Type II CSI Feedback Basis Selection for MU-MIMO Overhead Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Type II precoder schemes in NR wireless communication systems lead to increased CSI feedback overhead and UE precoder search complexity, limiting MU-MIMO performance and high layer representations.

Innovation Solution

A method for expressing precoder vectors as a linear combination of spatial-domain and frequency-domain components, allowing for reduced feedback overhead by selecting frequency-domain components from a subset of candidate components and transmitting non-zero coefficients.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If Type II precoder schemes are used to improve MU-MIMO performance, then channel state information accuracy is improved, but CSI feedback overhead increases

Engineering Contradiction:
Improvechannel state information accuracyVSAvoidCSI feedback overhead
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent extracts only the essential frequency-domain components from the complete precoder representation. By identifying and transmitting only non-zero coefficients of the frequency-domain basis vectors, the system removes redundant information while preserving the critical channel state information needed for accurate precoding.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies different levels of detail to different parts of the precoder representation. Frequency components with non-zero coefficients are represented with full precision, while zero-coefficient components are omitted entirely. This localized quality approach ensures that only the most significant channel characteristics are transmitted with high fidelity.

Inventive Principle:
Principle #3Local quality

2Measurement precision

If Type II precoder schemes are used to improve MU-MIMO performance, then channel state information accuracy is improved, but UE precoder search complexity increases

Engineering Contradiction:
Improvechannel state information accuracyVSAvoidUE precoder search complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the precoder search process into two stages: first selecting from a limited set of frequency-domain basis vectors, then determining coefficients for the selected basis. This segmentation reduces the overall search space by breaking down the complex precoder optimization into manageable sub-problems with smaller search spaces.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces dynamic adaptation in the precoder representation by allowing the UE to select frequency-domain basis vectors that best match the current channel conditions. This dynamic selection process enables the system to adapt to varying channel characteristics without exhaustively searching all possible precoders.

Inventive Principle:
Principle #15Dynamics

3Productivity

If frequency-selective precoding is implemented to match channel variations, then data rate is improved, but feedback overhead increases

Engineering Contradiction:
Improvedata rateVSAvoidfeedback overhead
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The patent creates a compressed representation of the frequency-selective precoder by copying only the essential characteristics (non-zero coefficients) rather than transmitting the complete precoder for each sub-band. This copying approach maintains the ability to track channel variations while significantly reducing the amount of feedback data required.

Inventive Principle:
Principle #26Copying

Data Source

PatentEP3874605B1Hybrid common/independent fd-basis for type ii CSI enhancement
Publication Date: 2026.04.01 TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)
  • EP3874605B1 patent drawingFigure 1
  • EP3874605B1 patent drawingFigure 2
  • EP3874605B1 patent drawingFigure 3

AI summary

A method performed by a wireless device (510, 700, 1000) for transmitting a channel state information (CSI) report for a downlink channel comprises obtaining (601) a first set of candidate frequency-domain components and determining (602) a set of spatial-domain components. The method comprises determining (603) a second set of candidate frequency-domain components as a subset of the first set of candidate frequency-domain components. The method comprises determining (604), for each spatial-domain component of the set of spatial-domain components, a spatial-domain component-specific set of frequency-domain components as a subset of the second set of candidate frequency-domain components. The method comprises transmitting (605), to a network node (560, 900), the CSI report.