Botnets detection for keeping the security of computer systems based on fuzzy clustering

Abstract

Botnets have been detected as the most important internet threat in recent years which are developing and spreading constantly. Botnets detection is a new and challenging research domain in security section of computer nets. Because detection of an attack isn't considered as a normal situation or a definite Botnet attack and we can't decide definitely, therefore this article intends to count each intrusion with one degree of attack and give the action initiative to the organization for regulating the sensitiveness measure for the attack of intrusions perception. Also in this research a combined approach based on evolutional algorithm of colonial competition and fuzzy clustering (fuzzy C-Mean) has been presented in order to detect Botnet. For all simulations, programming in MATLAB 2013 B environment has been used. The used Botnet data collection in this article was MCFP. The results indicate the superiority of suggested method over similar basic methods.