Abstract:
Presented work focuses onto proposal, implementation and evaluation of the new method for detection and type identification of SYN flood (DoS) attacks. The method allows distinguishing type of detected SYN flood attacks – random, subnet or fixed. Based on Counting Bloom filter, the attack detection and identification algorithm is proposed, implemented and evaluated in KaTaLyzer network traffic monitoring tool. Proof of correctness of the approach for TCP SYN flood attack detection and type identification is provided – both in practical and theoretical manners. In practice, new module for KaTaLyzer is implemented and TCP attacks are detected, identified and network administrator is notified about them in real-time.
