5G Neighbor Relation Conflict Prediction via ML
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing adoption of 5G technology in cellular networks leads to more frequent changes in network environments, resulting in a higher frequency of updates to neighbor relation tables. This, combined with the smaller coverage area of 5G cells and the growing number of connected devices, causes an exponential increase in handovers, which can lead to PCI conflicts.
Innovation Solution
The implementation of an ensemble machine learning approach to predict neighbor relation conflicts by identifying new 5G cells and suggesting their addition to neighbor relation tables, while also warning of potential PCI conflicts. This involves instructing terminals to perform measurements, analyzing measurement reports, and using classification models to determine if adding new 5G cells will cause conflicts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If 5G cells are deployed to increase network capacity and coverage, then network productivity and service capability are improved, but the frequency of neighbor relation table updates and handovers increases exponentially, leading to PCI conflicts
Solution Approach 1:
The system performs preliminary actions by proactively identifying new 5G cells through measurement reports before conflicts occur. The classification model predicts potential PCI conflicts in advance, allowing the network to take preventive measures such as adjusting neighbor relation tables or modifying PCI allocation before actual conflicts arise during handovers.
Solution Approach 2:
The system implements feedback mechanisms by continuously monitoring measurement reports from terminals and using the classification model to analyze potential conflicts. The results feed back into network configuration adjustments, creating a closed-loop system that adapts to changing network conditions and prevents PCI conflicts through iterative optimization of neighbor relation tables.
2Area of stationary object
If the number of 5G cells is increased to cover the same area, then network coverage and capacity are improved, but the complexity of managing neighbor relations and predicting conflicts increases
Solution Approach 1:
The system enables self-service by automatically identifying new 5G cells through terminal measurement reports and using the classification model to predict conflicts without requiring manual intervention. The system autonomously generates predictions and provides recommendations for configuration adjustments, reducing the operational burden on network administrators while managing increasing cell densities.
Solution Approach 2:
The system replaces manual mechanical processes of neighbor relation management with an automated machine learning-based classification model. Instead of manually configuring and updating neighbor relation tables for each new 5G cell, the system uses automated measurement collection, analysis, and prediction to manage complexity, substituting human-operated mechanical processes with intelligent automated systems.
3Adaptability or versatility
If handover frequency is increased due to smaller 5G cell coverage areas, then network adaptability and service continuity are improved, but the likelihood of PCI conflicts and network instability increases
Solution Approach 1:
The system applies preliminary anti-action by predicting potential PCI conflicts before they occur during handovers. The classification model analyzes measurement reports and identifies configurations that would lead to conflicts, allowing the system to take counter-actions such as adjusting neighbor relation tables or modifying PCI assignments in advance, thereby preventing conflicts that would otherwise disrupt handover processes and network stability.
Data Source
AI summary
Neighbor relation conflict prediction is performed by operations including receiving, from a serving MCG of a terminal, a measurement report of the terminal including a plurality of signal measurements associated with an SCG represented by a PCI and an ARFCN, identifying an unlisted SCG among the plurality of signal measurements, identifying one or more nearby MCG within a threshold distance of the serving MCG, counting a number of SCG in the NRT of each nearby MCG having the PCI and the ARFCN of the unlisted SCG, applying a classification model to the counted number of SCG and an MCG-PCI-ARFCN identifier representing the serving MCG, the PCI, and the ARFCN to obtain a binary value indicating whether provisioning the unlisted SCG with the serving MCG and the plurality of nearby MCG will result in PCI conflict.