Abstract not found

There is often the need to update an installed Intrusion Detection System (IDS) due to new attack methods or upgraded computing environments. Since many current IDSs are constructed by manual encoding of expert security knowledge, changes to IDSs are expensive and slow. In this paper, we describe a data mining framework for adaptively building Intrusion Detection (ID) models. The central idea is to utilize auditing programs to extract an extensive set of features that describe each network connection or host session, and apply data mining programs to learn rules that accurately capture the behavior of intrusions and normal activities. These rules can then be used for misuse detection and anomaly detection. Detection models for new intrusions or specific components of a network system are incorporated into an existing IDS through a meta-learning (or co-operative learning) process, which produces a meta detection model that combines evidence from multiple models. We discuss the strengths...

Intrusion detection (ID) is an important component of infrastructure protection mechanisms. Intrusion detection systems (IDSs) need to be accurate, adaptive, and extensible. Given these requirements and the complexities of today’s network environments, we need a more systematic and automated IDS development process rather than the pure knowledge encoding and engineering approaches. This article describes a novel framework, MADAM ID, for Mining Audit Data for Automated Models for Intrusion Detection. This framework uses data mining algorithms to compute activity patterns from system audit data and extracts predictive features from the patterns. It then applies machine learning algorithms to the audit records that are processed according to the feature definitions to generate intrusion detection rules. Results from the 1998 DARPA Intrusion Detection Evaluation showed that our ID model was one of the best performing of all the participating systems. We also briefly discuss our experience in converting the detection models produced by off-line data mining programs to real-time modules of existing IDSs. Categories and Subject Descriptors: C.2.0 [Computer-Communication Networks]: General—Security and protection (e.g., firewalls); C.2.3 [Computer-Communication Networks]:

This paper presents a taxonomy of intrusion detection systems that is then used to survey and classify a number of research prototypes. The taxonomy consists of a classification first of the detection principle, and second of certain operational aspects of the intrusion detection system as such. The systems are also grouped according to the increasing difficulty of the problem they attempt to address. These classifications are used predictively, pointing towards a number of areas of future research in the field of intrusion detection. 1 Introduction There is currently a need for an up-to-date, thorough taxonomy and survey of the field of intrusion detection. This paper presents such a taxonomy, together with a survey of the important research intrusion detection systems to date and a classification of these systems according to the taxonomy. It should be noted that the main focus of this survey is intrusion detection systems, in other words major research efforts that have resul...

The Intrusion Detection System architectures commonly used in commercial and research systems have a number of problems that limit their congurability, scalability or efficiency. The most common shortcoming in the existing architectures is that they are built around a single monolithic entity that does most of the data collection and processing. In this paper, we review our architecture for a distributed Intrusion Detection System based on multiple independent entities working collectively. We call these entities Autonomous Agents. This approach solves some of the problems previously mentioned. We present the motivation and description of the approach, partial results obtained from an early prototype, a discussion of design and implementation issues, and directions for future work.

Network-based attacks are becoming more common and sophisticated. For this reason, intrusion detection systems are now shifting their focus from the hosts and their operating systems to the network itself. Network-based intrusion detection is challenging because network auditing produces large amounts of data, and dierent events related to a single intrusion may be visible in dierent places on the network. This paper presents a new approach that applies the State Transition Analysis Technique (STAT) to network intrusion detection. Network-based intrusions are modeled using state transition diagrams in which states and transitions are characterized in a networked environment. The target network environment itself is represented using a model based on hypergraphs. By using a formal model of both the network to be protected and the attacks to be detected the approach is able to determine which network events have to be monitored and where they can be monitored, providing automatic suppo...

An articial immune system (ARTIS) is described which incorporates many properties of natural immune systems, including diversity, distributed computation, error tolerance, dynamic learning and adaptation and self-monitoring. ARTIS is a general framework for a distributed adaptive system and could, in principle, be applied to many domains. In this paper, ARTIS is applied to computer security, in the form of a network intrusion detection system called LISYS. LISYS is described and shown to be eective at detecting intrusions, while maintaining low false positive rates. Finally, similarities and dierences between ARTIS and Holland's classier systems are discussed. 1 INTRODUCTION The biological immune system (IS) is highly complicated and appears to be precisely tuned to the problem of detecting and eliminating infections. We believe that the IS provides a compelling example of a massively-parallel adaptive information-processing system, one which we can study for the purpose o...

Sharing data between widely distributed intrusion detection systems offers the possibility of significant improvements in speed and accuracy over isolated systems. In this paper, we describe and evaluate DOMINO (Distributed Overlay for Monitoring InterNet Outbreaks); an architecture for a distributed intrusion detection system that fosters collaboration among heterogeneous nodes organized as an overlay network. The overlay design enables DOMINO to be heterogeneous, scalable, and robust to attacks and failures. An important component of DOMINO’s design is the use of active sink nodes which respond to and measure connections to unused IP addresses. This enables efficient detection of attacks from spoofed IP sources, reduces false positives, enables attack classification and production of timely blacklists. We evaluate the capabilities and performance of DOMINO using a large set of intrusion logs collected from over 1600 providers across the Internet. Our analysis demonstrates the significant marginal benefit obtained from distributed intrusion data sources coordinated through a system like DOMINO. We also evaluate how to configure DOMINO in order to maximize performance gains from the perspectives of blacklist length, blacklist freshness and IP proximity. We perform a retrospective analysis on the 2002 SQL-Snake and 2003 SQL-Slammer epidemics that highlights how information exchange through DOMINO would have reduced the reaction time and false-alarm rates during outbreaks. Finally, we provide preliminary results from our prototype active sink deployment that illustrates the limited variability in the sink traffic and the feasibility of efficient classification and discrimination of attack types. 1

Network-based attacks have become common and sophisticated. For this reason, intrusion detection systems are now shifting their focus from the hosts and their operating systems to the network itself. Network-based intrusion detection is challenging because network auditing produces large amounts of data, and different events related to a single intrusion may be visible in different places on the network. This paper presents NetSTAT, a new approach to network intrusion detection. By using a formal model of both the network and the attacks, NetSTAT is able to determine which network events have to be monitored and where they can be monitored.

This dissertation explores an immunological model of distributed detection, called negative detection, and studies its performance in the domain of intrusion detection on computer networks. The goal of the detection system is to distinguish between illegitimate behaviour (nonself ), and legitimate behaviour (self ). The detection system consists of sets of negative detectors that detect instances of nonself; these detectors are distributed across multiple locations. The negative detection model was developed previously; this research extends that previous work in several ways. Firstly, analyses are derived for the negative detection model. In particular, a framework for explicitly incorporating distribution is developed, and is used to demonstrate that negative detection is both scalable and robust. Furthermore, it is shown that any scalable distributed detection system that requires communication (memory sharing) is always less robust than a system that does not require communication...