Affiliations: Department of Computer Science and Engineering,
Arizona State University, USA | Department of Industrial Engineering, Arizona State University, USA
Note: [] Corresponding author: Nong Ye, Associate Professor, Department
of Industrial Engineering, Arizona State University, Box 875906, Tempe, Arizona
85287, USA. Tel.: +1 480 965 7812; Fax: +1 480 965 8692; E-mail:
nongye@asu.edu
Abstract: An intrusion into an information system tries to compromise the
security of the system. Intrusion Detection Systems (IDSs) attempt to detect
these intrusions. This paper discusses what an IDS requires from the
target information system and how the IDS detects intrusions into the target
information system. Specifically, we describe the architecture of a distributed
host-based IDS developed at the Information and Systems Assurance Laboratory,
Arizona State University. At each host machine in the information
system we install an event data collector that collects and filters data of
events from the host machine. The Centralized IDS Server receives
the processed data and sends them to Individual Technique
Servers. These Individual Technique Servers use different intrusion
detection algorithms covering both anomaly detection techniques and signature
recognition techniques. Each Individual Technique Server determines
an intrusion warning (IW) level for each event. The Centralized IDS
Server then integrates the IW levels from the Individual Technique Servers into
a composite IW level, and provides it to the security administrator.