HYBRID LSTM-CNN FRAMEWORK TO DETECT AND MITIGATE DDOS ATTACKS IN CLOUD INFRASTRUCTURE

Abstract

Distributed Denial of Service (DDoS) attacks remain a critical threat to cloud infrastructure, targeting availability by overwhelming servers with malicious traffic. Traditional detection methods struggle with scalability and real-time response in dynamic cloud environments. This paper proposes an AI-based framework integrating machine learning (ML) and deep learning (DL) models to detect and mitigate DDoS attacks in real time. We leverage anomaly detection, traffic pattern analysis, and automated mitigation strategies to secure cloud resources. Experimental results using datasets like CICDDoS2019 demonstrate a detection accuracy of 98.7% with a false-positive rate below 2%. The framework also integrates mitigation mechanisms such as traffic filtering and resource scaling using cloud-native tools (e.g., AWS Shield, Kubernetes). This work advances cloud security by combining explainable AI (XAI) models with adaptive mitigation policies. This paper explores the possibility of the combination of Long Short-Term Memories (LSTM) and Convolutional Neural Network (CNN), framework (ie. an AI driven model), to diminish the probability chance of Distributed Denial of Service Attack (DDoS) in Cloud servers.