Network forensics is addressed to deal with cybercrime. The main purpose of a network forensics system is reconstructing evidences of network attacks. In order to reconstruct evidence, the network attack is firstly identified. Therefore, network attack detection solutions play an important role in network forensics. There are two main types of network attacks: network level and application level. Network level attack detection solutions focus on the information in the headers of network packets. While, application level attack detection solutions investigate the data fragments carried out in the packet payloads. We propose an approach based on Shannon entropy and machine learning techniques to identify executable content for anomaly-based network attack detection in network forensics systems. Experimental results show that the proposed approach provides very high detection rate.
|Name||2014 11th International Conference on Fuzzy Systems and Knowledge Discovery, FSKD 2014|
|Conference||11th International Conference on Fuzzy Systems and Knowledge Discovery 2014|
|Period||19/08/14 → 21/08/14|