Different approach to intrusion detection touted
No need to update attack signatures?
A security firm has put together two intrusion detection products to create technology it says takes a different approach to defending against hack attacks.
CentraxICE, from security integrator Articon-Integralis, is positioned as a product which provides "defence in depth" from hack attacks beyond that offered by firewalls. It is designed to defend against packet floods -- attempts to break into systems by bombarding an organisation's Web server with traffic and thereby overwhelming an organisation's defences.
The product combines CyberSafe Centrax as a host-based Intrusion Detection System (IDS), and BlackICE Sentry from Network ICE, in a single product.
Articon-Integralis said CentraxICE has a single console and claims its elements are as well integrated as a complete IDS from a single vendor, while offering the better performance needed for high-speed networks.
Richard Barber, group technical advisor at Articon-Integralis, said that CentraxICE scores over competitive products because it "uses protocol analysis techniques instead of simply comparing packets to a database of known attack signatures." He added that this approach also reduces false positives.
Kenneth De Spiegeleire, of Internet Security Systems, whose main business is in the IDS market, said the most "effective and reliable technology used in most IDS still seems to be pattern recognition, despite issues concerning false positives, which are almost unavoidable".
He was sceptical about claims that CentraxICE needed no updates on attack signatures which he compared to developing anti-virus software that never needs updating.
The IDS market is going towards providing anomaly detection, he added, but combining this signature recognition and threshold setting to produce a next-generation IDS product is still some way off. Anomaly detection or other artificial-intelligence based technologies in particular will be hard to develop.
"To do this you need to develop software that will 'interpret' captured packets or data in real-time and perform an instant evaluation whether they could contain buffer overflows, DoS packets, exploits of system configurations, and dozens of other exploit categories -- whether they have been discovered yet or not," De Spiegeleire said.
IDS, which provide security management, gather and analyse information from various areas within a computer network to identify possible security breaches.
The passive component is called host-based intrusion detection and includes inspection of the system's configuration files to detect inadvisable settings or weak passwords. The active component, network-based intrusion detection, is designed to record and provide alerts on attacks by crackers. ®