Predicting DoS/DDoS Attacks Using Deep Learning Models

Abstract

Distributed Denial of Service (DDoS) and Denial of Service (DoS) attacks are top cybersecurity threats that compromise the performance and security of computer networks. These malicious attacks aim to overload or block access to network resources for legitimate users, leading to significant disruptions in organizational operations and negatively influencing user experiences. Deep learning techniques present an innovative solution against the growing menace of DoS and DDoS attacks, offering promising capabilities in identifying complex patterns within large datasets for effective attack detection and mitigation. Our research project is directed towards creating an advanced detection system capable of accurately distinguishing between legitimate and malicious traffic, thereby providing a robust defense against the disruptions caused by DoS and DDoS attacks. Utilizing cutting-edge deep learning methodologies, including Long Short-Term Memory (LSTM) networks and Convolutional Neural Networks (CNN), we aim to bolster cybersecurity by strengthening network resilience in the face of ongoing cybercriminal threats.