Precoding Matrix Feedback via Weighted Vector Combination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The current dual-stage codebook structure in LTE FDD systems for precoding matrix feedback is not sufficiently precise, limiting the accuracy of channel information feedback and downlink communication quality.
Innovation Solution
A method that involves sending a reference signal on multiple antenna ports, allowing the first device to measure and feed back channel information including identification of vectors and weighted combination factors, enabling the second device to generate a more precise precoding matrix through weighted combination of vectors, thereby improving link adaptation and system performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a dual-stage codebook structure is used for precoding matrix feedback, then the feedback mechanism is established, but the feedback precision is insufficient
Solution Approach 1:
The precoding matrix feedback is segmented into two stages: W1 selection from a first codebook and W2 selection from a second codebook. This segmentation allows independent optimization of each stage, improving overall feedback precision while managing complexity through modular structure
Solution Approach 2:
The patent introduces an additional dimension to the codebook structure by adding the W2 stage with column selection and co-phase information. This transforms the single-stage feedback into a multi-dimensional feedback mechanism, enhancing precision by providing finer granular control over precoding matrix selection
2Measurement precision
If only one vector is selected from M vectors using W2, then the feedback mechanism is simple, but the precoding matrix feedback precision is not sufficient
Solution Approach 1:
Instead of selecting only one vector from M vectors, the patent enables selection of multiple vectors through the W2 stage. This partial extension beyond single-vector selection improves precoding precision by capturing more channel characteristics, while the information overhead is managed through efficient encoding of column selection and co-phase information
3Measurement precision
If multiple vectors are combined with weighted combination factors, then the precoding precision is improved, but the information overhead increases
Solution Approach 1:
The patent changes the parameter representation by introducing weighted combination factors (amplitude and phase) that continuously adjust the contribution of each selected vector. This parameter-based approach achieves high precision by fine-tuning vector combinations, while the information overhead is managed through quantization and efficient coding schemes for the combination factors
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
A channel information sending method, a data sending method, and a device are provided, to improve feedback precision of a precoding matrix. A first device includes: a receiving module, configured to receive a reference signal; a processing module, configured to measure the reference signal to obtain first channel information and second channel information; and a sending module, configured to send the first channel information and the second channel information, where the first channel information includes identification information of M first vectors, M is an integer not less than 2, the second channel information includes information about a weighted combination factor used for performing weighted combination on N first vectors in the M first vectors, N is a positive integer not greater than M, and the first channel information and the second channel information are used to constitute a precoding matrix. When generating the precoding matrix, the second device may perform weighted combination on the M first vectors based on the weighted combination factor indicated by the received second channel information, instead of selecting only one eigenvector from a plurality of eigenvectors, so that the generated precoding matrix is more precise.