Traffic analysis and resource adaptation in large-scale 5G multi-layer edge networks

In this research, we propose automating network management through data-driven intelligence,\r\nwith a particular focus on anomalies and network traffic during specific events or periods.\r\nWe analyze a large dataset collected by Orange mobile network operator in France with the\r\ngoal of forecasting mobile demand for different classes of services. To model the underlying\r\nnetwork infrastructure, we introduce a model for the underlying network based on a hierarchy\r\nof virtualization layers and slices. Building on this model, we propose algorithms to optimize\r\nthe resources allocated to network slices and traffic distribution within the operator’s network.\r\nNetwork performance is evaluated as the fraction of time the mobile traffic is within the\r\ncapacity of the network. Our results demonstrate that dynamic reallocation of resources among\r\nslices, and dynamic load balancing (traffic shaping) between nodes notably improves network\r\nperformance. These results provide insights into critical aspects related to future 5G network\r\nmanagement.