CSI Codebook Indexing With Mixed Granularity Feedback

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing wireless communication systems face challenges with high CSI feedback overhead and reduced accuracy due to increased antenna dimensions and correlated signals, which hinder data throughput and increase physical dimensions, making them vulnerable to environmental effects.

Innovation Solution

A method for CSI reporting using a codebook representation with multiple sub-indices, where each matrix is indexed by a first and second sub-index with different time-frequency granularity, exploiting antenna correlations to reduce overhead and improve accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If the number of transmitter and receiver antennas is increased to meet high data throughput requirements, then the data throughput is improved, but the CSI feedback overhead increases further

Engineering Contradiction:
Improvedata throughputVSAvoidCSI feedback overhead
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The codebook is segmented into multiple subcodebooks, each representing different spatial correlation characteristics. Instead of feedback for all antennas, the system segments the feedback into: 1) subcodebook index indicating which subcodebook is used, 2) first subset of coefficients with first quantization precision, and 3) second subset of coefficients with second quantization precision. This segmentation reduces the total feedback overhead while maintaining throughput performance.

Inventive Principle:
Principle #1Segmentation

2Productivity

If the number of antenna elements is increased, then the spatial multiplexing gain is improved, but the physical dimensions of transmitter and receivers increase, making them more vulnerable to environmental effects

Engineering Contradiction:
Improvespatial multiplexing gainVSAvoidenvironmental vulnerability
Core Design Contradiction:
ProductivityVSObject-affected harmful factors

Solution Approach 1:

The patent changes the parameter representation from individual antenna coefficients to correlation-based parameters. By modeling the channel using a subset of coefficients that capture the essential spatial correlation characteristics, the system achieves the same spatial multiplexing gain with fewer physical antennas, thereby reducing environmental vulnerability while maintaining productivity.

Inventive Principle:
Principle #35Parameter changes

3Loss of information

If quantization is used to reduce CSI feedback overhead, then the feedback overhead is reduced, but the CSI reporting accuracy decreases

Engineering Contradiction:
Improvefeedback overheadVSAvoidCSI reporting accuracy
Core Design Contradiction:
Loss of informationVSMeasurement precision

Solution Approach 1:

Different quantization precisions are applied to different subsets of channel coefficients based on their local importance. The first subset of coefficients (representing dominant spatial correlations) is quantized with higher precision, while the second subset (representing less critical correlations) uses lower precision. This local quality differentiation reduces overall feedback overhead while preserving essential CSI accuracy.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP3726738B1Reporting of channel state information
Publication Date: 2026.03.04 HUAWEI TECH CO LTD
  • EP3726738B1 patent drawingFigure 1~2
  • EP3726738B1 patent drawingFigure 3~4
  • EP3726738B1 patent drawingFigure 5~6

AI summary

Reporting of channel state information (CSI) between a transmit node (TN) and a receive node (RN) in wireless communications is disclosed. For a TN being arranged for multiple transmit antenna transmissions and the CSI being represented by K matrix/matrices, Wi, from a codebook, C, comprising a plurality of matrices; wherein each matrix in said codebook, C, represents a state for the multi-antenna channel from the TN to the RN, K ≥ 1 and i = 0.. K -1; a first sub-index, k1, and at least one second sub-index, k2, are reported for each one of the K matrix/matrices, Wi, wherein each one of the K matrix/matrices, Wi, is indexed by the first and second sub-indices, and the first and second sub-indices have different time-frequency reporting granularity.