SCG Failure Prediction via ML for Traffic Redistribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In 5G non-standalone (NSA) wireless communications, secondary cell group (SCG) failures are difficult to predict, leading to inefficient data transmission and resource wastage due to high error rates, especially with millimeter wave frequencies or poor wireless environments, as existing mechanisms detect failures only after they occur, not before.
Innovation Solution
A user equipment (UE) employs a machine learning model to predict SCG radio link failures (RLF) by analyzing inputs such as block error rates, packet reordering events, and throughput metrics, allowing proactive rerouting of data traffic from the SCG to the master cell group (MCG) before actual failure occurs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data transmission continues on SCG links with deteriorating quality, then throughput may be maintained temporarily, but error rates increase and resources are wasted
Solution Approach 1:
The system performs preliminary actions by detecting early signs of SCG degradation (increased block error rates, packet reordering events) and proactively rerouting traffic to MCG before complete failure occurs. This prevents futile transmissions on degraded links while maintaining efficient resource utilization.
Solution Approach 2:
The system continuously monitors SCG link quality metrics including block error rates, packet reordering events, and throughput. This feedback mechanism enables dynamic detection of degradation patterns and triggers automated traffic rerouting decisions to maintain reliable communication.
2Reliability
If traffic is rerouted from SCG to MCG before failure, then transmission reliability improves, but network capacity utilization decreases
Solution Approach 1:
The system performs preliminary rerouting only when early degradation signs are detected, not preemptively for all SCG traffic. This selective approach maintains reliability by protecting vulnerable flows while preserving capacity utilization by keeping healthy SCG links active.
Solution Approach 2:
The system changes the routing parameter dynamically based on observed link quality metrics. When block error rates and reordering events exceed thresholds, the routing decision transitions from SCG-preferred to MCG-preferred, optimizing the balance between reliability and capacity utilization.
3Measurement precision
If machine learning models are deployed for RLF prediction, then failure detection accuracy improves, but device complexity increases
Solution Approach 1:
The machine learning model operates autonomously using inputs already collected by standard protocol stacks (block error rates, reordering events). The model self-manages prediction without requiring additional complex measurement infrastructure, reducing the complexity burden while maintaining high detection accuracy.
Data Source
AI summary
A method of wireless communication, by a user equipment (UE), includes setting up a secondary cell group (SCG) with a second radio access technology (RAT) that differs from a first RAT associated with a master cell group (MCG). The method also includes communicating wirelessly via the secondary cell group and the master cell group. The method further includes predicting a radio link failure (RLF) for the secondary cell group based on multiple inputs to a machine learning model. The method still further includes routing data transmission from the secondary cell group to the master cell group, after predicting the SCG RLF.