AI-Driven DDoS Mitigation at the Edge: Leveraging Machine Learning for Real-Time Threat Detection and Response
Sahil Arora, Pranav Khare, S. V. Gupta
Abstract
As cyber threat actors develop increasingly sophisticated strategies, cutting-edge cyber security is necessary for industry organizations and government agencies. A security threat model must consider these developments since they might bring new cyber dangers. Many attacks have been cropping up on the Internet as it has grown. Today's most prevalent attacks are viruses, distributed denial of service (DDoS) attacks, service interruptions, code injection, and spoofing. Several new techniques for DDoS detection have been proposed, including AI-based machine learning and deep learning tactics. The most recent developments in cybersecurity's use of AI and ML are covered extensively in this article. Using the most current dataset, CICDDoS2019, this research compares and contrasts the performance of deep learning models that identify DDoS attacks, such as RNNs and GRUs. To measure how well the model detects DDoS assaults, use evaluation measures, including recall, accuracy, precision, and F1 score. Models exhibit comparable performance on the CICDDoS2019 dataset, as shown by the experimental findings, which show an accuracy score of 99.9%. The study highlights the effectiveness of AI- driven methods in improving cybersecurity resilience in network infrastructures to new and changing edge DDoS assaults.