Some results on intrusion and anomaly detection using signal processing and NEAR system

Intrusion detection using traffic analysis can offer interesting results in its extended form of anomaly detection. Applying specific methods (special data collection, signal processing) can also lead to detection of other unwanted situations in the network. However, there is a delicate balance between rightful detection and false positives. The paper extracts some real-life examples where additional analysis beyond simple processing is required to take the proper decision.