Maximum entropy based IDS using header and payload information

In this work, a new maximum entropy based intrusion detection system that combines data from payload and packet headers, is proposed. This combined method uses sniffed data from a network to detect intrusions to or any other anomaly in that network. This method is tested with DARPA 1999 intrusion detection evaluation (IDEVAL) Dataset, which is the largest dataset available including whole payloads.