Flow-based Front Payload Aggregation
Tobias Limmer, Falko Dressler · 2009
We present and discuss a new monitoring technique that we call Front Payload Aggregation (FPA). Instead of being limited to either analyzing single packets for signature-based attack detection or exploiting statistical flow information for anomaly detection, FPA combines the advantages of both approaches. Exploiting the fact that most attack signatures can be found in the very first packets of a connection, we collect payload information from these few packets (we take the first n payload Bytes) and associate it to the corresponding flow data. Thus, intrusion detection can still be performed with a high degree of confidence and the monitoring system becomes efficient w.r.t. processing performance and attack resilience.