DEVELOPMENT OF FRAMEWORK FOR DDOS ATTACK DETECTION IN NETWORK DEVICES USING MACHINE LEARNING

Abstract

The exponential growth of internet users poses a significant challenge to safeguarding online resources against security threats. The escalating frequency of Denial of Service (DoS) attacks further intensifies these concerns, underscoring the urgent need for sophisticated cyber-defense mechanisms. Addressing this imperative, our study presents an innovative machine learning-based system engineered to detect Distributed Denial of Service (DDoS) attacks. By harnessing the predictive capabilities of Logistic Regression, K Nearest Neighbor, and Random Forest algorithms, our approach fortifies defenses against evolving cyber threats. To gauge the efficacy of our models, extensive experiments were conducted utilizing the recently updated NSL KDD dataset. The results unveil the exceptional accuracy of our proposed system in identifying DDoS attacks, surpassing prevailing state-of-the art detection methods. These findings underscore the pivotal role of our research in bolstering cyber-security resilience amidst the mounting challenges posed by the digital landscape.