Massive MIMO Channel Measurement Feedback Overhead Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing Massive MIMO technology faces high feedback overheads in channel measurement, particularly due to the need for terminal devices to determine and feedback complex precoding matrices, which increases computational complexity and pilot overheads.
Innovation Solution
A method where terminal devices receive precoded reference signals based on predetermined angle or delay vectors, allowing them to simplify channel measurement by feeding back only delay vectors and weighting coefficients, reducing feedback overheads while ensuring precision through reciprocity-based channel adaptation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If terminal devices determine and feed back complete precoding matrices through downlink channel measurement, then precoding precision is improved, but feedback overheads and computational complexity increase significantly
Solution Approach 1:
The precoding matrix is segmented into multiple beam vectors and corresponding weighting coefficients. Instead of feeding back the complete precoding matrix, the terminal device feeds back only the indices of selected beam vectors and their weighting coefficients, which are then used to reconstruct the precoding matrix at the network device side. This segmentation significantly reduces feedback overhead while maintaining precoding precision.
Solution Approach 2:
The essential components (beam vectors and weighting coefficients) are extracted from the complete precoding matrix for feedback. The network device already has the beam vectors from codebook configuration, so only the indices and weighting coefficients need to be fed back, extracting and transmitting only the necessary information to reconstruct the precoding matrix.
2Measurement precision
If terminal devices determine and feed back complete precoding matrices, then precoding precision is improved, but computational complexity at terminal device increases
Solution Approach 1:
The complex precoding matrix determination process is segmented into beam vector selection and weighting coefficient determination. The terminal device only needs to select beam vectors from the codebook and determine their weighting coefficients, rather than computing the complete precoding matrix, significantly reducing computational complexity while maintaining precision.
Solution Approach 2:
The computationally intensive task of determining the complete precoding matrix is replaced by extracting only the essential elements (beam vector indices and weighting coefficients) that are needed to reconstruct the precoding matrix at the network device, reducing terminal device computational burden.
3Measurement precision
If network device uses more reference signal ports for channel measurement, then channel measurement precision is improved, but pilot overheads increase
Solution Approach 1:
The reference signal ports are segmented and associated with specific beam vectors. Instead of requiring separate reference signal ports for each element of the precoding matrix, the system uses a reduced set of reference signal ports corresponding to the beam vectors, with the precoding information derived from beam vector indices and weighting coefficients.
Solution Approach 2:
The need for numerous reference signal ports is reduced by extracting and using only the essential channel information contained in the beam vectors and weighting coefficients. The network device can reconstruct the precoding matrix using fewer reference signal measurements combined with the codebook-based beam vectors.
Data Source
AI summary
One example method includes receiving one or more precoded reference signals, where the one or more precoded reference signals are obtained by precoding one or more reference signals based on K angle vectors. First indication information can then be generated, where the first indication information is used to indicate at least one delay vector and P weighting coefficients corresponding to P angle-delay pairs, the at least one delay vector and the P weighting coefficients are determined based on the one or more precoded reference signals, each of the P angle-delay pair includes one of the K angle vectors and one of the at least one delay vector, and the P angle-delay pairs and the P weighting coefficients are used to determine a precoding matrix, where P and K are positive integers, and where P≥1, K≥1. The first indication information can then be sent.