VRAN RU Spatial Compression via Covariance Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The legacy '7-2' split in virtual radio access networks (VRAN) faces challenges such as outdated channel estimates due to periodic SRS transmission, neglect of interference environments, and significant fronthaul bandwidth overhead due to SRS loading.
Innovation Solution
A spatial compression scheme is implemented in the Radio Unit (RU) using antenna covariance matrix estimation and decomposition, allowing for two-stage compression of SRS and PUSCH signals, which reduces the number of streams transmitted over the fronthaul interface.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If SRS is transmitted periodically for channel estimation, then fronthaul bandwidth is reduced through spatial compression, but channel estimates become outdated in high mobility scenarios
Solution Approach 1:
The system performs spatial compression of SRS signals before transmission over the fronthaul interface, preparing the signals in advance to reduce bandwidth requirements. The antenna covariance matrix is estimated and decomposed to create compressed representations that maintain channel estimation accuracy while reducing the number of streams transmitted.
Solution Approach 2:
The system changes the parameter of spatial compression by using antenna covariance matrix estimation and decomposition to reduce the number of SRS streams from all antennas to a compressed set. This parameter change allows maintaining channel estimate accuracy while reducing fronthaul bandwidth requirements.
2Reliability
If all antenna streams are transmitted over fronthaul for SRS, then beamforming performance is maintained, but fronthaul bandwidth requirements increase significantly
Solution Approach 1:
The system extracts only the essential information from all antenna streams by using spatial compression based on antenna covariance matrix decomposition. Instead of transmitting all 64 antenna streams, the system extracts and transmits a compressed representation that maintains beamforming performance while significantly reducing bandwidth requirements.
Solution Approach 2:
The system applies dimensionality reduction by transforming the full-rank antenna covariance matrix into a compressed representation with fewer dimensions. The spatial compression reduces the number of streams from the full antenna array to a reduced set, achieving bandwidth savings while preserving the essential beamforming characteristics.
3Quantity of substance
If SRS is used for beamforming weight calculation, then spatial compression is achieved, but interference environment is not accounted for
Solution Approach 1:
The system makes the spatial compression mechanism universal by applying the same antenna covariance matrix estimation and decomposition process to both SRS and PUSCH signals. This multi-functional approach allows the system to handle different signal types with a unified compression framework, improving beamforming accuracy while maintaining spatial compression benefits.
4Reliability
If additional streams are sent over fronthaul to mitigate UE mobility, then beamforming performance is maintained, but fronthaul bandwidth increases
Solution Approach 1:
The system dynamically adapts the spatial compression parameters based on channel conditions and mobility scenarios. By using antenna covariance matrix estimation that captures the current channel state, the system can adjust the compression ratio and number of streams to transmit, maintaining beamforming performance while minimizing fronthaul bandwidth requirements under varying mobility conditions.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
A method of spatial compression in a Radio Unit for MIMO, wherein the spatial compression is determined from a number of antennas to a number of streams in one or more stages, the method includes computing an antenna covariance matrix from a received signal over multiple resource elements; performing a matrix decomposition based on the computed covariance matrix to produce one or more compression vector; andgenerating a spatially-compressed signal by performing spatial compression on the matrix decomposition in one or more stages based.