MIMO Demodulation via Noise De-correlation for Frequency-Selective Channels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wideband MIMO receivers face performance degradation due to non-flat fading and inter-symbol/inter-chip interference, as existing demodulation algorithms are not optimized for correlated noise and assume diagonal channel matrices, leading to suboptimal signal recovery.
Innovation Solution
Implementing ML or near-ML demodulation after noise de-correlation and successive interference cancellation, using an equivalent MIMO channel matrix and noise covariance matrix constructed from equalizer output, to improve signal recovery in frequency-selective channels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If linear equalizer is used to suppress ISI/ICI in wideband MIMO systems, then interference is reduced, but performance is far from optimal due to slicing function assumptions and correlated noise
Solution Approach 1:
The patent introduces an intermediary processing step between equalization and demodulation: noise de-correlation. The equalizer output with correlated noise is transformed into a de-correlated signal with uncorrelated noise, enabling subsequent ML demodulation to achieve optimal performance. This intermediary transformation resolves the contradiction by making the noise suitable for ML algorithms while preserving the interference suppression benefits of equalization.
Solution Approach 2:
The patent changes the statistical parameter of the noise from correlated to uncorrelated through the de-correlation transformation. By applying a transformation matrix to the equalizer output, the noise covariance matrix is converted to an identity matrix, changing the noise correlation structure to match the requirements of ML demodulation algorithms.
2Device complexity
If slicing function is applied after equalization, then device complexity is reduced, but performance degrades because it assumes diagonal channel matrix which is generally not true
Solution Approach 1:
The patent performs preliminary action by de-correlating the noise before demodulation. This preprocessing step transforms the correlated noise into uncorrelated noise and prepares the signal in a form suitable for ML demodulation, avoiding the performance degradation that would result from directly applying slicing to the equalizer output.
3Reliability
If ML demodulation is directly applied to equalized wideband signal, then signal recovery performance should be optimal, but performance degrades because noise components are highly correlated
Solution Approach 1:
The patent introduces noise de-correlation as an intermediary step that transforms the equalizer output to satisfy the i.i.d. noise assumption of ML demodulation algorithms. This intermediary transformation enables the direct application of ML demodulation to achieve optimal performance while maintaining algorithm compatibility.
Solution Approach 2:
The patent changes the noise correlation parameter from highly correlated to uncorrelated through de-correlation transformation. This parameter change makes the noise statistics compatible with ML demodulation requirements, enabling optimal signal recovery performance.
Data Source
AI summary
A method is described that enables maximum-likelihood (ML) demodulation for MIMO communications over frequency-selective channels. An equalizer is typically employed to suppress inter-symbol interference (ISI) due to frequency-selectiveness of the channel, but the noise of the equalizer output can be highly correlated such that standard ML-MIMO demodulations cannot directly apply. The method comprises first constructing equivalent post-equalization MIMO channel and noise covariance matrix, and then de-correlating the equalizer output so that ML or near-ML MIMO demodulations can be applied to improve the performance. Additionally, successive ISI cancellation (SIC) is described for further performance improvement.