Download a high resolution photo of professor Kenneth Frampton and graduate students Isaac Amundson and Stephen Williams.
NASHVILLE, Tenn. ñ The modern world is filled with the uncoordinated
beeping and buzzing of countless electronic devices, so it was only a
matter of time before someone designed an electronic network with the
ability to synchronize dozens of tiny buzzers in much the same way that
frogs and cicadas coordinate their nighttime choruses.
"Several years ago, I was on a camping trip and we pitched our tent in
an area that was filled with hundreds of tree frogs," says Kenneth D.
Frampton, an assistant professor of mechanical engineering at
Vanderbilt University, who dreamed up the project. "The frogs were so
loud that I couldn’t get to sleep. So I began listening to the chorus
and was fascinated by how the pattern of synchronized calling moved
around: Frogs in one area would croak all together for a while, then
gradually one group would develop a different rhythm and drift off on
its own."
Last summer’s emergence of cicada brood X brought back that memory and
prompted Frampton to assign undergraduates Efosa Ojomo and Praveen
Mudindi ñ working under the supervision of graduate student Isaac
Amundson ñ with the task of simulating this complex natural behavior
using a wireless distributed sensor network. They presented the results
of their project on Nov. 16 at the annual meeting of the American
Acoustical Society in San Diego.
Consulting the literature about animal vocalizations, the engineers
discovered that a number of different theories have been advanced to
explain such naturally occurring synchronized behaviors. They may have
evolved cooperatively in order to maximize signal loudness, to confuse
predators or to improve call features that attract potential mates. Or
they may have evolved competitively in order to mask or jam the calls
of nearby animals.
"Whichever theory is true, it is clear that these behavior patterns are
complex and offer an interesting inspiration for group behaviors," says
Frampton.
One thing that these behaviors have in common is that they are produced
by groups of animals who are in communication with each other but who
are acting on their own. Networks consisting of nodes that communicate
with each other but act independently according to simple rules are
becoming increasingly popular and were the obvious system to use.
"There is a great deal that we do not yet know about the group behavior
of such systems," says Frampton. "So, in addition to being a lot of
fun, the synchronized calling experiment is adding to our understanding
of the behavior of this kind of network."
The engineers began with a wireless network of 15 to 20 "Motes," a
wireless network designed by computer scientists at the University of
California-Berkeley and manufactured commercially by Crossbow Inc.
These are small microprocessors equipped with wireless communications.
The researchers added a microphone and a buzzer to each node.
To mimic synchronized calling behaviors, the researchers first
programmed a single leader, dubbed the alpha node, to begin calling
(buzzing) with an arbitrary duration and frequency. The alpha node was
set so it called at this rate regardless of any other calling in its
vicinity. The remainder of the devices, referred to as beta nodes, were
programmed differently. They were instructed to listen with their
microphones and when they hear a call that is sufficiently loud, to
estimate its duration and frequency and then begin calling in synch
with the detected call.
"Although this behavioral algorithm is quite simple, it produces some interesting group behaviors," Frampton reports.
When all is quiet and an alpha node begins calling, at first only those
beta nodes nearby hear the call and respond. Then, as more betas swell
the chorus, nodes farther away hear the call and join in. In this
fashion, synchronized calling gradually spreads concentrically out from
the alpha node until all the nodes are synchronized.
A second interesting behavior occurs when a beta node "hiccups" and
starts buzzing out of synch with its neighbors. Such hiccups can be
caused by measurement noise, operating system jitter and other factors.
Occasionally, when such a hiccup occurs, neighboring nodes
resynchronize to the errant node. Normally, these transients quickly
disappear as the wayward group resynchronizes with the larger group.
The most interesting behavior pattern appeared when the researchers
introduced a third kind of node that they labeled omega. This node was
programmed identically to an alpha node but set to a different duration
and frequency. When introduced into the array, an omega node begins to
attract neighboring nodes to its call cycle. Unlike the hiccup case,
however, the omega group does not resynchronize with the original
group. Rather, the omega node eventually recruits a growing number of
nodes to its calling cycle until a "balance of power" is reached with
the alpha node. The eventual balance between the two groups depends
strongly on the initial arrangement of the sensors.
"While this is a rather whimsical application of a sensor network, it
demonstrates the unique system behaviors that can arise in truly
distributed processing," says Frampton. Even when nodes follow very
simple rules, the behavior of the group can be quite complex. Although
this project is not likely to improve knowledge on synchronized calling
in nature, it does demonstrate the types of complex behavior patterns
that will be important for future developments in sensor networks,
Frampton says.
For more news about Vanderbilt research, visit Exploration, Vanderbilt’s online research magazine, at www.exploration.vanderbilt.edu.
Media contact: David F. Salisbury, (615) 343-6803
david.salisbury@vanderbilt.edu