Reflection based Distributed Denial of Service Attack Detection System

An attacker can use a reflection amplification attack to increase the amount of destructive traffic they can generate while also hiding the source of the assault. The target is overwhelmed by this type of distributed denial-of-service (DDoS) attack, which causes system and service disruption or outage. On a regular basis, these attacks are expanding at an exponential rate. As a result, presenting an effective intrusion detection system capable of detecting these threats in a timely and efficient manner has become vital. The study offers an intrusion detection system that detects various reflection-based DDoS assaults using the CICDDoS2019 dataset. The J48 classifier is used to evaluate the proposed system, which uses InfoGain (IG) filter-based feature selection techniques. The suggested approach achieves an 80.4597 percent reduction in the dataset's original characteristics. Portmap, LDAP, NetBIOS, and MSSQL threats are all successfully and efficiently detected by the framework. The results of the recommended technique are compared to those of earlier approaches on the identical CICDDoS2019 dataset.