Precoding Vector Feedback With Space-Frequency Units for Lower Overhead
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing feedback methods for precoding vectors in Massive MIMO systems result in high overheads due to the increasing number of transport layers and frequency domain units, leading to inefficient communication and increased resource utilization.
Innovation Solution
A method is proposed where a terminal device generates a CSI report indicating M space-frequency units and their weighting coefficients to determine a precoding vector, using a two-part approach to reduce feedback overheads by implicitly indicating phase information and optimizing the allocation of quantization bits based on the amplitude of weighting coefficients.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a two-level feedback manner including wideband feedback and sub-band feedback is used to indicate precoding vectors, then the network device can obtain accurate precoding vectors for each transport layer, but feedback overheads multiply as the quantity of transport layers and sub-bands increases
Solution Approach 1:
The patent segments the precoding feedback into two parts: beam vectors (spatial domain) and frequency domain vectors. The beam vectors are shared across multiple frequency domain units, while frequency domain vectors capture per-unit variations. This segmentation allows accurate representation of precoding vectors across multiple transport layers and sub-bands without multiplying feedback overhead, as the beam vector component is reused rather than independently reported for each layer or sub-band.
Solution Approach 2:
The beam vectors serve a universal function across multiple frequency domain units and transport layers. By indicating beam vectors at a wider granularity (applicable to multiple frequency domain units), the same spatial basis vectors are reused across different frequency resources, reducing the need for redundant feedback. This multi-functional use of beam vectors resolves the contradiction between maintaining precision across layers/sub-bands and reducing overall feedback quantity.
2Measurement precision
If a larger quantity of sub-bands is used to describe channel changes in frequency domain, then the approximation precision of precoding vectors improves, but feedback overheads increase significantly
Solution Approach 1:
The patent segments frequency domain representation into frequency domain vectors that capture channel characteristics across multiple frequency domain units. Instead of independently reporting full precoding vectors for each sub-band, the system uses a compact frequency domain vector representation that can be combined with beam vectors to reconstruct precoding vectors for multiple sub-bands, reducing feedback overhead while maintaining approximation precision.
Solution Approach 2:
The patent introduces a frequency domain vector dimension that complements the spatial beam vector dimension. By representing channel characteristics in this additional frequency domain dimension, the system can describe channel changes across multiple sub-bands using a compact vector representation rather than explicit per-sub-band precoding vectors, thereby improving approximation precision without proportionally increasing feedback overhead.
Data Source
AI summary
The present disclosure relates to methods and communications apparatuses. One example method include generating and sending a channel state information (CSI) report. The CSI report indicates M space-frequency units and a weighting coefficient of a part or all of the M space-frequency units, each of the M space-frequency units corresponds to one space domain vector and one frequency domain vector, and a weighted sum of the part or all of the M space-frequency units is usable for determining a precoding matrix of a plurality of frequency domain units.