Journal of Computer Science 7 (7): 1094-1104, 2011
ISSN 1549-3636
© 2011 Science Publications
1094
Towards an Integrated Intrusion Detection Monitoring in
High Speed Networks
Hassen Sallay
Department of Computer Sciences,
College of Computer and information Sciences
Imam Mohamad ibn Saud University, Riyadh, Saudi Arabia
Abstract: Problem statement: Security Management has become a critical aspect for large scale
distributed systems. Particularly, recent Distributed Intrusion Detection Systems (DIDS) schemes in
High Speed Networks (HSN) have raised new serious management problems and challenges.
Increasing the effectiveness of IDS monitoring is primordial to satisfy the restrictive constraints in
such large multi-domains environment for narrow context of HSN. Approach: We consider the
intrusion detection monitoring as a two facets entity: one at local level (single domain) and another at
the global one (multi-domains). Through the local level, evolution of single domain intrusion detection
process (vulnerability data collection, alert generation and sensor configuration according to some
improvement scenarios) can be monitored. The global level represents evolution of multi-domain
intrusion detection process as well as the eventual security defending process through overall network
(policy generation, load balancing operations and global alert correlation). Differentiating these two
facets, leads to the design of a scalable intrusion detection management solution. Results: The
effectiveness of DIDS management in HSN had been studied and an IDS scalable monitoring
architecture for multi-domains had been proposed. Several scenarios of Snort IDS showed an
improvement on the performance of real-time detection. An integration of a set of tools provided a
convivial IDS monitoring platform. Conclusion: To satisfy the constraints of Intrusion detection
process in term of real-time and efficiency in HSN we need to monitor efficiently the IDS process. In
this context, the management framework outlined is more appropriate, convenient and efficient. The
herein proposed architecture, the snort IDS improvement techniques and the integrated platform
played a crucial role in improving of IDS real-time monitoring.
Key words: Intrusion Detection Systems (IDS), high speed networks, management architecture,
PBNM monitoring, Snort Benchmarking, Integrated monitoring
INTRODUCTION
The goal of security management is to control the
access to sensitive information and resources based on
security policies to prevent and defend against
intentional or unintentional attacks in the network.
Among the well-known network defending techniques
stay the Intrusion Detection Systems (IDS). The IDS
scans the incoming or outgoing network traffic in order
to detect the malicious or suspicious activities. As
networks become faster such that the High Speed
Networks (HSN), there is a need for IDS to perform
security analysis techniques that can keep up with the
increased network throughput otherwise it becomes a
network bottleneck. Efficient management of
Distributed IDS (DIDS) is both a crucial requirement
and a major challenge for security services. Building
management solutions which can address scalability,
efficiency and real-time constraints for IDS is a key to
their successful deployment. But, due to the differences
between high speed networks and usual networks, the
potentially huge number of packets evolved over time, the
design and implementation of the afore mentioned
management services are much more complex to achieve.
Related work: Several efforts are driven in the
literature in the context of intrusion detection process.
(Wang and Liu, 2008) establishes a whole work model
on the basis of intrusion detection techniques and
proposes to establish a data warehouse of intrusion
detection to provide an efficient and stable system.
(Roschke et al., 2009) presents an extensible IDS
management architecture to manage security event and
correlation. It is based on the virtualization concept
which integrates and handles different types of sensors
or collects and synthesizes alerts generated from
multiple hosts located in a loosely coupled
environment. Yu et al. (2004), an intrusion alert
management system based on a collaborative
architecture design for multiple intrusion detection