Characteristics based Detection of Internet Worms using Combined Machine Learning Methods and Worm Containment

Abstract

With the rapid growth of Internet today many Internet based applications are evolving Internet is an open network accessed by all The major challenge in Internet is security. Many attacks newlineand vulnerabilities affect the network Among the various attacks Internet worms are vulnerable because they infect a large number of hosts within a short period of time and from that infected hosts newlinethey further initiate attacks like distributed denialofservice phishing and spyware through their propagation. Internet worms like CodeRed in 2001 Slammer in 2003 WittySasser in 2004 newlineStorm in 2007 Conficker in 2008 and StuxNet in 2010 to 2012 have created prominent damages to the hosts Within the period of five years 400000 computers got infected due to Blaster worm in 2003 newlineConficker worm damaged approximately 13 million IP addresses This number may increase year by year To overcome these damages and to defend against these attacks effective defense mechanism is necessary Therefore detection and containment of Internet worms are the need of the day