Elaboração de dataset para detecção de ataque DDoS

Abstract

The Distributed Denial of Service (DDoS) attack involves sending a large volume of requests to a specific target, overwhelming it and resulting in system crashes and slowness due to the excessive load of requests. Intrusion Detection System (IDS) tools, such as Snort, can assist in detecting this type of attack by allowing the analysis of the entire existing network traffic. However, these tools are limited to rules of previously known attacks, which hinders their ability to identify new attacks. The objective of this work is the creation of a dataset containing attack requests and normal traffic in the Internet Control Message Protocol (ICMP), collected and stored by Snort. The Random Tree classification algorithm is used to evaluate its effectiveness in identifying attacks. The results obtained demonstrate the dataset’s effective detection of a DDoS attack.