AI Model Selection for Channel Estimation in 5G UEs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing complexity of channel environments and user equipment mobility in beyond 5th generation mobile communication systems poses challenges for effective channel estimation, as existing AI models are not adaptable to varying scenarios.
Innovation Solution
The proposed solution involves an information processing method where user equipment (UE) determines the appropriate AI model for channel estimation based on the DMRS pattern, with configuration information sent by the base station indicating the number and type of AI models corresponding to each DMRS pattern.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If one same AI model is used to perform channel estimation in all scenarios, then the device complexity is reduced, but the adaptability to varying channel environments and mobility conditions deteriorates
Solution Approach 1:
The patent segments the channel estimation process by dividing it into multiple AI models, each specialized for specific DMRS patterns. Instead of using a single general-purpose AI model, the system creates multiple specialized models (first AI model for first DMRS pattern, second AI model for second DMRS pattern, etc.), allowing each model to be optimized for its specific pattern while maintaining overall system adaptability.
Solution Approach 2:
The patent implements dynamic selection of AI models based on the detected DMRS pattern. The system dynamically determines which AI model to use by detecting the DMRS pattern first, then selecting the corresponding specialized AI model. This dynamic adaptation allows the system to respond to varying channel environments and mobility conditions without increasing the complexity of individual models.
2Adaptability or versatility
If multiple AI models are introduced to handle different DMRS patterns, then the adaptability to diverse scenarios is improved, but the device complexity increases
Solution Approach 1:
The patent applies local quality by making each AI model specialized for a specific DMRS pattern rather than creating one complex universal model. Each AI model has local expertise in handling its designated pattern, which reduces the complexity burden on individual models while collectively providing comprehensive coverage for diverse scenarios.
Solution Approach 2:
The patent introduces an intermediary detection mechanism that identifies the DMRS pattern and selects the appropriate AI model. This intermediary layer (pattern detection and model selection process) manages the complexity of having multiple AI models by providing a systematic way to choose the right model, preventing the system from becoming unwieldy despite supporting multiple specialized models.
3Measurement precision
If AI methods are used for channel estimation in high mobility scenarios, then the measurement precision is improved, but the reliability deteriorates due to high variability in channel environment
Solution Approach 1:
The patent changes the parameter of AI model selection based on the detected DMRS pattern and channel conditions. By adjusting which AI model is deployed according to the specific DMRS pattern and mobility scenario, the system maintains measurement precision across varying conditions while ensuring reliability through appropriate model matching rather than relying on a single model in all situations.
Data Source
AI summary
An information processing method, a communication device, and a storage medium. The information processing method is performed by a UE, and includes: according to a number of AI models corresponding to a DMRS pattern, using the AI model corresponding to the DMRS pattern to perform channel estimation.