Intrusion prevention

Intrusion prevention, at its most basic, requires some way of recognizing intrusions in real time, of being able to handle both known and unknown types of attacks, and then having a way of blocking the incursions. This can happen in two ways:

* Network-based intrusion prevention puts a device directly on the network in a critical data path that inspects all the traffic allowed through by the firewall, which is the first line of perimeter defense. Those prevention products use various methods to spot trouble, such as looking for the characteristic signatures of known viruses or comparing the current traffic to a baseline of normal traffic behavior. If the devices detect anomalies, they block the traffic from continuing onto the network.

* Host-based intrusion prevention places intelligent agents on each host computer or server to flag intrusions by comparing the behavior of systems against

expected norms. If deviations occur, the systems then have some way of blocking the procedures that are causing the anomalous behavior without affecting the machine's normal operations.