DDoS Attack Intrusion Detection System with CNN and LSTM Hybridization

Distributed Denial of service (DDoS) threats are among the most harmful events to network security. DDoS attacks are considered among the most common types of network attacks. When servers get attacked by these attacks., users experience issues when they try to access those servers for assistance. This caused the need for an effective means of identifying DDoS attacks., which is essential. Research has established that deep learning and machine learning are valuable methods for detecting DDoS attacks. The model has been verified employing the NSLKDD dataset. The seven levels of this type of development enable enhanced performance compared to LSTM and standard CNN. Modules for recognizing and handling defects have been included in this system. Convolutional Neural Network (CNN) and Long Short-Term Memory (CNN-LSTM) techniques are implemented in the anomaly detection model to recognize irregular patterns in traffic effectively. The proposed model utilizes flow rule directions from the controller to detect irregular traffic. It employs IP tracing to connect back to the attacker.