WLAN CSI QR Decomposition That Preserves Relative Phase
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing WLAN sensing technologies fail to reserve relative phase information between subcarriers and pulses during QR decomposition, leading to ineffective target attribute information extraction such as target delay, angle, and Doppler shift.
Innovation Solution
Perform conjugate transpose and QR decomposition on channel state information to obtain a matrix Q that reserves relative phase information, allowing for effective sensing by indicating attribute information like delay, AOD, and Doppler shift.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of substance
If QR decomposition is performed on channel state information using existing rules, then feedback overheads are reduced, but relative phase information between subcarriers and pulses is lost
Solution Approach 1:
The patent changes the parameter representation by introducing a new decomposition format where the channel state information is decomposed into a unitary matrix Q and an upper triangular matrix R, but with modified phase reference rules. Specifically, the phase of certain elements in R is set to zero (reference phases) rather than using conventional QR decomposition rules, thereby preserving relative phase information while maintaining compression.
Solution Approach 2:
The patent introduces an intermediary reference phase mechanism. By selecting specific elements in the R matrix as phase references (setting their phases to zero), the system creates a reference framework that allows relative phase information to be preserved in the Q matrix without requiring full channel state information feedback. This intermediary reference system enables the preservation of sensing-critical phase relationships.
2Measurement precision
If relative phase information is preserved in QR decomposition, then target attribute information extraction becomes possible, but feedback overheads increase
Solution Approach 1:
The patent extracts only the essential information needed for sensing by decomposing the channel state information into Q and R matrices, where Q contains the relative phase information and R contains magnitude information with reference phases. By extracting and separately handling the phase reference elements, the system preserves sensing-critical information while minimizing feedback overhead through selective information retention.
Solution Approach 2:
The patent segments the channel state information into distinct components: the unitary matrix Q containing phase information, the upper triangular matrix R containing magnitude information, and identified reference phase elements. This segmentation allows the system to selectively feedback only the necessary components for sensing (primarily Q with reference phase indicators) rather than the complete channel state information, reducing overhead while maintaining measurement precision.
3Productivity
If conventional QR decomposition is used, then channel state information is compressed, but sensing capability is lost
Solution Approach 1:
The patent performs preliminary action by pre-identifying and marking reference phase elements in the R matrix before feedback. By determining which elements will serve as phase references in advance and setting their phases to zero, the system prepares the decomposition structure to preserve relative phase information. This preliminary configuration ensures that sensing capability is maintained from the outset while compression is applied through the structured Q-R decomposition format.
Data Source
AI summary
A communication method includes a first device that determines channel state information based on a first frame that is used for channel measurement and that is sent by a second device, performs conjugate transpose and QR decomposition to obtain a matrix R and a matrix Q, and sends, to the second device, the matrix Q indicating attribute information of a target path.