CSI Feedback Overhead Reduction in Massive MIMO Codebook Design

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current channel state information (CSI) feedback mechanisms in Massive MIMO systems, particularly in NR, face challenges in reducing signaling overheads due to the need to report a large number of beam superposition coefficients, which is not efficiently managed by existing codebook designs.

Innovation Solution

The proposed method reduces feedback overheads by generating and using codebook indication information that includes fewer pieces of beam information for each transport layer, allowing the receive-end device to construct precoding vectors using a reduced set of beam information, thereby minimizing the number of beam superposition coefficients fed back.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a plurality of beam superposition coefficients are reported for each transport layer in NR Massive MIMO, then the accuracy of channel state information is improved, but the feedback overhead increases significantly

Engineering Contradiction:
Improvechannel state information accuracyVSAvoidfeedback overhead
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent extracts and reports only the necessary beam superposition coefficients rather than all possible coefficients. Specifically, it identifies and reports only those coefficients that are essential for reconstructing the precoding matrix, thereby reducing feedback overhead while maintaining channel state information accuracy.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent segments the feedback process into two parts: W1 matrix indicating beam directions and W2 matrix indicating beam superposition coefficients. This segmentation allows the system to separately optimize the feedback of beam selection information and beam combination information, reducing overall feedback overhead.

Inventive Principle:
Principle #1Segmentation

2Loss of information

If all beams are reported for each transport layer, then the completeness of CSI feedback is improved, but the signaling overhead increases

Engineering Contradiction:
ImproveCSI feedback completenessVSAvoidsignaling overhead
Core Design Contradiction:
Loss of informationVSQuantity of substance

Solution Approach 1:

The patent applies partial action by reporting only a subset of beam superposition coefficients rather than all possible coefficients. It identifies the minimum necessary set of coefficients required to reconstruct the precoding matrix accurately, avoiding the overhead of reporting redundant information.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent changes the parameter representation by using a compact indexing scheme for beam superposition coefficients. Instead of reporting all coefficients explicitly, it uses index values that reference pre-defined codebook entries, thereby reducing the number of bits required for feedback.

Inventive Principle:
Principle #35Parameter changes

3Quantity of substance

If the number of reported beams is reduced to 1 beam in one polarization, then the feedback overhead is reduced, but the channel state information accuracy deteriorates

Engineering Contradiction:
Improvefeedback overheadVSAvoidchannel state information accuracy
Core Design Contradiction:
Quantity of substanceVSMeasurement precision

Solution Approach 1:

The patent merges the feedback of multiple beams across different polarizations into a unified codebook structure. By combining beam information from both polarizations in the W1 matrix and their superposition coefficients in the W2 matrix, it achieves accurate CSI feedback with reduced overhead compared to reporting all beams separately.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP3633891B1Method for feeding back channel state information, method for receiving channel state information, sending device, and receiving device
Publication Date: 2022.03.02 HUAWEI TECH CO LTD
  • EP3633891B1 patent drawingFigure 1
  • EP3633891B1 patent drawingFigure 2
  • EP3633891B1 patent drawingFigure 3~4

AI summary

Embodiments of this application disclose channel state information feedback and receiving methods, a transmit-end device and a receive-end device, and relate to the field of communications technologies. The method includes: generating and sending, by a transmit-end device, codebook indication information of K transport layers, where K is an integer greater than or equal to 2; the codebook indication information includes: L pieces of beam information used by the K transport layers, and indication information used to indicate beam information associated with each of the K transport layers; L is an integer greater than or equal to 2, and a quantity of pieces of beam information associated with at least one of the K transport layers is less than L. Implementing a channel state information feedback and receiving technology provided in this application helps reduce feedback overheads of channel state information.