Network Anomaly Detection Through The Use Of AI Based Technologies

Abstract

Networking is one of the most fundamental aspect of computer infrastructure along with Servers and Storage. The impact of Artificial Intelligence on Computer Networking has been profound from the early days (Mistry et al., 2024) of networking.
Artificial intelligence is used to make the network more efficient (Umoga et al., 2024)and effective. Cloud computing-based analysis with AI for Edge computing (Umoga et al., 2024) is analysed with advanced AI and analytics methods. The growth and importance of networking as a domain in the past decade has coincided with the explosion of AI technologies.
This has led to building AI for networking as well as networking for AI as two separate adjacencies. In this work AI for networking is examined with focus on classical and generative AI based technologies for network traffic classification and anomaly detection. The results are also compared with traditional methods like neural network based as well as classical statistical methods for anomaly detection methodologies. The work aims to provide the benefit of AI based technologies for intrusion detection and prevention which can be used to build a secure and robust network. The work looks at different class of machine learning technologies with multiple class of traffic and provides valuable insights. As part of this work close to twenty different machine learning algorithms along with Ten different publicly available dataset and provides the best combination for network traffic classification and anomaly detection. The research will provide notable insights to build a system for network anomaly detection as well as intrusion detection for the next generation of large scale and complex networks.