The detection of RCS worm epidemics

  • Authors:
  • Kurt Rohloff;Tamer Başar

  • Affiliations:
  • BBN Technologies, Cambridge, MA;University of Illinois at Urbana-Champaign, Urbana, IL

  • Venue:
  • Proceedings of the 2005 ACM workshop on Rapid malcode
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper discusses the problem of automatically detecting the existence of Random Constant Scanning (RCS) worm epidemics on the Internet by observing packet traffic in a local network. The propagation of the RCS worm is modelled as a simple epidemic. An optimal hypothesis-testing approach is presented to detect simple epidemics under idealized conditions based on the cumulative sums of log-likelihood ratios. It is shown that there are limitations on the ability of this optimal method to detect several important subclasses of RCS worm epidemics even under idealized conditions.