Desenvolvimento de um IDS baseado em técnicas de Data Mining.

Abstract

We are in the digital age, where most of the electronic devices are connected, transmitting and receiving information through a network. Generally, this information is of utmost importance and value, as in the case of large companies that are constant targets of people and malicious systems, which cause losses and problems. In this way, attack detection is of utmost importance to network administrators, since many methods of protection (access control, firewall policies, and data encryption), may not be sufficient. IDS - Intrusion Detection System have become essential as a complement to the infrastructure of the current networks, as well as new detection techniques integrated with these tools. This work proposes the use of Data Mining algorithms to generate SYN Flood attack detection patterns, based on data from the KDDCup’99 dataset. These generated patterns are used for the development of an IDS to detect this type of attack. The results showed that the IDS developed was effective in detecting SYN Flood attacks.