Predictive Network Node Capacity Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current self-configuration and self-optimization techniques in 5G wireless networks are reactive, responding to current network conditions and UE traffic load, leading to significant delays in addressing performance degradation, which continues during the adjustment period.
Innovation Solution
Implement predictive methods for network nodes to anticipate modifications in coverage and capacity by analyzing metrics and UE measurements, allowing proactive adjustments and coordination between nodes to prevent performance issues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If reactive self-optimization techniques are used to respond to current network conditions, then network nodes can adjust to current traffic patterns, but significant delays occur in addressing performance degradation
Solution Approach 1:
The patent implements predictive self-optimization by using machine learning models to forecast future network traffic patterns and capacity requirements. Network nodes proactively adjust coverage and capacity parameters before actual degradation occurs, eliminating the reactive delay inherent in traditional approaches. The system predicts future states based on historical data and trends, enabling preventive actions rather than corrective responses.
2Reliability
If network nodes continuously monitor and adjust coverage and capacity, then service quality can be maintained, but network complexity increases
Solution Approach 1:
The patent implements self-optimization where network nodes autonomously perform monitoring, prediction, and adjustment operations without requiring centralized control or manual intervention. Each node uses its own collected data and embedded machine learning models to make independent decisions about coverage and capacity optimization, reducing the complexity of centralized management while maintaining high service quality through distributed intelligence.
Data Source
AI summary
Embodiments include methods for a first network node of a wireless network. Such methods include determining a predicted modification in coverage and/or capacity, during a subsequent time period, of one or more of the following served by the first network node: one or more cells, and one or more reference signal (RS) beams. Such methods also include sending, to a second network node of the wireless network, a first message comprising an indication of the predicted modification in coverage and/or capacity. Other embodiments include complementary methods for the second network node, as well as network nodes configured to perform such methods.