AI Model for CSI Feedback Quantization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current machine learning-based channel state information (CSI) feedback techniques require separate models for each quantization scheme and CSI reporting payload size, limiting scalability and efficiency in mobile communication systems.
Innovation Solution
A method using a two-sided machine learning model that supports various quantization schemes and CSI reporting payload sizes with a single model, enabling efficient compression and decompression of CSI feedback information through vector and scalar quantization configurations, and allowing for training of AI models to adapt to different quantization schemes and payload sizes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If separate trained models are used for each quantization scheme and CSI reporting payload size, then CSI feedback accuracy is maintained, but device complexity and system scalability deteriorate due to requiring multiple models
Solution Approach 1:
The patent implements a universal AI model architecture that can handle multiple quantization schemes (uniform, non-uniform, vector quantization) and various CSI reporting payload sizes through a single model. The model uses configurable parameters to adapt to different quantization configurations without requiring separate trained models for each scheme, thereby reducing device complexity while maintaining CSI feedback accuracy.
2Measurement precision
If the entire channel information is transmitted, then CSI feedback accuracy is improved, but the amount of transmitted information increases causing radio resource overhead
Solution Approach 1:
The patent extracts only the essential channel state information features using an AI-based compression approach. Instead of transmitting entire channel information matrices, the system uses neural networks to identify and transmit only the most critical channel characteristics, thereby reducing the volume of transmitted information while preserving the accuracy needed for effective CSI feedback.
Solution Approach 2:
The patent changes the representation parameters of channel information by transforming detailed channel matrices into compressed feature vectors through AI processing. The system adjusts the information representation from high-dimensional channel state data to low-dimensional essential features, reducing transmitted information volume while maintaining feedback accuracy through intelligent parameter transformation.
3Adaptability or versatility
If a single AI model supports multiple quantization schemes, then system scalability is improved, but the difficulty of training and configuring the model increases
Solution Approach 1:
The patent implements a dynamic model configuration system where the AI model can adapt its parameters and processing modes based on the selected quantization scheme. The system dynamically adjusts its internal operations to match the requirements of uniform quantization, non-uniform quantization, or vector quantization without requiring complete retraining, thereby reducing training complexity while maintaining multi-scheme support capability.
Solution Approach 2:
The patent performs preliminary configuration of the AI model with multiple quantization scheme capabilities during model initialization. The model is pre-configured with knowledge of different quantization approaches and can select the appropriate processing mode in advance based on system requirements, reducing the complexity of real-time adaptation and simplifying the training process compared to having to train separate models for each scheme.
Data Source
AI summary
A method of a terminal may comprise: receiving a channel state information (CSI) request message from a base station, the CSI request message including first indication information indicating one of quantization configuration information of a first artificial intelligence (AI) model configured in the terminal to encode and transmit CSI; quantizing generated CSI feedback information based on the first indication information; and transmitting a CSI report message including the quantized CSI feedback information to the base station.