Sistema de ciberdefensa dinámica basada en algoritmos evolutivos para la prevención de ataques informáticos

Abstract

Además de realizar una labor determinada de forma eficiente, los servicios informáticos deben ser capaces de evitar los ataques y de detectar los que haya. Una técnica de defensa consiste en convertirse en un objetivo móvil, que varíe el perfil de forma que los atacantes no lo reconozcan. Mediante algoritmos evolutivos trataremos de configurar diferentes servicios de forma que se maximice la diversidad, a la vez que se optimice la seguridad y las prestaciones.

Cyber attacks are one of the biggest problems for many organizations. That organizations are investing so many resources in detecting cyber-attacks, but they still have serious difficulties to prevent them. The common way to perform an attack is using a list of known vulnerabilities. Many of these vulnerabilities can be caused by a bad configuration. A simple protection against computer security threats can be implemented by properly tuning existing system software without the needed of expensive security solutions. This project shows a low-cost technique to prevent computer attacks. This technique consists of constantly modifying the configuration of a server. Using a genetic algorithm, many different possible configurations are mutated to find an optimal solution. This optimal configuration is regularly applied, so the systems information previously collected by a potential attacker is no longer effective, achieving a simple protection layer while optimizing the service configuration.