cell phone towers and phone lines knocked out by Hurricane Katrina,
police in New Orleans at one point resorted to communicating via a
single radio channel on their walkie talkies. The Mississippi National
Guard even employed what the Associated Press called “ancient
battlefield tactics,” having runners sprint between commanders to get
If recent events like these have
tragically demonstrated the need for new communications technology that
can function independently of infrastructure, then an ambitious,
multidisciplinary project led by CITRIS-affiliated researcher J.J.
Garcia-Luna-Aceves is aimed at figuring out how to build just that.
Called Dynamic Ad-Hoc Wireless Network (DAWN), the project brings
together scientists from seven leading universities to develop the
underlying technology needed for scalable, ad-hoc or peer-to-peer
wireless networks. The hope is that these networks could be relied upon
for communication in rapidly changing environments like disaster areas
or war zones.
“In the future, every rescue worker could
have a wireless internetworking node (WIN) and this node would be able
to relay information from other nodes and also store information until
it gets close enough to other nodes to relay the information,” explains
Garcia-Luna-Aceves, Jack Baskin Professor of Computer Engineering at UC Santa Cruz.
main advantage of using WINs in such scenarios is that they would
connect to one another rather than having to relay information back
through a single hub, eliminating dependence on vulnerable
infrastructures like cell phone towers and land lines. The resulting
ad-hoc network would be designed to accommodate the influx of
additional nodes as well as the exit of others. They could also send
and receive other types of communications, such as text messages and
photographs, with built-in memory sticks storing information whenever
the network became unavailable.
To make all this happen,
though, DAWN researchers first need to establish the underlying theory
and protocols that will enable wireless networks to scale up in this
fashion. “This project is really looking at the science of ad-hoc
networks. The fact is that in military networks and even in commercial
settings, we don’t know how to make networks scale and we’ve yet to
take advantage of all these new physical layer technologies like
storage and more sophisticated receivers,” says Garcia-Luna-Aceves.
of the biggest challenges will be getting WINs to configure themselves
automatically. In a typical internet connection, a client connects to a
main server, but in ad-hoc wireless networks, there is no main server;
the nodes connect directly to one another. As anyone who has set up a
home internet connection knows, it's often necessary to take a number
of configuration steps in order to get the service to work. On the
other hand, nodes in constantly changing network would need to
automatically do this as one rescue worker’s node came into or lost
contact with another’s node.
Other challenges include
security. Particularly in a military operation, you don’t just want
anybody with a node to be able to leap onto the network. There are
practical matters to consider as well, such as trade-offs between the
size of the devices, the number and types of functions it can perform,
and the amount of power it needs to function–plus, getting the
protocol stack to line up so that all the various functions work
In order to bring about results in a timely
manner, the project is being split up among researchers at seven top
universities, UC Santa Cruz, Massachusetts Institute of Technology,
Stanford University, UC Berkeley, UCLA, University of Illinois at
Urbana-Champaign, and the University of Maryland. Collaborators from
industry and the military are providing the teams with realistic
scenarios to use in their simulations. The project is funded by a
five-year, $5.2 million U.S. Department of Defense grant to be shared
by the seven universities. The kick-off meeting was held on July 26.
For more information:
"UCSC to lead ambitious multidisciplinary research project on wireless communication networks" by Tim Stephens (UCSC press release, April 4, 2005)