Beamformee Steering Matrix Generation via SVD Submatrix Extraction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing beamformees require additional control circuits for convex optimization, increasing design complexity and manufacturing cost due to row normalization distorting the steering matrix in beamforming processes.

Innovation Solution

A beamformee that executes a first singular value decomposition operation on a channel coefficient matrix to generate a steering matrix, using submatrices and iterative updates to avoid distortion, eliminating the need for additional control circuits.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If convex optimization technique is used to generate steering matrix, then beamforming performance is maintained, but device complexity and manufacturing cost increase significantly

Engineering Contradiction:

Improvebeamforming performanceVSAvoidcontrol circuits

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent extracts the essential computational core (singular value decomposition) from the complex convex optimization framework, retaining the beamforming performance while eliminating the need for additional control circuits. The SVD-based method focuses only on the critical matrix decomposition operation needed to generate the steering matrix.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent replaces expensive, complex control circuitry with a simpler, computationally-efficient algorithmic approach. The SVD-based steering matrix generation uses standard processing circuits already present in the beamformee, avoiding the need for specialized hardware components.

Inventive Principle:

Principle #27Cheap short-living objects (Disposable)

2Ease of operation

If row normalization operation is executed on steering matrix, then output power equality is achieved, but steering matrix distortion occurs

Engineering Contradiction:

Improveoutput power controlVSAvoidbeamforming performance

Core Design Contradiction:

Ease of operationVSReliability

Solution Approach 1:

The patent performs preliminary normalization during the SVD decomposition process itself, rather than applying row normalization to the final steering matrix. By incorporating normalization into the matrix factorization step, the algorithm ensures that the steering matrix columns have unit norm without distorting the matrix structure or degrading beamforming performance.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11405078B1Device for implementing beamforming in wireless networks

Publication Date: 2022.08.02 NXP USA INC

AI summary

A beamformee executes a singular value decomposition operation on a channel coefficient matrix to generate a diagonal matrix and a unitary matrix. The beamformee extracts a submatrix from each of the diagonal matrix and the unitary matrix. Each of a number of rows and a number of columns of the submatrix extracted from the diagonal matrix is equal to a number of spatial streams associated with a beamformer. Further, the submatrix extracted from the unitary matrix has a number of rows and a number of columns equal to a number of beamformer antennas of the beamformer and the number of spatial streams, respectively. The beamformee generates a subspace matrix that is a product of the submatrix extracted from the diagonal matrix and a conjugate transpose of the submatrix extracted from the unitary matrix. Further, the beamformee generates a steering matrix based on the subspace matrix.