Neumann Series Approximation for MU-MIMO Matrix Inversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Massive MIMO systems face high computational complexity due to large inverse matrix calculations in Zero-Forcing (ZF) based methods, leading to processing delays and performance degradation from approximation errors in hardware implementation.
Innovation Solution
The method calculates the Signal-to-Interference Ratio (SIR) caused by inverse matrix approximation errors and adaptively selects the truncation order of the Neumann Series (NS) and the number of multiplexed User Equipment (UEs), modifying Channel Quality Indicators (CQI) and Modulation and Coding Schemes (MCS) to optimize system performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If exact inverse matrix calculation is used in ZF based detection or precoding, then detection accuracy and precoding performance are improved, but computational complexity increases significantly
Solution Approach 1:
The patent uses a low-complexity approximate inverse matrix calculation method instead of exact inversion. The Neumann series approximation with truncation provides a computationally feasible solution that sacrifices minimal accuracy for significant complexity reduction, enabling hardware implementation in massive MIMO systems.
Solution Approach 2:
The patent introduces a truncation order parameter N to control the approximation level of the Neumann series. By adjusting N, the system can balance between computational complexity and detection accuracy/precoding performance, allowing adaptive optimization based on channel conditions and system requirements.
2Reliability
If exact inverse matrix calculation is used, then system performance is improved, but processing delay increases
Solution Approach 1:
The approximate inverse matrix method dramatically reduces computation time compared to exact inversion, enabling real-time processing in massive MIMO systems while maintaining acceptable performance through the Neumann series approximation with sufficient truncation order.
3Device complexity
If Neumann Series truncation is used to reduce complexity, then computational complexity is reduced, but approximation error increases
Solution Approach 1:
The truncation order N is optimized to achieve the desired balance between complexity and accuracy. The patent provides methods to determine appropriate N values that ensure approximation errors remain below acceptable thresholds while maintaining computational feasibility for hardware implementation.
Solution Approach 2:
The patent incorporates feedback mechanisms to monitor system performance and adjust the truncation order N dynamically. This allows the system to maintain optimal performance by adapting the approximation level based on actual channel conditions and performance requirements.
4Measurement precision
If higher truncation order N is used, then approximation accuracy is improved, but computation resource increases
Solution Approach 1:
The patent optimizes the truncation order N to achieve the minimum required approximation accuracy while minimizing computation resources. By carefully selecting N based on system requirements and channel conditions, the patent avoids unnecessary computational overhead while ensuring sufficient approximation quality for reliable detection and precoding.
Data Source
AI summary
This invention presents methods for signal detection and transmission in MU-MIMO wireless communication systems, for inverse matrix approximation error calculation, for adaptively selecting the number of multiplexed UEs in a MU-MIMO group, for adaptively choosing a modulation and channel coding scheme appropriate for the quality of MU-MIMO channels with the approximation error of matrix inverse being incorporated.