Home > Security

Computerworld - Once upon a time, network security infrastructure consisted only of a firewall deployed at the perimeter. This worked fairly well when there was limited interaction between internal and external networks, the internal users were trusted and the value of the network-available assets was limited.
However, there has been considerable change in recent years. Network-aware applications and interactions between networks have greatly increased, and while access is being granted on a greater scale to these business-critical functions, attackers and their tools have become more sophisticated.
Fortunately, many organizations have augmented their security infrastructures to handle these changes. Using a number of tools, including virus-detection systems, vulnerability-assessment scanners, and encryption- and intrusion-detection systems (IDS), companies have made an effort to both detect and prevent security threats to their networks. Early versions of these security tools had trouble detecting certain types of threats and were unnecessarily complex, IDSs in particular. Such challenges rendered IDSs difficult to deploy, frustrating to use and possible to evade.

	Brian Hernacki is an architect at the Symantec Research Lab. He can be reached at brian_hernacki@symantec.com.

How anomaly detection works
To address these shortcomings, several products now support a technique known as anomaly detection. While anomaly detection is far from new, there has been considerable confusion over what it is and how it works.
By definition, an anomaly is something that's different, abnormal or not easily classified. Therefore, the concept of anomaly detection in computer security involves the discovery of an abnormality in something (a network, host, set of users, for example) when compared against expected behavior.
One of the key differences between anomaly detection and other forms of detection is that, rather than defining what is either not allowed or bad, it defines what is allowed or good. Many traditional forms of detection rely on comparing observed behavior and noting when something known to be "bad" is seen. These are often referred to as misuse-detection systems; however, this nomenclature can be confusing since anomaly systems also detect misuse. A more accurate name, perhaps, is explicit-detection systems. These systems operate well when the number of possible bad behaviors is small and doesn't change very rapidly. However, in larger systems with greater variation, these two conditions often do not hold. It then becomes an onerous task to maintain the list of what is bad.
Anomaly detection is a more proactive system that relies on having some definition of allowed behavior and then noting when observed behaviors differ. This operates well when it's easier or more efficient to define what is allowed