They certainly have a Windows XP and Vista. They are even in OS X. They, computer worms, can replicate and spread through the network and removable drives, for many years been a headache for developers of security systems. In an effort to stop worms, scientists have turned to rather exotic schemes. Professor of Computer Science, an expert in computer security Errin Falp from Wake Forest University (Wake Forest, NS, USA) talks about a new development in this area, the creators of which were inspired by ants. He said that "in nature, ants are very well may reflect an external threat. They quickly establish a sort of protective barrier, but once the offender is stopped, quickly returned to normal behavior. We tried to achieve this behavior in a computer system."

WFU Scientists have created a digital "ants" ─ utilities that migrate across the network from one computer to another in search of a threat. When someone discovers a threat, others are going to it, using the so-called "hive mind". This approach allows us to quickly identify and decontaminate dangerous files. This development could open a new era of anti-virus. Typically, anti-virus programs are large and require many resources, but small anti-virus programs are not always as reliable as we would like.

"Ant" approach is fundamentally different from anything available before. Instead of a monolithic anti-virus program, individual for each computer used by nearly 3000 kinds of ants, each of which is able to recognize a specific type of malware. Moving ant-anti-viruses on a network akin to the movements of ants in the forest, they gather on the infected computer and leave it after cleaning. To maintain a constant number of ants in the network using a "patrol" utility, and for the control of the colony, monitoring and maintenance of certain functions is the "manager" tool.

When testing the system successfully detected the worm, the researchers introduced into a computer network. Being planned larger-scale testing of this system. Researchers believe that such an approach would be most promising for large homogeneous networks (with the same hardware and software on each workstation), such as a state-owned enterprises, large corporations and universities.