ML-Based SCG Coverage Prediction for 5G NR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 5G NR technology faces challenges in accurately determining whether a user equipment (UE) is within the coverage area of a secondary cell group (SCG), leading to inefficient SCG measurement configurations and potential reductions in LTE throughput and increased battery power consumption.
Innovation Solution
Implementing a machine learning (ML) model trained on master cell group (MCG) information and historical UE data to indicate SCG coverage, allowing UEs and base stations to communicate effectively based on ML model indications, thereby optimizing SCG measurement configurations and reducing unnecessary measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional SCG measurement configuration methods are used, then measurement coverage can be determined, but accuracy is insufficient and trial-and-error methods are required
Solution Approach 1:
The patent applies preliminary action by using machine learning models to predict SCG coverage status before actual measurements are performed. The system pre-determines whether a UE is likely to have SCG coverage based on trained models, allowing the network to proactively configure measurements only when necessary, thereby improving accuracy while reducing the trial-and-error complexity of traditional methods.
Solution Approach 2:
The patent introduces machine learning models as intermediary components between the network and measurement processes. These models act as mediators that process MCG information and historical data to predict SCG coverage status, enabling more accurate determination without requiring complex trial-and-error measurement configurations.
2Measurement precision
If SCG measurements are performed frequently to ensure accurate coverage determination, then coverage accuracy improves, but battery power consumption increases
Solution Approach 1:
The system performs preliminary coverage assessment using machine learning models before initiating actual SCG measurements. By predicting coverage status in advance based on MCG information and historical data, the network can avoid unnecessary measurements when coverage is unlikely, thus reducing battery power consumption while maintaining measurement accuracy when needed.
Solution Approach 2:
The patent applies partial action by performing measurements only when predicted to be necessary, rather than continuously or excessively measuring. The machine learning model filters out unnecessary measurement scenarios, allowing the system to conduct partial measurements only when coverage determination is actually needed, thereby reducing energy consumption.
3Measurement precision
If trial-and-error methods are used for SCG measurement configuration, then coverage determination can be achieved, but LTE throughput is reduced
Solution Approach 1:
The patent uses machine learning models to perform preliminary coverage assessment before initiating SCG measurements. This allows the network to determine in advance whether measurements are necessary, avoiding the trial-and-error process that causes delays and reduces LTE throughput, while still achieving accurate coverage determination when measurements are performed.
Solution Approach 2:
The machine learning models serve as intermediaries that eliminate the need for trial-and-error measurement configurations. By predicting SCG coverage status based on trained models, the system can directly determine coverage accuracy without the time-consuming iterative process that impedes LTE throughput.
Data Source
AI summary
A UE may receive information associated with an MCG. The UE may train, based on at least one of the information associated with the MCG or historical information of the UE for an SCG, an ML model that indicates whether a location of the UE is within a coverage area of the SCG. The UE may communicate with a base station based on an indication of the ML model. The indication of the ML model may be indicative of whether the UE is within the coverage area of the SCG.