Recurring Busy-Period Detection in Network Segments for RRM Scheduling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current radio resource management (RRM) systems in wireless networks do not dynamically consider busy periods, leading to disruptive reconfigurations and client disconnects due to manual or interval-based updates that fail to account for dynamic network variables and outliers.
Innovation Solution
Implement an AI-RRM system that uses machine learning to analyze network utilization data, identify reoccurring busy periods, and defer non-urgent RRM updates to these periods, incorporating a buffer to adjust for changing patterns and discarding outlier data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If RRM updates are performed based on performance triggers or intervals, then network configuration optimization is achieved, but network disruptions and client disconnects occur during high traffic periods
Solution Approach 1:
The system dynamically identifies busy periods by analyzing historical network utilization data and adapts RRM update scheduling based on detected patterns. The network controller learns from past traffic patterns to automatically adjust when to perform updates, transitioning from static interval-based scheduling to dynamic pattern-based scheduling that responds to actual network conditions
Solution Approach 2:
The system performs preliminary identification of busy periods by analyzing historical network utilization data before scheduling RRM updates. By detecting patterns in advance and predicting future busy periods, the system can proactively schedule updates during identified non-busy periods, preventing disruptions before they occur
2Reliability
If RRM updates are scheduled during non-busy periods, then network disruptions are reduced, but the system complexity increases due to pattern recognition requirements
Solution Approach 1:
The network controller performs self-learning by automatically analyzing its own historical network utilization data to identify busy period patterns. The system serves itself by generating the intelligence needed for optimal update scheduling from its own operational data, eliminating the need for external complex analysis systems
Solution Approach 2:
The system implements feedback loops where network utilization data is continuously collected, analyzed for patterns, and used to adjust future update scheduling decisions. The outcomes of updates are monitored and fed back into the pattern recognition system, allowing continuous refinement of busy period identification without manual intervention
Data Source
AI summary
The disclosed technology relates to determining a period in which a network is busier relative to other periods. The method may comprise analyzing network utilization data from a network segment of the network. The network utilization data may be collected over a first duration. The method may further comprise of determining, from the network utilization data, a baseline network utilization. The baseline network utilization may be a persistent quantification of network utilization present throughout the first duration. The method may further comprise identifying at least a reoccurring period that is characterized by network utilization data indicating consecutive network utilization that is above the baseline network utilizations; and configuring a network controller to designate the at least the reoccurring period as a busy period.