AI Beam Management Using ML-Assisted Reference Signal Reporting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing beam management methods in wireless communication systems, such as those in 3GPP NR, face inefficiencies in multiplexing radio resources due to varying requirements for different usage scenarios like eMBB, mMTC, and URLLC, leading to increased power consumption and resource overhead, particularly in beam selection and management.
Innovation Solution
Implementing artificial intelligence (AI) and machine learning (ML) models for beam management, where user equipment (UE) and base stations exchange messages and measurement results to enhance beam measurement and reporting, optimizing beam selection and reducing resource overhead through AI/ML processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional beam management methods are used in wireless communication systems, then beam selection and management can be performed, but power consumption increases and resource overhead increases due to varying requirements for different usage scenarios
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting beam management parameters based on usage scenarios. The base station configures different reference signal resource quantities and measurement parameters for eMBB, mMTC, and URLLC scenarios, optimizing the balance between beam selection accuracy and power consumption for each scenario's specific requirements
Solution Approach 2:
The patent segments beam management into scenario-specific configurations. Different usage scenarios (eMBB, mMTC, URLLC) are handled with customized reference signal resources, measurement configurations, and reporting parameters, allowing optimized power consumption and resource usage for each segment rather than a one-size-fits-all approach
2Reliability
If traditional beam management methods are used in wireless communication systems, then beam selection and management can be performed, but resource overhead increases due to varying requirements for different usage scenarios
Solution Approach 1:
The patent changes resource allocation parameters based on usage scenarios. The base station configures different quantities of reference signals, measurement parameters, and reporting configurations for eMBB, mMTC, and URLLC scenarios, reducing resource overhead by matching the configuration complexity to the actual requirements of each scenario
Solution Approach 2:
The patent segments resource allocation into scenario-specific configurations. Each usage scenario receives customized reference signal resources and measurement parameters, preventing uniform over-provisioning of resources and reducing overall resource overhead while maintaining necessary beam selection accuracy
3Productivity
If AI/ML models are implemented for beam management, then resource utilization improves and power consumption reduces, but system complexity increases due to model processing requirements
Solution Approach 1:
The patent introduces AI/ML models as intermediaries between raw measurement data and beam management decisions. The models process measurement results from reference signals and generate optimized beam selection recommendations, improving resource utilization while the base station handles the computational complexity of model training and inference
4Reliability
If AI/ML models are implemented for beam management, then beam selection accuracy improves, but device complexity increases due to model processing requirements
Solution Approach 1:
The patent uses AI/ML models as intermediary processing layers that enhance beam selection accuracy. The models take measurement data as input and produce optimized beam decisions, improving reliability while concentrating the computational complexity in the base station's model processing rather than in the UE's measurement and reporting functions
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided are a method and apparatus for providing a beam management technology using artificial intelligence and machine learning in a wireless communication system. The method may include receiving, from a base station, a message including one or more pieces of resource information for a reference signal and an indicator indicating a model process linked to the one or more pieces of resource information; deriving a measurement result by performing a measurement operation on the reference signal on the basis of the message and performing the model process linked to the indicator by using the measurement result; and transmitting, to the base station, a report message including at least one of the measurement result and model process result information.