MIMO Precoding Quantization Error Reduction via Multi-Codebook Feedback

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current MIMO systems face challenges in reducing quantization errors due to limited codebook sizes, which degrade link performance and increase complexity, while larger codebooks increase feedback overhead and computational requirements.

Innovation Solution

The method involves the receiver feeding back multiple indices of precoding matrices along with scalar coefficients related to the geometric structure, allowing the transmitter to generate a refined precoding matrix that reduces quantization errors, even with a small codebook, by interpolating or combining matrices from multiple codebooks.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the codebook size is increased to reduce quantization error, then link performance is improved, but feedback overhead and receiver complexity increase exponentially

Engineering Contradiction:
Improvequantization precisionVSAvoidfeedback overhead
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the precoding matrix selection process into two stages: first selecting a coarse precoding matrix from a first codebook, then refining it by selecting a second precoding matrix from a second codebook. This segmentation allows the system to achieve high quantization precision without requiring a single enormous codebook, thereby reducing feedback overhead while maintaining performance.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a hierarchical dimension to the codebook structure, organizing precoding matrices across multiple levels (first codebook for coarse selection, second codebook for refinement). This dimensional transformation enables the system to represent precoding matrices with higher precision using a more structured, multi-level approach rather than a single flat codebook, reducing the feedback burden.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Measurement precision

If the codebook size is increased to reduce quantization error, then link performance is improved, but computational complexity and memory requirements increase

Engineering Contradiction:
Improvequantization precisionVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent divides the computationally intensive codebook search into two manageable stages: a first search in the first codebook to identify candidate matrices, and a second search in the second codebook to refine the selection. This segmentation significantly reduces the computational burden compared to searching a single large codebook, while still achieving high quantization precision.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary action by first selecting a coarse precoding matrix from the first codebook before proceeding to refine it with the second codebook. This preliminary selection narrows down the search space and guides the subsequent refinement process, reducing overall computational complexity while maintaining high precision.

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If multiple codebooks are used to reduce quantization error, then link performance is improved, but system complexity and design burden increase

Engineering Contradiction:
Improvequantization precisionVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the precoding matrix representation into two distinct codebooks with specific functions: the first codebook provides coarse quantization with fewer matrices, while the second codebook provides fine quantization with more matrices. This segmentation allows each codebook to be optimized independently and simplifies the overall system design compared to creating a single large codebook.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies partial action by using a first codebook that provides sufficient coarse quantization on its own, then adds a second codebook for additional refinement. This approach uses only the necessary amount of codebook capacity at each level, avoiding the excessive complexity of a single comprehensive codebook while achieving the desired precision.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP2400705B1Multiple-input multiple-output systems and methods for wireless communication thereof for reducing the quantization effect of precoding operations utilizing finite codebooks
Publication Date: 2014.03.05 HTC CORP
  • EP2400705B1 patent drawingFigure 1
  • EP2400705B1 patent drawingFigure 2
  • EP2400705B1 patent drawingFigure 3

AI summary

A method for reducing the quantization effect of precoding operations utilizing a finite codebook in MIMO systems is provided. First, at the receiver side, downlink channel state information is obtained and a set of indices of precoding matrices within a plurality of finite codebooks are selected accordingly. The selected indices of precoding matrices for each of the finite codebooks and a set of scalar coefficients are transmitted from the receiver to the transmitter. Thereafter, at the transmitter side, at least a first and a second refined precoding matrices are generated based on the selected set of indices of precoding matrices for all of the finite codebooks, and the one or more scalar coefficients and a final precoding matrix is generated at least based on the first refined precoding matrix and the second refined precoding matrix. The final precoding matrix is applied for transmission between the transmitter and the receiver.