AI Channel State Information Feedback with Accuracy Indication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The accuracy of channel state information (CSI) coded and decoded by AI network models decreases with changes in channel quality, leading to reduced accuracy in communication systems.
Innovation Solution
A method where a terminal processes channel information using an AI network model, reports target channel characteristic information, and provides assistance information to a network side device to improve the accuracy of decoded CSI, allowing the device to determine the accuracy of recovered channel information and decide whether to update the AI models or use assistance information for recovery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If AI network models are used to code and decode channel state information, then processing efficiency is improved, but accuracy decreases with channel quality changes
Solution Approach 1:
The terminal device feeds back assistance information to the network side device, including channel characteristic information and accuracy indication information. This feedback mechanism allows the network side device to adjust AI model parameters or select different models based on the reported accuracy, thereby maintaining CSI accuracy while utilizing AI for efficient processing.
Solution Approach 2:
The system dynamically adapts to changing channel conditions by allowing the network side device to adjust AI model parameters or switch between different AI models based on the accuracy indication information received from the terminal. This dynamic adjustment ensures that the AI-based processing maintains high accuracy across varying channel quality conditions.
2Loss of information
If assistance information is reported to improve decoded accuracy, then information completeness increases, but communication overhead increases
Solution Approach 1:
The system extracts only the essential assistance information needed for accurate CSI decoding, such as channel characteristic information and accuracy indication information, rather than transmitting all raw channel data. This selective extraction reduces communication overhead while providing sufficient information for the network side device to improve decoding accuracy.
Solution Approach 2:
The terminal device changes the parameter representation by encoding channel characteristics into compact assistance information with accuracy indicators. This parameter transformation reduces the volume of transmitted data while preserving the essential information needed for accurate CSI recovery, thereby reducing overhead while maintaining information completeness.
Data Source
AI summary
This application discloses methods for assisting in reporting and for restoring channel characteristic information, a terminal, and a network side device. The method includes: processing, by a terminal, first channel information into target channel characteristic information by using a first AI network model; and sending, by the terminal, the target channel characteristic information to a network side device, and sending first information to the network side device. The first information includes at least one of first indication information or target assistance information. The first indication information indicates accuracy of second channel information recovered based on the target channel characteristic information or indicates information for assisting the network side device in determining the accuracy of the second channel information, and the target assistance information is used to assist the network side device in recovering the second channel information based on the target channel characteristic information.