Precoding Feedback Reduction via Codebook Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional codebook-based precoding techniques are poorly suited for multi-point transmission scenarios in wireless communications networks, leading to a large signaling load due to varying path losses between mobile stations and multiple transmitter sites, which consumes system resources and increases feedback bits required.
Innovation Solution
The proposed solution involves specifying a precoding vector as a two-step process, where a codebook is selected based on path loss conditions using log2(Q) bits and an index to the desired precoding vector within the codebook is specified using log2(L) bits, with updates to the codebook occurring at a slower rate than the precoding vector, reducing the overall feedback bits needed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional codebook-based precoding techniques are used for multi-point transmission, then the system can support multiple transmitter sites, but the signaling load increases due to varying path losses between mobile stations and multiple transmitter sites
Solution Approach 1:
The precoding vector specification is segmented into two independent parts: a codebook index indicating the codebook to use, and a vector index indicating the specific precoding vector within that codebook. This segmentation allows the system to support multi-point transmission by selecting appropriate codebooks for different path loss scenarios while maintaining efficient signaling through compact indexing.
2Adaptability or versatility
If a single precoding vector is specified from all possible precoding vectors, then the system can adapt to channel conditions, but the number of feedback bits increases due to the large number of possible codebook combinations
Solution Approach 1:
The feedback mechanism is segmented into two independent indices: a codebook index that identifies which codebook to use, and a vector index that identifies the specific precoding vector within that codebook. This reduces the total feedback bits compared to specifying a single precoding vector from all possible vectors across multiple codebooks, because the vector index only needs to cover vectors within one codebook rather than all vectors across all codebooks.
Solution Approach 2:
The system performs preliminary action by pre-defining multiple codebooks with different structures optimized for different path loss scenarios. The mobile station measures path losses and selects the appropriate codebook before selecting the specific precoding vector, which streamlines the feedback process and reduces the number of bits needed to convey the precoding information.
3Productivity
If codebook updates occur at the same rate as precoding vector updates, then the system can respond to changing conditions, but the signaling overhead increases
Solution Approach 1:
The system performs preliminary action by establishing multiple pre-defined codebooks that cover different path loss scenarios. The mobile station measures path losses and selects the appropriate codebook, which then remains valid for a period. During this period, only the vector index within the selected codebook needs to be updated frequently to track fast fading, while the codebook itself changes only when path loss conditions change significantly. This separation allows differential update rates that reduce overall signaling overhead.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Techniques for reducing the number of bits needed to specify the best precoding vector for each mobile station in a wireless communication network that employs multi-point transmission are disclosed. An exemplary method begins with the estimation of path loss between a mobile station and each of a plurality of geographically separated transmitter sites, each transmitter site having at least one transmitter antenna. Based on the estimated path losses, one of a plurality of pre-determined subsets (codebooks) of a pre-determined set of antenna precoding vectors is selected. A group index identifying the selected subset is then transmitted to the mobile station. Subsequently, a vector index is received from the mobile station, the vector index corresponding to a precoding vector in the selected subset, and data is transmitted to the mobile station, using the precoding vector applied to the transmitter antennas at the plurality of transmitter sites.