Premium
Evaluating intrusion prevention systems with evasions
Author(s) -
Särelä Mikko,
Kyöstilä Tomi,
Kiravuo Timo,
Manner Jukka
Publication year - 2017
Publication title -
international journal of communication systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.344
H-Index - 49
eISSN - 1099-1131
pISSN - 1074-5351
DOI - 10.1002/dac.3339
Subject(s) - computer science , computer security , intrusion , intrusion detection system , intrusion prevention system , evasion (ethics) , robustness (evolution) , biochemistry , chemistry , immune system , geochemistry , gene , immunology , biology , geology
Summary Intrusion prevention systems have become a common security measure in the past 20 years. Their promise is the possibility to prevent known attacks against vulnerable, unpatched devices inside enterprise networks. However, evasion techniques that enable the attacker to evade the eye of the intrusion prevention system are a potential problem for this capability. These techniques take advantage of the robustness principle that has guided designers to create systems that will try to recreate protocol content from any input they receive. In this work, we evaluated the effectiveness of 35 well‐known evasions against 9 commercial and 1 free, state‐of‐the‐art, intrusion prevention systems. We conducted 4 experiments with one million attacks against each device. Each system lets a significant amount (0.1%‐45%) of attacks pass through unrecognized. Our results show that most existing intrusion prevention systems are vulnerable against evasions.