Joint CSI Report Quantization for Multi-TRP Overhead Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing communication systems face significant challenges in managing the high overhead of CSI reports for multiple TRPs, leading to potential report drops due to limited uplink reporting resources, especially in 5G networks utilizing mmWave frequency spectrum.
Innovation Solution
A method and apparatus for reducing CSI reporting overhead by designating one CSI report as a reference and others as dependent, using differential quantization and encoding, thereby reducing the amount of data transmitted.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If multiple CSI reports are transmitted for multiple TRPs, then the channel state information coverage is improved, but the uplink reporting overhead increases
Solution Approach 1:
The patent segments CSI reports into two categories: reference CSI reports (independently decoded) and dependent CSI reports (differentially decoded). This segmentation allows the system to maintain complete CSI information for multiple TRPs while reducing overhead by exploiting correlations between reports. The reference reports contain full CSI information, while dependent reports contain only differential updates, thus resolving the contradiction between information completeness and overhead reduction.
Solution Approach 2:
The patent uses copying by having dependent CSI reports reference and reuse information from reference CSI reports. Instead of transmitting redundant full CSI data for each TRP, the system creates differential copies that point to the reference report for common information. This copying mechanism significantly reduces uplink overhead while maintaining the ability to reconstruct complete CSI information at the receiver, addressing the contradiction between reporting completeness and overhead quantity.
2Reliability
If all CSI reports are independently decoded, then the decoding reliability is improved, but the processing complexity increases
Solution Approach 1:
The patent segments the decoding process into two distinct paths: independent decoding for reference CSI reports and differential decoding for dependent CSI reports. This segmentation allows the system to maintain high reliability for reference reports through independent decoding while reducing overall processing complexity by using simpler differential decoding for dependent reports. The segmented approach resolves the contradiction by applying different decoding strategies to different report types based on their importance and correlation characteristics.
Solution Approach 2:
The patent applies local quality by using different decoding methods for different CSI reports based on their specific characteristics. Reference CSI reports receive the more reliable but complex independent decoding treatment, while dependent CSI reports use the simpler differential decoding approach. This localized application of decoding quality resolves the contradiction between overall reliability and processing complexity by optimizing each report type according to its specific needs and correlation with reference reports.
3Quantity of substance
If differential values are used in dependent CSI reports, then the reporting overhead is reduced, but the quantization precision may be compromised
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the quantization parameters for differential values in dependent CSI reports. By optimizing the number of bits allocated for differential quantization based on the actual correlation between reference and dependent reports, the system achieves the right balance between overhead reduction and precision maintenance. This parameter adjustment resolves the contradiction by adapting the quantization precision to the specific statistical properties of the CSI data, ensuring sufficient accuracy while minimizing overhead.
Data Source
AI summary
Techniques of generating CSI reports include designating at least one of the reports as a reference and the other reports as dependent on the reference. For example, a dependent CSI report may have a differential value of a CSI quantity rather than a raw value of the CSI quantity. Because generally for parameters in a CSI report, the differential quantity has a smaller range of values than the raw values of the CSI quantity, a quantization of the differential value of the CSI quantity will have fewer bits than the raw value; for example, for L1-RSRP, the raw value has 7 bits while the differential value has 4 bits. In another example, a dependent value of a CSI quantity is a flag indicating that the CSI quantity value is located in a reference CSI report.