Eigen-Beamforming Using Second-Order Statistics for SINR-Constrained MIMO
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current beamforming schemes in 5G massive MIMO systems face challenges with high computational complexity and sensitivity to estimation errors, particularly when using advanced algorithms that require large amounts of channel state information (CSI), which are often not available accurately.
Innovation Solution
A co-operating interference aware eigenbeamforming method that utilizes second-order statistics and minimizes total transmit power while maintaining a predefined signal-to-interference-plus-noise ratio (SINR) constraint, using beamforming weights optimized based on second-order statistics of radio channels between antenna arrays and terminal devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If advanced beamforming schemes using large amounts of CSI are employed, then beamforming performance is improved, but computational complexity increases
Solution Approach 1:
The patent extracts only the essential second-order statistical information (covariance matrices) from the full channel state information, discarding redundant details. This allows beamforming optimization to be performed using a condensed representation that requires significantly less computational processing while maintaining the critical interference characterization needed for performance.
Solution Approach 2:
The patent transforms the beamforming problem from using raw channel matrices to using second-order statistical parameters (covariance matrices). This parameter transformation changes the mathematical nature of the problem, enabling the use of eigenvalue decomposition on smaller, more manageable matrices that capture the essential spatial correlation properties without the full complexity of instantaneous channel realizations.
2Reliability
If advanced beamforming schemes requiring more CSI are used, then beamforming performance is improved, but sensitivity to estimation errors increases
Solution Approach 1:
The patent performs preliminary averaging to compute second-order statistics over time and frequency before the beamforming optimization is executed. This pre-processing step smooths out estimation errors and noise in the channel measurements, creating more robust covariance matrices that are less sensitive to instantaneous estimation inaccuracies while still capturing the underlying spatial correlation structure.
Solution Approach 2:
The patent introduces second-order statistical parameters (covariance matrices) as an intermediary between the raw channel measurements and the beamforming optimization. This intermediary representation acts as a buffer that filters out estimation errors while preserving the essential spatial correlation information needed for effective interference management and beamforming weight calculation.
3Device complexity
If second order statistics are used for beamforming optimization, then computational complexity is reduced, but information utilization is limited
Solution Approach 1:
The patent changes the parameter representation from full channel matrices to second-order statistical parameters, specifically covariance matrices. This transformation reduces the dimensionality of the problem from O(N²) elements in full channel matrices to O(N²) elements in covariance matrices that can be averaged over multiple samples, effectively compressing the information while preserving the essential spatial correlation properties needed for beamforming.
Solution Approach 2:
The patent creates a statistical copy (covariance matrix) of the channel characteristics that captures the essential spatial correlation structure without requiring the full instantaneous channel state. This copied representation can be reused across multiple beamforming optimizations and time instances, reducing the need for continuous full CSI acquisition and processing while maintaining effective interference management.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
Solutions to determine how to form beams when co-operating multiple-input multiple-output antenna panels are used are disclosed. Beams are formed according to sets of beamforming weights corresponding to minimized total transmit power from an antenna array of interest to served terminal devices. Optimization (302) to minimize total transmit power subject to a predefined constraint on a minimum allowable expected value of signal-to-interference-plus-noise ratio, SINR, is performed. In the optimization expected values of SINR for a terminal device are calculated based on based on the sets of beamforming weights for at least the served terminal devices, and weighted second order statistics, wherein a weighting factor for second order statistics between the served terminal devices and the antenna array is one and weighting factors for the other second order statistics less than one. The second order statistics are comprised in maintained (301) channel state information.