Video Streaming Traffic Prediction Using Machine Learning Techniques and Content Delivery Networks
Hamza Majid Hamid, Rana Fareed Ghani · 2025
The Internet has undergone significant transformations in recent years, leading to the emergence of numerous innovative applications and services centered on digital content consumption. To support this evolution, advanced frameworks integrating content delivery networks (CDNs) and artificial intelligence (AI) have been developed to optimize content delivery and enhance user experiences. These frameworks leverage machine learning models to improve traffic prediction, bandwidth management, and network optimization for video streaming platforms. The proposed system preprocesses video streaming data and employs AI to analyze traffic patterns during peak hours. By identifying underutilized channels and redirecting traffic to less congested routes, the system ensures faster content delivery, reduces delays, and prevents bottlenecks during high-demand periods. The results demonstrate significant improvements in prediction accuracy and reduced latency. The performance metrics for different models are as follows: in machine learning models, including Decision Tree (DT), Random Forest (RF), and Naïve Bayes (NB), for traffic prediction in CDN networks. DT achieved the highest accuracy of 99.36% with a testing time of 0.5907 seconds and a decision time of 0.018 milliseconds in P2P, proving its effectiveness in optimizing traffic distribution and enhancing network performance. These findings highlight the system's ability to address video streaming challenges, particularly during peak traffic periods, providing.