Amid chaos lie the keys to building a better wireless network - at least that�s the theory a band of university researchers will spend the next five years trying to turn into a small-scale reality for the US Army.
The US$4.5 million project, led by the University of California at San Diego, will exploit the chaotic energy of lasers to drive data signals among five mobile communications units. If successful, the network will be the first demonstration of multiple point communications along a network based on the principles of chaos.
"The absolute outcome is that chaos could be organized to give us efficient use of channel capacity," said Henry Abarbanel, professor of physics at UCSD. And this could be a big help in solving the ever-present problem of finding ways to get more bandwidth, Abarbanel explained. The proposed network can potentially transmit data at hundreds of megabits per second and even faster; the only ceiling is the transmission capacity of the equipment, said Greg VanWiggeren, research assistant at Georgia Institute of Technology�s Optical Chaos Laboratory.
What makes chaos so attractive for a network is its nonlinear nature. Typically, a communications system will rely on the transmission of linear signals, a process that requires much power to reduce instances of nonlinear signals, or noise, along the network. But building a network that takes advantage of the nonlinear nature of chaos eliminates this need to dampen any noise because it is no longer a detriment, but part of the system.
So developers save money because they don�t have to tune or enhance components, and devices save on power because they don�t have to dampen noise. These aspects alone could catch the eye of communications companies, some of whom, like Qualcomm, are taking part in the UCSD project. Other university partners include University of California at Los Angeles and Stanford.
"Chaotic [energy] is intrinsically broadband and low cost," noted Abarbanel. "We are building this with off-the-shelf parts, with hundred-dollar parts and not thousand-dollar parts."
The genesis for this project is a theory that Abarbanel devised as the result of observations he made of various projects at the UCSD Institute for Nonlinear Science. The theory holds that chaotic systems could be synchronized and that applications could be built upon them.
In a wireless network, a transmitter and receiver must be operating at the same frequency for data to reach its intended destination. In the case of a chaos-based network, the energy intensity of these devices must be the same. For transmitters and receivers to fall into synch, one has to pick up the behavior of the other. For example, if two grandfather clocks were set next to each other, the oscillation of their pendulums would become the same as the vibrations from each traveled to the other through the floor. This signal is what would pull the pendulums into synch.
This same principle explains how the transmitters and receivers of a chaos-based network would synch, explained VanWiggeren, whose work with Georgia Tech physics professor Rarjarshi Roy is a fiber-optic application of Abarbanel�s theory.
The fiber-optic chaos network would handle data thusly: The sender attaches a data message onto a chaotic signal which is sent on to a receiver. If all is in synch, the receiver will split the signal, sending to one photodiode a copy of the chaotic signal and the message. A second photodiode will simultaneously receive the chaotic signal as it would appear had no message been attached. In this final step, the chaotic energy is subtracted out, and the message is left.
"This shows that we�re not restricted to sine waves [of radio waves] and zeroes and ones," explained VanWiggeren, noting that there could be privacy benefits as well.
"Chaos is faster than information, so the [chaos] signal makes it difficult to determine whether there is a message attached or not. I know only from comparing the signals at the photodiodes."
But VanWiggeren and Abarbanel are not ready to call the chaos network the next big development in cryptography; it has yet to be tested for security. For now, the main emphasis will be on seeing whether the network can work for multiple users and in the daily throes of Army maneuvers.
"We need to see how this will work outside a physics lab," Abarbanel said.
