system at a building the department routinely sets on fire for training exercises.
Along with mobile units, this system included several PC-based receiving stations set outside the building. These contained communications cards and GPS receivers. Firefighters in the building wore small prototype MANET-equipped devices, developed by MeshNetworks. Using triangulation within the MANET and GPS at the receiving stations, the system calculated the firefighters' positions and tracked them on PC screens.
The pilot system located firefighters to within about 15 feet, said Bill Godfrey, deputy chief of training and information technology at the department in Winter Park, Fla. "It's still not right on top of him, but for our purposes, and for the techniques we use in the fire service, that's certainly better than what we've got now."
Godfrey said the department is still evaluating several points about the system: how precisely PCs must be placed around the perimeter of a fire scene, for instance, and how building construction affects system performance. These are important questions if fire departments are to use such a system in actual emergencies, he said. "The real issue is, can you dynamically throw this up in two minutes or less on a fire scene and get the same information when the building is not pre-engineered?"
Even if the answer is no, the technology could still save lives, Godfrey said.
Unfortunately trainees who become disoriented during live burn exercises aren't always located and rescued in time. "This technology could eliminate that variable for training," he said. "Given the cost-benefit ratio of doing that, I think the fire service will be looking at that pretty hard."
If Orange County Fire Rescue does deploy MeshNetworks' system for location, it will probably take advantage of its other capabilities as well. "We will certainly use the fact that it's carrying a data signal to do video backhaul from the structure," Godfrey said. The department could also use the system as a backup to its voice radio system and establish video conferences in the field, he said.
Justify the Cost?
Government agencies must look long and hard at applications like these before deploying a MANET, given current costs, observed Richard Reynolds, a member of the telecommunications team in the Delaware Department of Technology and Information. "The needs have to be developed and justified, as a typical cost factor that was presented to me was $100,000 to deploy a one square mile pilot system."
As with any wireless network, security is another important concern. MeshNetworks designed its MANET to operate with any encryption technology the user deploys to protect data from end to end, Rotundo said.
To ensure only authorized users join a MANET, NIST is exploring various authentication techniques. These might include "smart passes and passwords, and also biometric identification through fingerprints or the image of your iris," said Moayeri.
Developers need to establish standards to promote interoperability among different MANET devices, Moayeri said. "The other thing we want to look at is interoperability with legacy equipment," he added. That cannot mean simply inserting MANET cards into existing radios, since many agencies are still using old, analog systems.
But even if building a MANET means buying special-purpose equipment, in at least some cases, the need will justify the investment. "It is such a life safety issue for the fire service," Godfrey said, "that if the technology were available to locate firefighters and it didn't integrate with anything, fire departments would still want it, need it and use it."