PMI Codebook Basis Selection for Fast-Changing MIMO Channels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Closed loop MIMO in wireless communication networks experiences increased signaling overhead and channel aging issues, particularly in high-speed environments, necessitating improved precoding matrix feedback to handle fast-changing channels.
Innovation Solution
The method involves selecting time-domain (TD) or Doppler-domain (DD) basis vectors independently for different groupings of spatial-domain (SD) and frequency-domain (FD) basis vectors for different groupings of spatial-domain (SD) and frequency-domain (FD) basis vectors to enhance precoding matrix information feedback, reducing signaling overhead by grouping and selecting inactive spatial beams or frequency components, and using bitmap matrices to indicate non-zero coefficients.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If frequent feedback reporting is triggered to handle channel aging in high-speed environments, then the validity of precoding matrix information is improved, but signaling overhead increases
Solution Approach 1:
The codebook is segmented into multiple codebook groups, each associated with different time-domain basis vectors. Instead of reporting the entire precoding matrix frequently, the UE reports only the selected codebook group index and basis vector indices, which are then used to reconstruct the precoding matrix at the gNB. This segmentation reduces the amount of feedback data while maintaining accuracy in fast-changing channels.
Solution Approach 2:
The gNB pre-configures multiple codebook groups with different time-domain basis vectors before feedback reporting. The UE selects from these pre-prepared options based on current channel conditions, eliminating the need to report complete precoding matrices frequently. The preliminary preparation of codebook groups enables efficient adaptation to channel changes with minimal signaling.
2Reliability
If complete precoding matrix information is reported to assist transmitter precoding configuration, then precoding performance is improved, but feedback overhead increases
Solution Approach 1:
The invention extracts only the essential parameters needed for precoding reconstruction: codebook group index, time-domain basis vector indices, and linear combination coefficients. The complete precoding matrix is not reported; instead, only the key components that define the precoding structure are fed back. This extraction approach maintains precoding accuracy while dramatically reducing feedback overhead.
Solution Approach 2:
The feedback mechanism changes from reporting complete matrix elements to reporting compressed parameters (indices and coefficients). The precoding information is transformed from a full matrix representation to a parameterized representation using codebook groups and basis vectors, reducing the dimensionality of feedback while preserving the essential precoding characteristics needed for accurate reconstruction.
3Measurement precision
If minimum time delay for feedback reporting is maintained to ensure processing accuracy, then measurement reliability is improved, but responsiveness to channel changes deteriorates
Solution Approach 1:
The codebook groups are dynamically configured with different time-domain basis vectors to match current channel conditions. The gNB can adapt the codebook configuration based on estimated mobility and channel dynamics, allowing the feedback mechanism to respond quickly to channel changes while maintaining measurement accuracy through appropriate basis vector selection.
Solution Approach 2:
Multiple codebook groups are pre-configured with different time-domain characteristics before channel changes occur. When channel conditions change, the UE can immediately select from these pre-prepared groups without waiting for reconfiguration, enabling rapid adaptation to channel variations while maintaining measurement precision through the appropriate pre-configured basis vectors.
Data Source
AI summary
Methods and systems are disclosed to enhance reporting of precoding matrix indication (PMI) associated with fast changing channel response by independently selecting time-domain (TD) or Doppler-domain (DD) basis vectors for different groupings of spatial-domain (SD) and frequency-domain (FD) basis vectors. A UE may generate PMI by applying TD/DD basis vectors in addition to the SD and FD basis vectors to increase the validity time of the PMI. To address the increased signaling overhead associated with the TD/DD components, the UE may group selected spatial beams and FD components to drop inactive spatial beams or inactive FD components based on the LC coefficient matrix. The UE may independently select TD or DD basis vectors for different paired groups of spatial beams and FD components and report the sub-selection of the spatial beams, FD components. and TD or DD component for each group, and the reduced dimension of the LC coefficient matrix.