CSI Compression via Spatial and Frequency Transformations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current communication systems face challenges in efficiently compressing and decompressing channel state information (CSI) in wireless communications, particularly in MIMO and Massive MIMO systems, due to the trade-off between accuracy and overhead, especially in Frequency Division Duplexing (FDD) scenarios where channel reciprocity is absent, leading to increased overhead and reduced channel capacity.
Innovation Solution
A computation device and method that perform spatial and frequency-to-time transformations on CSI to generate compressed CSI, exploiting correlations in both dimensions, and a corresponding restoring device that reverses these transformations to recover the original CSI, allowing for reduced digital representation size without significant loss in accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the number of choices in the codebook is increased to provide finer resolution, then precoder selection accuracy is improved, but CSI report overhead increases
Solution Approach 1:
The channel state information is segmented into two parts: short-term CSI (precoding matrix indicator PMI) and long-term CSI (channel quality indicator CQI and rank indicator RI). The long-term CSI components are compressed and fed back less frequently, while short-term PMI provides fine-grained precoder selection. This segmentation allows high precision in precoder selection without proportionally increasing overall feedback overhead.
Solution Approach 2:
Long-term channel characteristics (CQI and RI) are estimated and compressed in advance, before the actual precoder selection is needed. These pre-compressed long-term parameters are stored and reused across multiple short-term feedback instances, eliminating the need to repeatedly transmit redundant channel information and reducing overall feedback overhead while maintaining accurate precoder selection.
2Measurement precision
If CSI is transmitted frequently to maintain accuracy, then channel state accuracy is improved, but channel capacity is reduced
Solution Approach 1:
The system performs preliminary estimation of long-term channel characteristics (CQI and RI) that remain relatively stable over time. These pre-estimated parameters are compressed and fed back less frequently, allowing the system to maintain accurate channel state information without requiring frequent full CSI transmissions, thus preserving channel capacity for data transmission.
Solution Approach 2:
CSI feedback is segmented into short-term (PMI, updated frequently) and long-term (CQI and RI, updated less frequently) components. This segmentation allows the system to maintain channel state accuracy through frequent updates of critical precoding information while reducing overall feedback frequency for stable channel parameters, thereby minimizing the impact on channel capacity.
3Quantity of substance
If the size of digital representation of CSI is reduced, then overhead is reduced, but accuracy of CSI is compromised
Solution Approach 1:
Long-term CSI parameters (CQI and RI) are estimated and compressed in advance before transmission. This preliminary compression allows the system to represent these stable channel characteristics using fewer bits without significant accuracy loss, as the parameters change slowly and can be accurately captured at lower resolution compared to fast-varying short-term channel information.
Solution Approach 2:
Different compression levels are applied to different CSI components based on their importance and variability. Short-term PMI, which requires high precision for accurate precoder selection, is maintained at higher resolution. Long-term CQI and RI, which are more stable and less sensitive to quantization, are compressed to lower resolution, achieving overall size reduction while preserving critical accuracy requirements.
Data Source
AI summary
The invention relates to generating compressed channel state information and restoring the channel state information from the compressed channel state information. A computation device for compressing channel state information, CSI, representing a channel transfer function H having a spatial dimension and a frequency dimension comprises a transforming unit configured to perform a spatial transformation and a frequency-to-time transformation subsequently and in any order on the channel transfer function H to obtain a transformed channel transfer function HT, and a compressing unit configured to select values of the transformed channel transfer function HT and to generate compressed channel state information, CCSI, based on the selected values.