Space-Frequency Equalization Covariance Estimation for Multi-Antenna Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional impairment covariance estimation methods in wireless communication systems with multiple antennas suffer from performance gaps, especially in scenarios with a large number of receiving antennas, leading to significant noise impact and reduced accuracy.
Innovation Solution
A method that transforms raw impairment covariance estimates into the time domain, applies a triangle window for masking, determines a threshold based on maximum delay spread, and then transforms back to the frequency domain to form an impairment covariance matrix, which is used for equalization weighting in wireless communication systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If multiple basic impairment covariance estimates are averaged to reduce noise impact, then noise reduction is achieved, but performance gap with ideal IRC increases when number of receiving antennas increases
Solution Approach 1:
The patent segments the covariance estimation process into distinct stages: initial covariance calculation, time-domain transformation, windowing operation, and thresholding. This segmentation allows selective processing of different frequency components, preserving useful signal information while removing noise, thereby resolving the contradiction between noise reduction and performance maintenance in multi-antenna scenarios
Solution Approach 2:
The patent transforms the covariance estimates from frequency domain to time domain, applies windowing and thresholding operations, then transforms back to frequency domain. This dimensional transformation enables effective noise filtering while preserving the structural properties needed for accurate IRC performance, especially when the number of receiving antennas is large
2Ease of manufacture
If non-parametric covariance estimation is used in IRC, then implementation simplicity is maintained, but accuracy deteriorates in severe multipath situations with multiple antennas
Solution Approach 1:
The patent modifies the covariance estimation process by introducing time-domain transformation and thresholding parameters. The threshold parameter ρth, determined based on noise power estimation and confidence level, dynamically adjusts the filtering strength. This parameter-based approach maintains implementation simplicity while significantly improving estimation accuracy in multipath scenarios with multiple antennas
Data Source
AI summary
The present invention relates to a method for generating impairment covariances for equalization in a receiver of a wireless communication system, as well as an covariance estimator, a receiver and a wireless communication system associated therewith, where the receiver is equipped with multiple antennas, and an impairment covariance matrix is used to calculate equalization weighting vector for signals transmitted by a transmitter, the method comprising the steps of: calculating a raw impairment covariance estimate between a first antenna and a second antenna of the multiple antennas on each of subcarriers allocated to the transmitter in frequency domain, transforming the raw impairment covariance estimates into time domain, masking the transformed impairment covariance estimates by a triangle window with a width defined by a maximum delay spread, determining a threshold based on the transformed impairment covariance estimates for the subcarriers and thresholding the masked impairment covariance estimates with the determined threshold, and transforming the thresholded impairment covariance estimates into frequency domain to form an impairment covariance matrix for calculating the equalization weighting vector. The impairment covariance estimation according to the present invention provides a performance improvement in equalization.