Wake Forest Researchers Turn to Evolution, Ants To Combat Viruses

• By Joshua Bolkan
• 02/15/12

A team at Wake Forest University is working on a genetic algorithm to proactively discover more secure computer configurations.

Genetic algorithms are search heuristics inspired by evolution and often used for optimization. They start with a population of solutions that are evaluated for fitness and then allowed to create offspring with a chance of mutation. The offspring then become the new population of solutions in the next run of the algorithm. During each cycle of the process, solutions that are determined to be more fit are given a greater chance to create offspring.

Funded by a grant from the Pacific Northwest National Laboratory (PNNL), Computer Science Professor Errin Fulp and graduate student Michael Crouse, are trying to use such an algorithm to "improve defense mechanisms of similar computing infrastructures with minimal human interaction," according to information released by the university.

"Typically, administrators configure hundreds and sometimes thousands of machines the same way, meaning a virus that infects one could affect any computer on the same network," said Crouse. "If successful, automating the ability to ward off attacks could play a crucial role in protecting highly sensitive data within large organizations."

Fulp explained that most attacks occur after a reconnaissance phase during which a virus or other threat looks for defenses and the best way into a system. But when the threat returns later, if any thing has changed, it's much less likely to attack.

"Just as one might try to prevent a home robbery, our goal is to create a 'moving target defense' that detects cyber threats when they first case the house," said Fulp. "If we can automatically change the landscape by adding the technological equivalent of security cameras or additional lighting, the resulting uncertainty will lower the risk of attack."  

In a separate PNNL project, Fulp and Crouse have turned to ants for help in combating viruses that attack the nation's electrical grid.

"The idea is to deploy thousands of different types of digital ants, each looking for evidence of a threat," Fulp said. "As they move about the network, they leave digital trails modeled after the scent trails ants in nature use to guide other ants. Each time a digital ant identifies some evidence, it is programmed to leave behind a stronger scent. Stronger scent trails attract more ants, producing the swarm that marks a potential computer infection."

"Fulp is also currently leading a group of Wake Forest faculty in creating an academic center dedicated to the study and teaching of bio-inspiration and biomimicry," according to information released by Wake Forest.

"Increasingly, conventional solutions are unable to scale to the information or processing required by the massive influx of data. Furthermore, the complexity of many computer systems will necessitate solutions that are more scalable, robust, and flexible than traditional methods can offer," says Fulp. "Designs found in nature can serve as a source of inspiration, providing robust and efficient methods that are better suited to address these complex problems."