MIMO Precoding Weights via Gram Matrix Inversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional MIMO systems face high computational complexity in calculating uplink receive combining/detection and downlink transmit precoding/beamforming parameters, leading to performance trade-offs and inefficiencies in processing time.
Innovation Solution
A method using the matrix inversion lemma to recursively calculate zero forcing beamforming weights by adding or removing user terminal channels, exploiting the Gram matrix structure to reduce computational complexity while maintaining full performance potential.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional MIMO systems use MMSE or MVU channel estimation techniques with matrix inversion, then channel estimation accuracy is improved, but computational complexity increases to cubic order in the product of number of antennas and pilot sequence length
Solution Approach 1:
The patent segments the channel estimation process by separating pilot signal processing from data signal processing, and by dividing the matrix operations into smaller manageable components. The Gram matrix is constructed from pilot signals first, then inverted once to be reused for multiple user detections, avoiding redundant cubic complexity operations for each user.
Solution Approach 2:
The patent performs preliminary channel estimation and Gram matrix inversion using pilot signals before actual data transmission. The precomputed Gram matrix inverse is then reused for multiple user detections, avoiding repeated cubic complexity matrix inversions and reducing overall computational burden while maintaining estimation accuracy.
2Device complexity
If polynomial expansion techniques are used to reduce ZF computational complexity, then implementation complexity is reduced, but performance trade-offs occur due to approximation errors
Solution Approach 1:
The patent enables the system to adaptively select between exact matrix inversion and polynomial expansion methods based on available computational resources and performance requirements. The Gram matrix structure allows the system to self-optimize by choosing the most appropriate computation method for current conditions, balancing complexity and performance automatically.
Data Source
AI summary
The present disclosure relates to a method and a network node for calculating transmitter precoding weights and receiver combining weights for a multiple input multiple output (MIMO) antenna system. Channel responses are estimated at the network node for user terminals accessing the network node on a carrier. Zero forcing beamforming weights are determined for the carrier by adding one of the user terminals at a time in a calculation of an inverse of a Gram matrix containing parameters of the channel responses.