Role of Machine Learning Ensemble in DDoS Intrusion Detection

The most common attacks over the last ten years have been Distributed Denial of Service (DDoS) attacks. In order to combat these attackers' novel DDoS attack patterns and techniques, a Network Intrusion Detection System (NIDS) should be smoothly configured. In this paper, we provide a NIDS can recognize existing and novel DDoS attack types. Our NIDS key feature is how using ensemble models, it integrates various classifiers with the idea that each classifier can concentrate on a particular type of intrusion. This creates a more effective defense against intrusions. Additionally, we analyse DDoS attacks in-depth and validate the reduced feature set using this domain knowledge to significantly improve accuracy. The NIDS we propose can successfully identify 99.2% of DDoS attacks after testing it with the NSL-KDD dataset with a smaller feature set. We compare our findings to those of other methods. Our NIDS has the flexibility to accommodate novel DDoS attack patterns.