CSI Feedback Codebook Using Joint Spatial-Frequency Sparsity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing dual-domain compressed codebook in MIMO technology for channel state information feedback in LTE and 5G systems has high overheads due to the sparse feature of the channel in the angle-delay domain, limiting system performance, and the use of discrete Fourier transform codebooks restricts sparseness and increases feedback overheads.
Innovation Solution
A channel state information feedback method using a first codebook based on spatial-frequency domain column vectors to represent channels in a joint spatial-frequency domain, reducing feedback overheads by exploiting the sparse features of the channel and optimizing the codebook design through various formulations that incorporate phase differences, superposition coefficients, and port selection matrices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If a dual-domain compressed codebook is used for CSI feedback, then feedback overhead is reduced by compressing spatial and frequency domains separately, but the overhead remains large because sparse features of the channel are not fully exploited
Solution Approach 1:
The patent merges the spatial domain and frequency domain into a joint spatial-frequency domain for codebook design. Instead of separately compressing spatial and frequency domains as in conventional dual-domain codebooks, the invention creates a unified codebook structure that jointly represents both domains, enabling more efficient exploitation of channel sparsity and reducing feedback overhead while maintaining system performance.
2Ease of manufacture
If DFT codebooks are used for spatial and frequency domains as specified in protocol, then implementation is standardized, but sparseness of the combination coefficient matrix is limited and system performance is reduced
Solution Approach 1:
The patent changes the codebook parameters from conventional DFT-based codebooks to a joint spatial-frequency domain codebook with different structural parameters. The new codebook uses a unified basis that jointly captures spatial and frequency characteristics, changing the fundamental parameters of the codebook design to enable better sparsity exploitation and improved system performance while maintaining implementation feasibility.
3Device complexity
If separate spatial domain vectors and frequency domain vectors are fed back, then feedback structure is simple, but feedback overhead is large because combination coefficients for each subband must be separately fed back
Solution Approach 1:
The patent merges the separate feedback of spatial domain vectors, frequency domain vectors, and combination coefficients into a unified joint spatial-frequency domain codebook representation. This consolidation eliminates the need for separate feedback of combination coefficients for each subband, reducing feedback overhead while maintaining a manageable feedback structure through the unified codebook approach.
Data Source
AI summary
This application discloses a channel state information feedback method and a communication apparatus. The method includes: determining channel state information based on a first codebook, and sending the channel state information. The first codebook is determined based on a channel represented by one or more spatial-frequency domain column vectors, and the one or more spatial-frequency domain column vectors indicate the channel in joint spatial-frequency domain. In this solution, the first codebook may be determined based on the channel represented by the one or more spatial-frequency domain column vectors. Because the one or more spatial-frequency domain column vectors may represent the channel in joint spatial-frequency domain, that is, the first codebook represents the channel based on joint spatial-frequency domain, a joint spatial-frequency domain feature may be fully used, to explore a sparse feature of the channel, thereby improving system performance.