On random-inspection-based intrusion detection

  • Authors:
  • Simon P. Chung;Aloysius K. Mok

  • Affiliations:
  • Department of Computer Sciences, University of Texas at Austin, Austin, TX;Department of Computer Sciences, University of Texas at Austin, Austin, TX

  • Venue:
  • RAID'05 Proceedings of the 8th international conference on Recent Advances in Intrusion Detection
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

Monitoring at the system-call-level interface has been an important tool in intrusion detection. In this paper, we identify the predictable nature of this monitoring mechanism as one root cause that makes system-call-based intrusion detection systems vulnerable to mimicry attacks. We propose random inspection as a complementary monitoring mechanism to overcome this weakness. We demonstrate that random-inspection-based intrusion detection is inherently effective against mimicry attacks targeted at system-call-based systems. Furthermore, random-inspection-based intrusion detection systems are also very strong stand-alone IDS systems. Our proposed approach is particularly suitable for implementation on the Windows operating system that is known to pose various implementation difficulties for system-call-based systems. To demonstrate the usefulness of random inspection, we have built a working prototype tool: the WindRain IDS. WindRain detects code injection attacks based on information collected at randominspection points with acceptably low overhead. Our experiments show that WindRain is very effective in detecting several popular attacks against Windows. The performance overhead of WindRain compares favorably to many other intrusion detection systems.