Cautious virus detection in the extreme

  • Authors:
  • John Case;Samuel E. Moelius, III

  • Affiliations:
  • University of Delaware;University of Delaware

  • Venue:
  • Proceedings of the 2007 workshop on Programming languages and analysis for security
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

It is well known that there exist viruses whose set of infected programs is undecidable. If a virus detector is to err on the side of caution with respect to such a virus, then it must label some perfectly innocent programs as being infected by the virus. Can there exist a virus whose set of infected programs is so unwieldy that any cautious virus detector must label all but finitely many programs as being infected by the virus â聙聰 even when infinitely many programs are not infected by the virus? Although such viruses can exist, strong theoretical evidence is presented that such a virus is unlikely to be encountered in the real world. Several of our proofs employ infinitary self-reference arguments