Reproduce “A Review of Network Traffic Analysis and Prediction Techniques” on OpenResearch

About this paper

Manish Joshi, Theyazn Hassn Hadi

A comprehensive review from North Maharashtra University examines diverse network traffic analysis and prediction techniques, detailing methodologies such as data mining, time series models, soft computing, and neural networks. The work provides a structured overview of their application in network management and security, emphasizing the critical roles of preprocessing and hybrid approaches.

Key insights

• Data mining techniques, such as clustering and classification, are fundamental for network traffic analysis, especially in intrusion detection, supported by robust preprocessing.
• Hybrid models, combining different analysis or prediction techniques, consistently demonstrate superior performance by leveraging the strengths of multiple approaches.
• While linear models are applicable for stable traffic patterns, nonlinear and hybrid time series models are more effective at capturing the complex, bursty, and self-similar characteristics of modern network traffic.

Summary by alphaXiv.