MIMO Codebook Design for Feedback Overhead Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Wireless communication systems face challenges in beamforming, particularly in closed-loop systems which require channel feedback, leading to overhead and sensitivity to feedback errors, and in open-loop systems which need constant phase calibration and uplink pilots, making them costly and sensitive to radio channel environments.

Innovation Solution

A systematic codebook design for closed-loop MIMO systems using complex Hadamard transformations and discrete Fourier transform matrices to construct unitary matrices with constant modulus properties, optimizing spectral efficiency and reducing channel quality index calculations, suitable for both single-user and multi-user MIMO scenarios.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If closed-loop transmit beamforming is used, then channel feedback overhead is reduced compared to open-loop, but the system becomes sensitive to feedback channel errors due to feedback delay or fast channel variation

Engineering Contradiction:
Improvefeedback overheadVSAvoidfeedback channel error sensitivity
Core Design Contradiction:
Loss of informationVSReliability

Solution Approach 1:

The system performs preliminary phase calibration in the uplink before downlink transmission. The base station receives uplink pilot signals from the mobile station, estimates the channel phase, and pre-compensates for phase errors. This preliminary action ensures that when downlink feedback is transmitted, the phase alignment is already optimized, reducing sensitivity to feedback delays and channel variations.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements a closed-loop feedback mechanism where the mobile station measures the downlink channel quality and feeds back channel quality indicators (CQI) to the base station. The base station uses this feedback to adaptively adjust the precoding matrix and beamforming weights, optimizing performance while managing feedback overhead through selective feedback of essential channel information.

Inventive Principle:
Principle #23Feedback

2Device complexity

If open-loop transmit beamforming is used, then phase calibration overhead is reduced, but the system requires constant uplink pilots which lead to excessive feedback overhead

Engineering Contradiction:
Improvephase calibration overheadVSAvoidfeedback overhead
Core Design Contradiction:
Device complexityVSLoss of information

Solution Approach 1:

The system merges uplink and downlink channel estimation by exploiting channel reciprocity in TDD systems. The base station uses uplink pilot signals to estimate both the uplink channel and the downlink channel (assuming reciprocal channels). This combined estimation approach eliminates the need for separate downlink pilots and reduces feedback overhead, as the base station already has channel information from uplink measurements.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The mobile station performs self-service by autonomously selecting appropriate precoding matrices from a codebook based on its channel measurements and feedback conditions. The mobile station calculates channel quality indicators and selects the best precoding matrix index (PMI) without requiring complex base station processing, thereby reducing overall system overhead and complexity.

Inventive Principle:
Principle #25Self-service

3Productivity

If closed-loop system with channel feedback is used, then spectral efficiency is improved, but additional overhead is required for feedback transmission

Engineering Contradiction:
Improvespectral efficiencyVSAvoidfeedback overhead
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The system changes feedback parameters adaptively based on channel conditions. When channel conditions are stable, the system reduces feedback frequency and uses compressed feedback formats. When channel conditions change rapidly, the system increases feedback frequency but uses differential feedback (only reporting changes) to minimize overhead. The precoding matrix feedback is quantized to essential bits, transmitting only the most significant channel information needed for beamforming optimization.

Inventive Principle:
Principle #35Parameter changes

4Productivity

If codebook size is increased to improve channel matching accuracy, then spectral efficiency gains increase, but channel quality index calculations become more complex

Engineering Contradiction:
Improvespectral efficiencyVSAvoidchannel quality index calculation complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The codebook is segmented into multiple sub-codebooks organized by transmission rank and antenna configuration. Instead of searching through one large codebook, the mobile station first determines the transmission rank based on channel conditions, then selects from the corresponding sub-codebook. This segmentation reduces the search space and calculation complexity while maintaining accurate channel matching, as each sub-codebook is optimized for specific transmission scenarios.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP2224607B1Method and Apparatus for Multiple Input Multiple Output (MIMO) Transmit Beamforming
Publication Date: 2018.08.08 SAMSUNG ELECTRONICS CO LTD
  • EP2224607B1 patent drawingFigure 1
  • EP2224607B1 patent drawingFigure 2
  • EP2224607B1 patent drawingFigure 3

AI summary

A wireless communications network is provided. The wireless communications network comprises a plurality of base stations. Each one of said base stations is capable of wireless communications with a plurality of subscriber stations. At least one of said plurality of base stations comprises a processor configured to select a codeword from a codebook and precode data with the selected codeword, and a transmitter configured to transmit the precoded data. Rank 1 of the codebook is selected from the following algorithm: