AI-Based CSI Reporting for Low-Loss 6G Channel Feedback

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing CSI reporting methods in wireless communication systems, such as codebook-based approaches, suffer from information loss and resource inefficiency, particularly in high-data-rate and low-latency scenarios like 6G communication systems.

Innovation Solution

Implementing an AI-based autoencoder model for compressing and reconstructing CSI data, with joint training of encoding and decoding models at the UE and BS, and adaptive updating of the AI model based on environmental changes to maintain performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If codebook-based CSI reporting methods are used, then channel state information can be transmitted, but information loss occurs and resource efficiency deteriorates

Engineering Contradiction:

ImproveCSI information lossVSAvoidresource efficiency

Core Design Contradiction:

Loss of informationVSProductivity

Solution Approach 1:

The patent replaces traditional codebook-based mechanical CSI reporting with AI-based neural network encoding/decoding. The UE uses a trained neural network encoder to compress CSI data, and the BS uses a corresponding decoder to reconstruct it, eliminating the need for predefined codebooks and reducing information loss while improving resource efficiency.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the fundamental parameter of CSI representation from discrete codebook indices to continuous neural network encoded vectors. This allows for more precise channel state representation and reduces quantization loss, while the learned compression achieves better resource efficiency.

Inventive Principle:

Principle #35Parameter changes

2Productivity

If traditional CSI reporting methods are used, then communication can proceed, but performance deteriorates in high-data-rate and low-latency scenarios

Engineering Contradiction:

Improvedata transfer rateVSAvoidperformance in 6G scenarios

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent performs preliminary training of the AI models offline before actual communication. The neural network encoder and decoder are pre-trained on channel characteristics data, so during real-time communication, the system can directly use the trained models without requiring online training, thus achieving high performance in 6G scenarios while maintaining reliability.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The AI-based CSI reporting system is self-adaptive to channel characteristics. The neural networks automatically learn and adapt to different channel conditions during the training phase, enabling the system to maintain high performance across varying data rates and latency requirements without manual intervention.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS20260005742A1Method and apparatus for reporting channel state information in wireless communication system

Publication Date: 2026.01.01 SAMSUNG ELECTRONICS CO LTD

AI summary

Provided is a 5th-generation (5G) or 6th-generation (6G) communication system for supporting higher data rates after the 4th-generation (4G) communication system such as long term evolution (LTE). According to an embodiment of the disclosure, a user equipment (UE) may compress channel state information (CSI) data estimated between a base station (BS) and the UE through an artificial intelligence (AI) model for encoding trained in association with an AI model for decoding in the BS, transmit the compressed CSI data to the BS, transmit, to the BS, the CSI data on which the compression is not performed, based on a preset condition, and receive information for updating the AI model for encoding when the BS determines to update the AI model for encoding based on the CSI data.