Evaluating the Impact of Automated Intrusion Response Mechanisms

  • Authors:
  • Thomas Toth;Christopher Kruegel

  • Affiliations:
  • -;-

  • Venue:
  • ACSAC '02 Proceedings of the 18th Annual Computer Security Applications Conference
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

Intrusion detection systems (IDSs) have reached a highlevel of sophistication and are able to detect intrusions witha variety of methods. Unfortunately, system administratorsneither can keep up with the pace that an IDS is deliveringalerts, nor can they react upon these within adequatetime limits. Automatic response systems have to take overthat task. In case of an identified intrusion, these componentshave to initiate appropriate actions to counter emerg-ingthreats. Most current intrusion response systems (IRSs)utilize static mappings to determine adequate response actionsin reaction to detected intrusions. The problem withthis approach is its inherent inflexibility. Countermeasures(such as changes of firewall rules) often do not only defendagainst the detected attack but may also have negative effectson legitimate users of the network and its services. Toprevent a situation where a response action causes moredamage that the actual attack, a mechanism is needed thatcompares the severity of an attack to the effects of a possibleresponse mechanism. In this paper, we present a networkmodel and an algorithm to evaluate the impact of responseactions on the entities of a network. This allows the IRS toselect the response among several alternatives which fulfillsthe security requirements and has a minimal negative effecton legitimate users.