Looking to create computer defenses that adapt well to the cat-and-mouse game played between computer users and cyber attackers, a team of researchers has turned to one of nature's most effective militias—ants. Computer scientists at Wake Forest University in Winston-Salem, N.C., and the Pacific Northwest National Laboratory (PNNL) in Richland, Wash., are studying whether software written to behave like an army of "digital ants" can successfully find and flag malicious software (or malware).

"In nature, we know that ants defend against threats very successfully," Wake Forest computer science professor Errin Fulp said in a prepared statement. "They can ramp up their defense rapidly, and then resume routine behavior quickly after an intruder has been stopped. We were trying to achieve that same framework in a computer system."

To prove that their "swarm intelligence" model could more quickly and thoroughly scan for malware, Fulp and his colleagues developed a way to divide up the process of searching for security threats across 64 computers networked together. As the digital ants sought out potential security problems, they left a digital trail of their progress, much the same way normal ants leave behind a scent that can be picked up and followed by other ants. When the researchers unleashed a worm on the network, the digital ants were able to find it.

This approach differs from conventional computer security software, which can for the most part be programmed to search only for known malware. Makers of this software often update it with descriptions of new viruses and worms, but this reactive model keeps computer users at least a step behind their adversaries. Fulp and his team hope that the sharing of information among the digital ants will lead to computer defense systems that can find malware written with slight variations in order to avoid detection.

Computer scientists are already studying programs that act like swarming ants to help alleviate telecommunications system bottlenecks. "The foraging of ants has led to a novel method for rerouting network traffic in busy telecommunications systems," Eric Bonabeau and Guy Thèraulaz wrote in an article, "Swarm Smarts," in Scientific American's 2008 special report on robots. Bonabeau is chief executive and chief scientific officer at Icosystem Corporation in Cambridge, Mass., while Thèraulaz is a research director at the Research Center on Animal Cognition of the National Center for Scientific Research (CNRS) at Paul Sabatier University in Toulouse, France.

Computer-maker Hewlett-Packard and the University of the West of England together invented a network routing technique in which antlike agents deposit bits of information, or "virtual pheromones," at telephone network nodes (or switching stations), according to Bonabeau and Theraulaz. These mark less congested areas of the network that could be used by phone companies to divert surges in traffic on the network.


Image ©iStockphoto.com/ Don Bayley