Intrusion Detection via Artificial Immune Networks

This paper has a two related aims: (1) introducing Artificial Immune Systems as a tool for the protection of computer security, and (2) introducing FIND, a prototypical IDS (Intrusion Detection System) based on this approach. FIND inputs a database previously recorded connection logs, and uses it to generate antibodies capable of recognizing if an incoming connection request of new type as either normal or attack. The learning capabilities of a supervised neural network make FIND capable to respond to the threats of an ever-changing network environment.