Network Slice PRB Re-Balancing for Variable 5G Traffic

The patent implements dynamic PRB configuration that automatically adjusts resource allocation based on real-time traffic patterns, time of day, location, and user types. This transforms the static configuration into a dynamic system that adapts to varying network conditions, thereby improving radio spectrum utilization efficiency without requiring manual reconfiguration and maintaining manageable complexity through automated algorithms.

The system employs self-service mechanisms where network slices automatically monitor their own traffic patterns and autonomously adjust their PRB allocations based on predefined policies and machine learning models. This eliminates the need for manual configuration while maintaining high adaptability to varying traffic conditions, as the system serves itself by making real-time optimization decisions.

The patent implements feedback mechanisms where the system continuously monitors traffic patterns, PRB utilization metrics, and network performance, then uses this feedback to automatically adjust configurations. Machine learning models analyze the feedback data and generate optimization decisions, creating a closed-loop control system that improves spectrum utilization while managing complexity through automated decision-making based on real-time network state.

The system applies partial automation by using machine learning models selectively for traffic pattern prediction and PRB allocation optimization, rather than attempting to automate all network control functions. The models focus specifically on predicting future traffic demands and suggesting optimal PRB configurations, leaving other network management tasks to traditional control mechanisms. This targeted approach achieves high prediction accuracy while limiting computational complexity to only the necessary predictive functions.