Neural Network Interference Whitener Selection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In wireless communication systems, particularly in 5G networks, interference whitening is challenging due to unknown signal-to-interference ratio (SIR) and interference distribution in the bandwidth part (BWP), which hinders effective mitigation of interference and noise.

Innovation Solution

A machine-learning-based system and method that employs neural networks to select an appropriate covariance matrix for interference whitening by extracting features from the received signal, allowing for adaptive selection based on SIR and interference distribution across resource blocks, enabling efficient interference mitigation without prior knowledge of these parameters.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional interference whitening methods are used, then interference mitigation can be achieved, but the system requires prior knowledge of SIR and interference distribution which is unknown in practice

Engineering Contradiction:

Improveinterference mitigation effectivenessVSAvoidadaptability to unknown interference conditions

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The neural network system performs self-learning by extracting features from received signals and automatically selecting appropriate covariance matrices without requiring external guidance or prior knowledge of interference characteristics. The system serves itself by adapting to unknown interference conditions through automated feature extraction and classification.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system uses feedback from signal processing to continuously improve its performance. By extracting features from received signals and using these features to select covariance matrices, the system creates a closed-loop feedback mechanism that adapts to changing interference conditions in real-time without requiring explicit knowledge of SIR or interference distribution.

Inventive Principle:

Principle #23Feedback

2Reliability

If multiple covariance matrices are selected based on feature extraction, then interference whitening performance improves, but system complexity increases

Engineering Contradiction:

Improveerror rate reductionVSAvoidneural network and feature extraction complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system segments the interference mitigation process into distinct manageable components: feature extraction module, neural network classification module, and covariance matrix selection module. This segmentation allows each component to be optimized independently and simplifies the overall system architecture while maintaining high performance.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system changes parameters dynamically by selecting different covariance matrices based on extracted signal features. Instead of using a fixed covariance matrix, the system adapts its parameters (covariance matrix selection) based on the extracted features, enabling optimal interference whitening for different interference conditions without requiring complex manual configuration.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20220318598A1Machine learning based interference whitener selection

Publication Date: 2022.10.06 SAMSUNG ELECTRONICS CO LTD

AI summary

A learning-based system and method for interference whitening method. In some embodiments, the method includes receiving a signal; extracting a first set of features from the signal; making a first selection, by a first neural network, based on the first set of features; and selecting a first covariance matrix, from a plurality of covariance matrices, based on the first selection.